{ "meta": { "title": "GTM World Model", "version": "3.3", "built": "2026-05-31", "language": "en-US", "purpose": "A comprehensive GTM decision map connecting 219 vocabulary terms, six tier/substrate/regime views, 17 extended layers and ten business-kind templates. Separates accounting, mathematical models, prediction, causal effects and qualified operating policies. The executable seed covers a limited MRR/retention scenario; other workflows remain specifications until local measurement and qualification.", "relationship_to_vocabulary": "The canonical glossary supplies stable term IDs; this model adds tier views, states, typed relationships, equations and policy contracts. Descriptive model names are not all validated glossary joins; validation reports unresolved names explicitly.", "core_thesis": "Stocks and flows require consistent units and timing. Conversion and economic summaries are conditional on buyer mix, business kind, policy and regime. Causal decisions compare feasible policies against alternatives or no action through incremental contribution; observations and plausible mechanisms alone do not identify their effects.", "how_to_read": "Read each expression with its assumptions, units, population, outcome horizon and evidence boundary. Preserve stable IDs. Curriculum C0-C8 and public editorial C1-C7 hubs are distinct namespaces. A measurable construct is not automatically measured, identified or deployment-qualified.", "changelog": [ "v1.0: three tiers + substrate over vocabulary v2 (195 terms); state vars, typed edges, equations, agent interface (LangChain worked example).", "v1.1: aligned to vocabulary v2.1 (219 terms). Added the Sense-Reason-Act-Learn cycle as the atomic unit of Tier 1; added agent_properties (inverted cost profile + compounding failure); added the FinOps->CAC edge to the economics tier; added coefficient-compression as a distributional effect on r_i; added governance instruments (task-to-role matrix, escalation paths, data-quality gates, audit logging) to the agent interface; added the GTM-AI maturity ladder and phased-rollout as adoption structure. Sources: Apollo, Poexis, Wyzard (+ LangChain).", "v3.2: C6+ Agentic-GTM formal extension. Formalizes eight prose theses (T8-T28 cluster) into 10 new/refined state variables (alpha, G, gamma, p, R, C_a, lambda, theta, kappa, B_agent) and 9 equations (E-A1..E-A9) scoped to Tier 1 (Execution) and the Operations & Systems substrate; does NOT revise Tiers 0/2/3/5 or the 17 base equations. Adds theses T8.1, T29 (governance maturity is a precondition for autonomy to lower cost, not a safety tax — with a worked sign-flip example), and T30 (agent adoption is a category externality no firm-level governance internalizes). Extends the measurement-gap register with 6 agentic parameters, all tagged unmeasurable_hypothesis. Every added edge carries an epistemic_type; the model refuses to assert the agentic-CAC sign (E-A8) until a public holdout exists.", "v3.3 (2026-10-05): research consistency, evidence boundaries, workflow qualification and nine-cluster roadmap; corrected public delivery and version controls." ], "revision_note": "v3.3 reconciles the 31-item research correction register, qualifies the nine agent equations, corrects stock/value, funnel, PLG and LTV algebra, and adds decision/evidence/measurement contracts plus nine cluster improvement plans. No proposed experiment is claimed as run.", "epistemic_honesty": "No universal retention dominance, funnel bottleneck ROI, 95% pre-contact value split, brand decay, PMF functional form, agent-governance threshold or hybrid superiority is asserted. Unmeasured coefficients remain unknown. Historical construction claims are superseded by the current equation, thesis and evidence records.", "vocabulary_terms_count": 219, "prior_version": "3.2; durable history: git log -- src/gtm/data/gtm_world_model.json", "posture_v3": "Preserve broad coverage with explicit observability, evidence and qualification states. Unknowns stay unknown; completeness is a decision map, not evidence that all its mechanisms are fitted or causal.", "runnable_seed": "gtm_engine_slice.py executes MRR flows and a qualified retention/LTV scenario with explicit assumptions and refusals; it is not an end-to-end GTM optimizer.", "v3_1_note": "Historical v3.1 added eight generalizations and ten business-kind templates. v3.3 corrects residual universal equilibrium, licensing and one-gate assumptions while preserving template IDs.", "extended": "2026-10-05", "v3_2_note": "v3.2 is ADDITIVE, not a rebuild. It closes the gap the v3.1 posture left open: the agentic claims (T8-T28) were carried as prose theses without state variables or equations of their own. v3.2 gives them the same treatment as the rest of the model — typed edges, epistemic tags, and a measurement-gap entry per unmeasured parameter — so an agentic GTM system can read the execution/governance tier as structured constraints, not narrative. Richer execution tier, same epistemic ceiling: the six new gap-register rows mean most added coefficients remain unmeasurable_hypothesis and may inform risk-awareness but never a forward agent action.", "agentic_extension_c6plus": "Legacy C6+ alias refers to the agentic extension, canonical curriculum C8. E-A1..E-A9 retain stable IDs with v3.3 policy/outcome/cost contracts. The former fatigue parabola and governance sign-flip are retired as operating claims; the executable retention/MRR seed remains limited.", "v3_3_note": "Research window 2026-04-05 through 2026-10-05: 65 current sources plus 12 older foundations. Reviewed evidence has heterogeneous access depth and scope. See research_update_v3_3, evidence_register, qualification_protocol, cluster_roadmap and proposed_studies.", "updated": "2026-10-05", "term_count": 219, "api_version": "1.0", "endpoint": "https://shalvisingh.com/api/gtm/model" }, "epistemic_types": { "identity": "An accounting equality or definition, exact only under its declared basis and mutually exclusive categories. It does not identify a behavioral response or a causal counterfactual.", "causal_regime": "A context-dependent causal mechanism or proposed structural model. Use intervention evidence and identification conditions to estimate an effect within support; the tag alone is not evidence that the effect has been identified.", "correlational": "Two variables move together, but the link is confounded (usually by product-market fit or the macro cycle) or directionally ambiguous. Must NOT be used for high-confidence counterfactuals. Useful for regime-detection and prior-setting, not point prediction. Example: high-NRR companies also have efficient CAC.", "latent_multiplier": "An unobserved construct or proposed modifier. Multiplicative form is a hypothesis, not guaranteed by latent status. Independent pre-outcome measurement and validation are needed to avoid circularity.", "estimator": "An estimated summary under explicit segmentation, timing, accounting, cost and statistical assumptions. LTV forecasts require a survival/contribution model; CAC ratios require complete eligible costs and acquired-customer denominators. No universal LTV correction factor follows from heterogeneity.", "conditional_coefficient": "A conditional observational probability or coefficient, useful within a defined population and information set. P(win|touch,X) is not E[Y(touch)-Y(no touch)|X] without identification; conditioning alone does not remove confounding.", "model_identity": "A mathematical result conditional on explicitly stated structural assumptions and domain. Exact algebra within a toy model does not establish that a company obeys it.", "accounting_definition": "A convention-defined ratio or measurement. Definitions can be exact on a fixed denominator and scope while interpretation and forecasts remain assumption-dependent.", "normative_constraint": "A declared policy, authority, risk or jurisdiction-dependent requirement. Enforce as a hard predicate where applicable; it is neither an accounting identity nor a fitted behavioral law.", "hypothesis": "A proposed, unvalidated mechanism or parameterization. Retain as a scenario or research target; never silently treat it as an observed or identified input." }, "tiers": { "tier_3_strategy": { "name": "Strategy / Configuration tier (slow-moving, human-owned)", "tempo": "quarters to years", "owned_by": [ "Founders", "Product Marketing", "GTM leadership" ], "what_it_decides": "WHO, WHAT, HOW and WHERE to sell under product, buyer and economic constraints. These are explicit policy choices rather than universally forced ACV moves.", "is_input_to": "The agent reads these as configuration; it does not get to change them. In LangChain's system, the ICP, the motion, and the warm/cold playbook are encoded in the agent's SKILL and do-not-send rules — decided by humans, handed to the agent.", "vocabulary_terms": [ "Go-to-Market (GTM) Strategy", "GTM Strategy vs Tactics", "GTM as a System", "Product-Market Fit", "Go-to-Market Fit", "Minimum Viable GTM", "90-Day Plan", "Launch", "Product Marketing", "TAM / SAM / SOM", "Market Opportunity", "Market Segmentation", "STP", "ICP", "Buyer Persona", "Buying Committee", "Firmographic Data", "Technographic Data", "Positioning", "Value Proposition", "Competitive Positioning", "Category Design", "Messaging Framework", "Proof Points", "Why Now", "Jobs to be Done", "Pricing & Packaging", "Freemium", "Free Trial", "GTM Motion", "Product-Led Growth", "Sales-Led Growth", "Marketing-Led / Inbound-Led", "Outbound-Led", "Channel/Partner-Led", "Community-Led", "Ecosystem-Led", "Content-Led", "Paid-Led", "Event-Led", "Hybrid Motion", "Land and Expand", "Bottom-Up Adoption", "Self-Serve", "4 Ps", "AIDA", "GTM Operating System", "Sales Methodology", "Sales Qualification Framework", "BANT", "CHAMP", "MEDDIC", "SPICED", "The Challenger Sale", "Marketing Strategy", "ABM", "Named Accounts", "GTM AI Maturity Ladder", "Agentic ABM" ], "key_outputs_to_lower_tiers": [ "ICP definition", "motion selection", "pricing", "messaging/playbook", "qualification criteria" ] }, "tier_2_economics": { "name": "Economics: ledgers, defined metrics and model-dependent forecasts", "tempo": "continuous; reviewed monthly/quarterly", "owned_by": [ "RevOps", "Finance", "GTM leadership" ], "what_it_decides": "Contribution, cash viability and feasible allocation on declared business-kind/cohort/period bases. Accounting equalities and defined ratios can be exact within conventions; predictive and causal interpretations require assumptions and evidence.", "is_input_to": "Sets the objective function and the guardrails the agent optimizes against. LangChain's agent implicitly serves this tier: by raising lead->opportunity conversion it improves the funnel term that feeds CAC efficiency.", "vocabulary_terms": [ "ARR", "MRR", "ACV", "TCV", "ARPA", "ASP", "Average Deal Size", "Bookings", "Billings", "CAC", "CAC Payback", "CPL", "LTV", "LTV:CAC", "Magic Number", "Rule of 40", "NRR", "GRR", "Churn", "Retention", "Expansion Revenue", "Cohort Analysis", "Sales Velocity", "Sales Cycle", "Win Rate", "Closed-Won / Closed-Lost", "Pipeline Coverage", "Forecast Accuracy", "Sales Forecasting", "Deal Health", "NPS", "Time to Value", "Stickiness", "North Star Metric", "Leading vs Lagging Indicators", "Vanity Metrics", "Attribution", "Self-Reported Attribution", "Dark Funnel", "Booking Rate", "Win/Loss Analysis", "Bowtie Model" ], "key_constraints_imposed": [ "Expected incremental contribution versus cost and downside", "Cash/runway and payment timing", "Delivery/review capacity, quality and service promises", "Applicable authority and compliance predicates" ], "the_one_identity": { "expression": "MRR_t = MRR_{t-1} + New_t + Expansion_t - Contraction_t - Churn_t + Other_defined_movements_t", "status": "exact recurring-value ledger on a declared convention", "leaks_when": "The ledger fails to reconcile when movements, boundaries, currency or conventions are inconsistent or incomplete. A difference from recognized revenue/cash alone is not a falsification; use an explicit bridge.", "provenance": "All four construction reports converge; Component A." }, "estimators": [ { "metric": "LTV", "naive_form": "ARPA*margin/churn", "hidden_assumption": "Stable cohort, explicit payment timing, survival/usage response, cost, discounting and censoring assumptions.", "corrected_form": "LTV=sum_{t>=0} E[net_contribution_t*1(active_t)]/(1+d)^t - nonrecurring_costs, with explicit first_payment; sBG q~Beta(alpha,beta), S(t)=B(alpha,beta+t)/B(alpha,beta)", "correction_magnitude": "For alpha=2.5,beta=6, mean initial churn is 2.5/8.5=29.41%. Expected paid periods including t=0 are 5; future periods excluding t=0 are 4. A geometric model matched to initial churn gives 3.4 including t=0 or 2.4 future periods. This example does not support a universal LTV overstatement or an invented 10%-churn comparator.", "provenance": "Fader-Hardie 2007 sBG / 2018 BdW (established); worked numbers from construction reports A" }, { "metric": "NRR", "hidden_assumption": "intra-account expansion only; ignores cross-sell attribution; no standard definition (Fastly published DBNR=126% and NRR=93% same period)", "note": "endogenous -- inflated by discounting and forced multi-year renewals; cannot be used to 'prove' Phi (the T7 circularity)" }, { "metric": "GRR", "hidden_assumption": "Defined beginning revenue cohort and contraction/churn convention; historical ratio does not require a stationary hazard, but projecting it does." }, { "metric": "CAC Payback", "hidden_assumption": "no expansion, no downgrades, constant gross margin; ignores TVM; numerator scope varies across Bessemer/KeyBanc/OpenView" }, { "metric": "Magic Number", "hidden_assumption": "fixed one-quarter lag between S&M spend and ARR growth" }, { "metric": "Rule of 40 / Rule of X", "hidden_assumption": "Practitioner health/valuation heuristic, not an identified operating objective. Growth multiplier and regime mapping are scenarios." } ], "provenance": "Construction reports A (all four converge); adversarial audit finding #1 (the single most damaging) + epistemic audit." }, "tier_1_execution": { "name": "Execution / Behavioral tier (fast, increasingly agent-owned)", "tempo": "per-lead, per-account, daily/weekly", "owned_by": [ "SDR/BDR", "AE", "CS/Deployed Eng", "Demand Gen", "GTM Engineer", "and increasingly: the agent" ], "what_it_decides": "The actual per-lead and per-account loop: detect signal -> decide whether to act -> gather context -> qualify -> draft/personalize -> route -> hand off -> follow up -> expand/retain. This is where the behavioral coefficients (conversion rates, response, reply rates) actually get produced, and where they DRIFT.", "is_the_layer_the_agent_runs": true, "langchain_mapping": "This entire tier is what LangChain's GTM agent automates: trigger on Salesforce lead -> do-not-send checks -> research (CRM/Gong/LinkedIn/Exa) -> load warm/cold skill -> draft -> human approval in Slack -> queue follow-ups; plus Monday account-intelligence for expansion/renewal/churn signals.", "vocabulary_terms": [ "Funnel", "TOFU / MOFU / BOFU", "Critical User Journey", "Lead", "Lead Lifecycle", "MQL", "MQA", "SAL", "SQL", "CQL", "PQL", "Opportunity", "Lead Handoff", "Lead Scoring", "Predictive Lead Scoring", "Demand Generation", "Demand Creation", "Demand Capture", "Lead Generation", "Nurture", "Gated Content", "Content Syndication", "Thought Leadership", "Inbound", "Speed-to-Lead", "Intent Data", "Intent Surge", "Buying Signal", "Buying Intent", "Signal-to-Action", "SEO", "PPC", "Retargeting", "Co-Marketing", "Webinar", "VAR", "Partner Enablement", "Marketplace", "Channel Strategy", "SDR / BDR", "Account Executive", "Account Manager", "Customer Success", "Sales Engineer", "Discovery", "Proof of Concept", "Quota", "Sales Stages", "Sales Sequence", "Cold Email", "Email Deliverability", "Domain Warm-Up", "Sales Enablement", "Sales Playbook", "Battlecard", "Case Study", "Multi-Threading", "Mutual Action Plan", "Territory Management", "Account Expansion", "Upsell", "Cross-Sell", "Advocacy", "Word-of-Mouth", "Onboarding", "Activation", "Feature Adoption", "Product Analytics", "Economic Buyer", "Champion", "Identify Pain", "Decision Criteria", "Decision Process", "Critical Event", "Pipeline", "Field Sales", "Inside Sales", "Deal Desk", "Sense-Reason-Act-Learn", "Next-Best-Action", "Signal-to-Revenue", "Signal Taxonomy", "Coefficient Compression", "Outer Flywheel" ] }, "cross_cutting_substrate": { "name": "Operations & Systems substrate (the rails all tiers run on)", "tempo": "always-on", "owned_by": [ "RevOps", "Marketing Ops", "GTM Engineer" ], "what_it_is": "The data, tooling, routing, and governance that let the other tiers function. Not a tier in the causal stack — it is the substrate. The AGENT itself is the newest member of this substrate.", "vocabulary_terms": [ "Revenue Operations", "Marketing Operations", "GTM Engineer", "Data Enrichment", "Lead Routing", "Stage Definitions", "Rules of Engagement", "SLA", "Operating Cadence", "QBR", "Sales-Marketing Alignment", "RACI / DRI", "CRM", "Marketing Automation Platform", "Sales Engagement Platform", "Clay", "Business Intelligence", "PLG CRM", "Agentic GTM", "GTM OS", "Task-to-Role Matrix", "Human-in-the-Loop", "Escalation Path", "Guardrails", "Data Quality Gate", "Audit Logging", "Compounding Failure", "Agent Cost Profile", "Agent FinOps", "Context Engineering", "Workflow Orchestration", "Subagent", "Learning Loop", "Phased Rollout" ], "langchain_mapping": "LangChain's agent IS a substrate component: it sits on Salesforce/Gong/BigQuery/Gmail (systems of record), enforces do-not-send Rules of Engagement, honors a 48h SLA, and its memory store is a new piece of RevOps infrastructure. Their lesson 'connect to systems of record from the start' is a statement about this substrate." }, "tier_0_prefunnel": { "name": "Pre-funnel buyer state and brand memory", "tempo": "slow (brand: quarters-years) to medium (buyer-state: weeks-months); mostly INVISIBLE to the seller", "owned_by": [ "Marketing/Brand", "Demand-gen", "Product (for PLG-led discovery)" ], "what_it_decides": "Prior need, experience, shortlist, roles, proof and constraints influence consideration and contact selection. Survey in-market/shortlist shares motivate measurement without identifying the share of causal value created before contact.", "is_input_to": "Defines pre-action covariates and hypothesized brand-to-consideration mechanisms; intervention effects remain separate.", "vocabulary_terms": [ "Brand Awareness", "Demand Creation", "Demand Capture", "Category Entry Points", "Mental Availability", "Share of Voice", "95-5 Rule", "Dark Funnel", "Buying Group", "Buying Committee", "Day-1 Shortlist", "Intent Data", "Signal-Based Selling", "Out-of-Market", "In-Market", "Thought Leadership" ], "key_constraints_imposed": [ "Preserve pre-action buyer-state observations and unknown values.", "Freeze eligibility and compare complete policy outcomes including no action where appropriate.", "Do not translate survey prevalence into seller-effect bounds or fixed headcount decay." ], "provenance": "Construction reports D+E (all four converge); 6sense BER 2024/2025; LinkedIn B2B Institute 95-5; Ehrenberg-Bass CEPs; adversarial audit blind-spot #4 (highest severity).", "evidence_caveat": "Retrospective commercial surveys motivate buyer-state measurement. Effects need randomization or justified exchangeability, positivity, consistency and interference handling; conditional effects are averages, not identifiable individual effects." }, "cross_cutting_regime": { "name": "Capital context and declared strategic objectives", "tempo": "slow (shifts with the rate/capital cycle over quarters-years)", "owned_by": [ "Board", "CFO", "Investors (exogenous to the firm)" ], "what_it_decides": "Capital conditions can constrain decisions and influence valuation. Operating objective weights are declared preferences, not causally determined by a fitted macro scalar in this model.", "definition": "Scalar index reducing a macro vector (10yr Treasury, Cloud forward multiples, VC deployment pace, median burn multiple) to a regime score in {ZIRP, tightening, high-rate}.", "functional_form": "U=w(context,mandate) dot declared outcomes, subject to hard constraints. Weights are declared scenarios/preferences, not fitted macro causal coefficients.", "measured_values": { "10yr_Treasury": "2021 ~1.45% (ZIRP) -> 2023 ~3.96% -> 2025 ~4.29% (Macrotrends)", "Cloud_100_multiple": "~34x peak 2021 (inferred) -> 26x 2023 -> 20x 2025 (Bessemer, -41% from peak)", "M_growth_weight": "Unfit scenario; no observed firm-level objective mapping supplied.", "implied_w_g/w_f": "Unfit scenario; specify units and mandate before comparison." }, "resolves": "Separate market multiples/rates, firm funding/runway/covenants and management objectives. Compare scenarios; no fixed rate-to-payback or rate-to-growth-weight mapping is identified.", "falsification": "A segment (e.g. bootstrapped profitable vertical SaaS) whose growth/FCF weighting did NOT shift 2021->2025 despite the rate spike would show Psi is not universal but a VC-backed artifact. No public panel of firm-level objective weights exists; specified-but-unrun.", "provenance": "Construction reports C (all four converge); Bessemer Rule-of-X 2024; Macrotrends; adversarial audit blind-spot #1.", "evidence_caveat": "Rule-of-X is a practitioner valuation heuristic. Units, scaling and firm mandate must be declared; causal operating policies do not follow from it." } }, "state_variables": [ { "symbol": "C", "name": "Logo and recurring-value stocks (separate units)", "tier": "economics", "type": "stock", "vocab": [ "ARR", "MRR" ], "note": "Customer counts and recurring monetary value are separate stocks; no shared expansion term for logos.", "canonical_symbol": "L_logo/M_recurring" }, { "symbol": "a", "name": "Acquisition inflow rate", "tier": "execution", "type": "flow_in", "vocab": [ "Bookings", "Win Rate", "Lead Generation" ], "note": "Acquisition inflow must state whether it counts logos or recurring currency per time.", "canonical_symbol": "a_logo/a_value" }, { "symbol": "delta", "name": "Churn rate", "tier": "economics", "type": "flow_out", "vocab": [ "Churn", "GRR" ], "note": "Logo loss and value contraction/churn require separate denominators and consistent periods.", "canonical_symbol": "delta_logo/delta_value" }, { "symbol": "g", "name": "Expansion rate", "tier": "execution", "type": "flow_modifier", "vocab": [ "Expansion Revenue", "Account Expansion", "Upsell", "Cross-Sell", "Land and Expand" ], "note": "Expansion changes recurring value; net value retention is defined on a fixed beginning cohort.", "canonical_symbol": "g_expansion" }, { "symbol": "V", "name": "Top-of-funnel volume", "tier": "execution", "type": "flow_source", "vocab": [ "Demand Generation", "Lead", "Intent Data" ], "note": "Raw demand entering the funnel.", "canonical_symbol": "V_eligible" }, { "symbol": "r_i", "name": "Stage conversion rates", "tier": "execution", "type": "coefficient", "vocab": [ "MQL", "SAL", "SQL", "Opportunity", "Lead Scoring", "Speed-to-Lead" ], "note": "Conditional stage rates depend on case mix and stage predicates. Agent mean/variance improvements require evidence; neither is universal.", "canonical_symbol": "r_stage" }, { "symbol": "p_bar", "name": "Average price / ACV", "tier": "strategy", "type": "coefficient", "vocab": [ "ACV", "ARPA", "ASP", "Pricing & Packaging" ], "note": "Price is one input to motion feasibility alongside contribution, cash, capacity and buyer requirements.", "canonical_symbol": "price_avg" }, { "symbol": "m", "name": "Gross margin", "tier": "economics", "type": "coefficient", "vocab": [ "CAC Payback", "LTV" ], "note": "Gross margin is a defined gross-profit share on its accounting basis. Contribution/net value may require additional variable delivery, service, return and failure costs; gross margin alone does not cover all of them.", "canonical_symbol": "margin_gross" }, { "symbol": "S", "name": "Sales+marketing spend", "tier": "economics", "type": "control_input", "vocab": [ "CAC", "Magic Number" ], "note": "Declare complete acquisition costs and denominator; agent costs may already be included.", "canonical_symbol": "spend_acquisition" }, { "symbol": "x", "name": "Market penetration / saturation", "tier": "strategy", "type": "state", "vocab": [ "TAM / SAM / SOM", "Market Opportunity" ], "note": "Specify counts versus fractions; reciprocal CAC saturation is an unfit response scenario.", "canonical_symbol": "penetration_x" }, { "symbol": "Phi", "name": "Product-market fit (latent)", "tier": "strategy", "type": "latent_multiplier", "vocab": [ "Product-Market Fit", "North Star Metric", "Retention", "NPS" ], "note": "Fit is multidimensional and latent; multiplicative response is a hypothesis, not its definition.", "canonical_symbol": "Phi_fit" }, { "symbol": "k(tau)", "name": "Lag/memory kernel", "tier": "execution", "type": "function", "vocab": [ "Dark Funnel", "Attribution", "Demand Creation", "Thought Leadership" ], "note": "Distributed-lag kernel. Why last-touch attribution structurally misattributes brand/demand-creation.", "canonical_symbol": "kernel_lag" }, { "symbol": "Psi", "name": "Macro regime scalar", "tier": "regime", "type": "exogenous_index", "vocab": [ "Rule of 40", "Burn Multiple" ], "note": "Macro conditions and management objective preferences are distinct; weights are unfit scenarios.", "canonical_symbol": "Psi_macro" }, { "symbol": "B", "name": "Buyer-state vector", "tier": "prefunnel", "type": "latent_state", "vocab": [ "Buying Group", "Day-1 Shortlist", "In-Market", "Intent Data" ], "note": "Pre-action buyer roles, need, shortlist and constraints help prediction but do not themselves identify effects.", "canonical_symbol": "B_buyer" }, { "symbol": "B_r", "name": "Brand stock", "tier": "prefunnel", "type": "stock", "vocab": [ "Brand Awareness", "Mental Availability", "Category Entry Points", "Share of Voice" ], "note": "Brand memory parameters are unknown for the local segment; linear stock and exponential lag are equivalent.", "canonical_symbol": "B_brand" }, { "symbol": "S", "name": "Switching-cost moat", "tier": "economics", "type": "latent_modifier", "vocab": [ "Switching Costs", "Vendor Lock-in" ], "note": "High GRR plus low satisfaction does not uniquely identify switching costs or a piecewise causal law.", "canonical_symbol": "S_switching" }, { "symbol": "k", "name": "PLG viral coefficient + loop state", "tier": "prefunnel", "type": "loop_state", "vocab": [ "Viral Coefficient", "Product-Led Growth", "Free-to-Paid Conversion" ], "note": "Separate referral generations, cumulative unique referrals, active users and paid customers; no universal B2B k<1 bound.", "canonical_symbol": "k_referral" }, { "symbol": "alpha", "name": "Autonomy level of an agentic execution chain", "tier": "execution", "type": "policy_summary", "vocab": [ "Agentic GTM", "Human-in-the-Loop" ], "note": "Descriptive fraction only. A complete policy conditions execute/review/repair/defer/stop on task, state, authority and capacity.", "canonical_symbol": "alpha_autonomy" }, { "symbol": "G", "name": "Governance / HITL maturity index", "tier": "substrate", "type": "exogenous_index", "vocab": [ "Human-in-the-Loop", "Governance" ], "note": "Summary descriptive index. Detection, enforcement, correction, provenance, containment and recovery need separate measured predicates.", "canonical_symbol": "G_governance_summary" }, { "symbol": "gamma", "name": "Effective error catch-rate of human review", "tier": "substrate", "type": "function", "vocab": [ "Human-in-the-Loop" ], "note": "Catch rate is criterion-specific; also measure false rejection, correction and reviewer-introduced defects.", "canonical_symbol": "gamma_catch" }, { "symbol": "p_step", "name": "Effective per-step error rate of an agent-executed chain", "tier": "execution", "type": "function", "vocab": [ "Agentic GTM" ], "note": "Residual shipped defects conditional on task, data, policy and configuration; no universal scalar error input.", "canonical_symbol": "p_step" }, { "symbol": "R_chain", "name": "Chain reliability across n sequential steps", "tier": "execution", "type": "function", "vocab": [ "Agentic GTM" ], "note": "Use conditional chain rule; independent homogeneous power is a toy case. Accepted outcome may include repairs.", "canonical_symbol": "R_chain" }, { "symbol": "C_a", "name": "Agent execution spend", "tier": "economics", "type": "flow_out", "vocab": [ "CAC", "Agentic GTM" ], "note": "All eligible attempt, retry, review, tool/data and maintenance cost with acquisition/delivery allocation stated.", "canonical_symbol": "cost_agent" }, { "symbol": "lambda_hitl", "name": "HITL-as-data-collection rate", "tier": "execution", "type": "coefficient", "vocab": [ "Human-in-the-Loop" ], "note": "Learning effectiveness must be observed through valid outcome linkage, implementation and held-out evaluation; unfit here.", "canonical_symbol": "lambda_learning" }, { "symbol": "theta", "name": "Category attention/congestion scenario", "tier": "prefunnel", "type": "loop_state", "vocab": [ "Brand", "Demand Generation" ], "note": "Attention/congestion hypothesis, distinct from forecast error. Positive decay bounds response to bounded input; constant adoption additionally gives a steady state. Net externality sign remains unknown.", "canonical_symbol": "theta_attention" }, { "symbol": "kappa", "name": "Aggregate agent-adoption share across the competitive category", "tier": "regime", "type": "exogenous_index", "vocab": [ "Agentic GTM" ], "note": "Category adoption is distinct from firm autonomy; interference and platform/rival response matter.", "canonical_symbol": "kappa_adoption" }, { "symbol": "B_agent", "name": "Agent-evaluator presence within a buying committee", "tier": "prefunnel", "type": "state", "vocab": [ "Buying Committee", "Buyer Persona" ], "note": "Role × task × phase matrix with platform/interface/version, fidelity, verification, authority and observed fulfillment; aggregate share optional.", "canonical_symbol": "B_AI" } ], "equations": [ { "id": "conservation_law", "name": "Separate logo and recurring-value stock balances", "epistemic_type": "model_identity", "expression": "dL/dt = a_logo - delta_logo*L; dM/dt = a_value - (delta_value-g_expansion)*M", "steady_state": "L*=a_logo/delta_logo for delta_logo>0; M*=a_value/(delta_value-g_expansion) is attracting only when delta_value>g_expansion", "reads_as": "L counts active logos; M measures recurring monetary value. New-logo inflow has units logos/time; new-value inflow has units currency/time. Expansion changes value without adding logos.", "implication": "With n=delta_value-g_expansion, M(T)=M0*exp(-n*T)+a_value*(1-exp(-n*T))/n for n!=0; at n=0, M(T)=M0+a_value*T. Near zero net attrition the long-run equilibrium is not a finite-horizon ROI comparison. Retention versus acquisition depends on feasible responses, costs, horizon and quality.", "vocab": [ "ARR", "Churn", "NRR", "GRR", "Expansion Revenue", "Retention" ], "agent_relevance": "Compare measured acquisition, retention and expansion policies using incremental net contribution on a common horizon. A denominator position alone does not establish the highest-return investment.", "latex": "\\frac{dL}{dt}=a_{\\mathrm{logo}}-\\delta_{\\mathrm{logo}}L,\\qquad \\frac{dM}{dt}=a_{\\mathrm{value}}-(\\delta_{\\mathrm{value}}-g_{\\mathrm{exp}})M", "steady_state_latex": "L^*=\\frac{a_{\\mathrm{logo}}}{\\delta_{\\mathrm{logo}}},\\qquad M^*=\\frac{a_{\\mathrm{value}}}{\\delta_{\\mathrm{value}}-g_{\\mathrm{exp}}}\\quad(\\delta_{\\mathrm{value}}>g_{\\mathrm{exp}})", "assumptions": [ "Continuous constant-rate toy model; rates and acquisition nonnegative; no capacity, saturation or price changes.", "At n=0,a_value=0 every constant M is an equilibrium, none attracting; at n<0,a_value=0, M=0 is an unstable equilibrium.", "A discrete ledger is exact on its basis; proportional constant-rate ODE behavior is an assumption." ], "units": { "L": "logos", "M": "currency of recurring value", "a_logo": "logos/time", "a_value": "currency/time", "delta_logo,delta_value,g_expansion": "1/time" }, "revision": "3.3", "curriculum_cluster": "C5", "evidence_ids": [ "R01", "R02", "R03", "R04", "R05", "R06", "R08", "R22", "R25", "R50", "R51", "R52", "R53", "R55", "R56", "R57", "R58", "R61", "R62", "R63" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "not_applicable_to_accounting_or_model_algebra", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "funnel_operator", "name": "The funnel as a log-linear operator", "epistemic_type": "identity", "expression": "R = V * (product of r_i) * p_bar <=> ln R = ln V + sum(ln r_i) + ln p_bar", "reads_as": "Cohort value on the declared booking/price basis equals eligible volume times nested stage conversion fractions times mean final-outcome value; logs are additive only for positive inputs. It is not an independent causal lever decomposition.", "implication": "All rates have unit elasticity for equal proportional improvements, holding other inputs fixed. For equal absolute rate changes, dR/dr_i=R/r_i, but investment return also depends on dr_i/dspend, downstream quality/rates, price, capacity and time. A low rate can be a useful filter; select the best feasible incremental contribution intervention, not min(r_i).", "vocab": [ "Funnel", "MQL", "SQL", "Opportunity", "Win Rate", "TOFU / MOFU / BOFU" ], "agent_relevance": "Reported opportunity gains are scoped observations, not proof that the lowest-rate stage is the best investment. Freeze eligibility and stage predicates, compare whole-cohort outcomes and measure feasible resource response.", "latex": "R=V\\,\\bar p\\prod_i r_i\\;\\Longleftrightarrow\\;\\ln R=\\ln V+\\ln\\bar p+\\sum_i\\ln r_i", "assumptions": [ "Same complete cohort, nested stage denominators, mutually consistent stage definitions and time horizon.", "The product is a descriptive accounting decomposition; stage interventions can change several rates and customer mix." ], "revision": "3.3", "curriculum_cluster": "C4", "evidence_ids": [ "R10", "R16", "R17", "R19", "R22", "R35", "R37", "R56", "R58", "R61" ], "measurement_status": "measurable_ex_post", "identification_status": "not_applicable_to_accounting_or_model_algebra", "qualification_status": "specification_not_deployment_qualified", "units": { "V": "eligible cohort entrants in declared horizon", "r_i": "conditional stage fractions on consistent nested cohorts", "p_bar": "mean value per final outcome on a declared booking/price basis", "R": "value implied by those final outcomes, not automatically recognized revenue or cash" }, "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "unit_economics", "name": "The unit-economics master inequality", "epistemic_type": "estimator", "expression": "LTV > CAC <=> (m * p_bar) / delta > S / a <=> (m * p_bar * a) / (delta * S) > 1", "reads_as": "Under constant contribution and homogeneous undiscounted hazard, LTV/CAC=m*p_bar*a/(delta*S). Its elasticity to churn is -1 and to acquisition +1; retention is not quadratic.", "implication": "Evaluate discounted incremental contribution and cash viability over a declared horizon. Payback, runway, capacity, service promises and authority can bind simultaneously. Historical ratio benchmarks are heuristics rather than universal pass/fail thresholds.", "vocab": [ "LTV", "CAC", "LTV:CAC", "CAC Payback", "Magic Number" ], "benchmark": { "LTV:CAC": "Historical SaaS heuristics such as 3:1 require cost/cohort/basis comparability; not universal viability rules.", "CAC_payback": "Choose a cash- and business-specific limit from runway, renewal, margin and obligations.", "Magic_Number": "A descriptive growth/spend heuristic with declared lag and annualization; not a causal spend recommendation." }, "v2_correction": "Retain estimator status; aggregate churn alone does not identify a lifetime contribution curve, and no universal 2-3x correction is justified.", "latex": "LTV > CAC \\;\\Longleftrightarrow\\; \\frac{m\\,\\bar{p}}{\\delta} > \\frac{S}{a} \\;\\Longleftrightarrow\\; \\frac{m\\,\\bar{p}\\,a}{\\delta\\,S} > 1", "revision": "3.3", "curriculum_cluster": "C5", "evidence_ids": [ "R01", "R02", "R03", "R04", "R05", "R06", "R08", "R22", "R25", "R50", "R51", "R52", "R53", "R55", "R56", "R57", "R58", "R61", "R62", "R63" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "saturation", "name": "Channel response and saturation scenarios", "epistemic_type": "hypothesis", "expression": "CAC(x) = CAC_0 / (1 - (x / TAM)^beta)", "reads_as": "This reciprocal form is one proposed finite-market response family, not a universal CAC curve. Here x is cumulative captured demand in the same units as TAM.", "implication": "Estimate marginal incremental contribution in tested spend support. Distinguish saturation, changing demand, creative fatigue, delivery and rival reactions. Use channel experiments with time paths and spillover checks; do not extrapolate the asymptote as fact.", "vocab": [ "TAM / SAM / SOM", "CAC", "Channel Strategy", "Paid-Led", "Content-Led" ], "caveat": "Requires 0<=x0 and beta>0; convexity is not guaranteed over the full domain for every beta. Parameters are unfit here. Other response families and endogenous budgets can fit observational data similarly.", "latex": "CAC(x) = \\frac{CAC_0}{1 - \\left(\\dfrac{x}{TAM}\\right)^{\\beta}}", "revision": "3.3", "curriculum_cluster": "C3", "evidence_ids": [ "R01", "R02", "R03", "R04", "R06", "R07", "R09", "R18", "R21", "R45", "R52", "R53", "R54", "R57", "R64", "R65" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "motion_threshold", "name": "Motion feasibility under economics and buyer requirements", "epistemic_type": "hypothesis", "expression": "human_sales_feasible when expected_incremental_discounted_contribution >= incremental_sales_and_service_cost, subject to cash, capacity, quality, buyer and authority constraints", "reads_as": "ACV is one input to contribution and cost-to-serve. Renewal, implementation, margin, purchase complexity, preference and opportunity cost also matter.", "implication": "PLG/hybrid/sales-led ACV bands are historical segment heuristics. Choose feasible motions by measured economics and requirements; hybrids and simultaneous gates are allowed.", "vocab": [ "GTM Motion", "Product-Led Growth", "Sales-Led Growth", "Hybrid Motion", "ACV", "PQL", "PLG CRM" ], "agent_relevance": "LangChain (developer-led, technical buyer, expansion-heavy) runs a hybrid motion: self-serve adoption + human outbound for higher-intent leads. Their agent serves exactly the hybrid seam — qualifying which leads warrant a human-approved touch.", "latex": "\\mathbb E[\\Delta\\mathrm{PV}(\\mathrm{contribution})]\\geq\\Delta\\mathrm{cost}\\quad\\text{subject to feasibility gates}", "revision": "3.3", "curriculum_cluster": "C3", "evidence_ids": [ "R01", "R02", "R03", "R04", "R06", "R07", "R09", "R18", "R21", "R45", "R52", "R53", "R54", "R57", "R64", "R65" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "attribution_convolution", "name": "Distributed lags and the attribution problem", "epistemic_type": "hypothesis", "expression": "r(t) = integral over tau of [ k(tau) * A(t - tau) ] dtau", "reads_as": "A convolution describes a proposed distributed response to earlier activity. A linear stock with decay has an equivalent exponential convolution solution.", "implication": "Last-touch credit is not incremental contribution. Self-reported attribution can suggest discovery mechanisms but does not identify a lag kernel or causal brand effect by itself.", "vocab": [ "Attribution", "Self-Reported Attribution", "Dark Funnel", "Demand Creation", "Thought Leadership" ], "latex": "r(t) = \\int k(\\tau)\\,A(t - \\tau)\\,d\\tau", "assumptions": [ "Kernel, outcome and exposure units, initial stock, competing activity, feedback and interference must be specified.", "Observational spend may follow demand. A pause or fitted convolution alone does not establish exogenous treatment." ], "revision": "3.3", "curriculum_cluster": "C3", "evidence_ids": [ "R01", "R02", "R03", "R04", "R06", "R07", "R09", "R18", "R21", "R45", "R52", "R53", "R54", "R57", "R64", "R65" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "pmf_multiplier", "name": "PMF interaction hypothesis", "epistemic_type": "hypothesis", "expression": "Compare R=Phi*f(GTM), R=f(Phi,GTM), additive and interaction alternatives on a declared outcome scale", "reads_as": "Fit may modify GTM response, but multiplicative versus additive structure requires identification and out-of-sample validation.", "implication": "Use independent pre-outcome fit measures, segment holdouts and feasible intervention comparisons. Low measured fit is a warning, not a proven universal zero-response threshold.", "vocab": [ "Product-Market Fit", "Go-to-Market Fit", "NRR", "Win Rate", "CAC", "North Star Metric" ], "contested": true, "v2_correction": "Both the original universal multiplier and the proposed switching-cost piecewise replacement remain hypotheses. Neither is established by co-moving metrics or dissatisfied retained customers.", "latex": "R=\\Phi f(\\mathrm{GTM})\\quad\\text{versus alternative response models}", "revision": "3.3", "curriculum_cluster": "C1", "evidence_ids": [ "R09", "R11", "R12", "R13", "R15", "R18", "R46", "R47", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "sales_velocity", "name": "Sales throughput approximation", "epistemic_type": "estimator", "expression": "Sales_Velocity = (num_opportunities * win_rate * average_deal_size) / sales_cycle_length", "reads_as": "A stationary pipeline approximation combines consistently defined opportunities, win share, deal value and cycle time. Stock denominators and completed-cohort outcomes must match.", "implication": "These are coupled quantities. A faster research/draft step does not necessarily shorten procurement or implementation. Measure complete throughput and queue delays rather than assuming four independent causal levers.", "vocab": [ "Sales Velocity", "Win Rate", "Average Deal Size", "Sales Cycle", "Multi-Threading", "Mutual Action Plan", "Speed-to-Lead" ], "latex": "\\text{Sales Velocity} = \\frac{n_{\\text{opps}} \\cdot \\text{win rate} \\cdot \\text{avg deal size}}{\\text{sales cycle length}}", "assumptions": [ "Stable entry/completion flows and comparable cohort/cycle definitions; severe seasonality, censoring and heterogeneity can invalidate the approximation." ], "revision": "3.3", "curriculum_cluster": "C4", "evidence_ids": [ "R10", "R16", "R17", "R19", "R22", "R35", "R37", "R56", "R58", "R61" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "mrr_walk", "name": "Recurring-value ledger reconciliation", "epistemic_type": "identity", "expression": "MRR_t = MRR_{t-1} + New_t + Expansion_t - Contraction_t - Churn_t + Other_defined_movements_t", "reads_as": "Ending MRR equals beginning MRR plus mutually exclusive movements on a fixed recurring-value convention; include FX, reactivation, scope changes or corrections as separate reconciliation terms when relevant.", "implication": "This is a recurring-value accounting identity, not recognized GAAP revenue or cash. Other ratios can also be exact definitions on their declared basis; their causal meaning and forecasts are separate.", "leaks_when": "The ledger fails to reconcile when movements, boundaries, currency or conventions are inconsistent or incomplete. A difference from recognized revenue/cash alone is not a falsification; use an explicit bridge.", "vocab": [ "MRR", "ARR", "Expansion Revenue", "Churn", "Contraction" ], "provenance": "Construction A (all four converge).", "latex": "MRR_t=MRR_{t-1}+\\mathrm{New}_t+\\mathrm{Exp}_t-\\mathrm{Contr}_t-\\mathrm{Churn}_t+\\mathrm{Other}_t", "reconciliation": "MRR_end - (MRR_start + New + Expansion - Contraction - Churn + Other_defined_movements) = residual; verify against an independently observed ending balance.", "revision": "3.3", "curriculum_cluster": "C5", "evidence_ids": [ "R01", "R02", "R03", "R04", "R05", "R06", "R08", "R22", "R25", "R50", "R51", "R52", "R53", "R55", "R56", "R57", "R58", "R61", "R62", "R63" ], "measurement_status": "measurable_ex_post", "identification_status": "not_applicable_to_accounting_or_model_algebra", "qualification_status": "specification_not_deployment_qualified", "assumptions": [ "Other includes separately defined reactivation, FX, scope changes or corrections when not already classified; no double counting.", "This is a recurring run-rate convention, distinct from revenue recognition and cash collections." ], "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "ltv_corrected", "name": "Corrected LTV under heterogeneous churn (Fader-Hardie sBG)", "epistemic_type": "estimator", "expression": "LTV=sum_{t>=0} E[net_contribution_t*1(active_t)]/(1+d)^t - nonrecurring_costs, with explicit first_payment; sBG q~Beta(alpha,beta), S(t)=B(alpha,beta+t)/B(alpha,beta)", "reads_as": "Survival-weighted expected net contribution requires a stated payment schedule, censoring and cohort model. Heterogeneity can change aggregate hazards without any individual behavior improvement.", "implication": "For alpha=2.5,beta=6, mean initial churn is 2.5/8.5=29.41%. Expected paid periods including t=0 are 5; future periods excluding t=0 are 4. A geometric model matched to initial churn gives 3.4 including t=0 or 2.4 future periods. This example does not support a universal LTV overstatement or an invented 10%-churn comparator.", "vocab": [ "LTV", "Churn", "Retention", "Cohort Analysis" ], "provenance": "Fader-Hardie 2007 (sBG) / 2018 (BdW), established; worked numbers from construction A.", "status": "sBG is an established model family under contractual, time-invariant individual dropout assumptions; no local fitted cohort parameters supplied here.", "latex": "\\mathrm{LTV}=\\sum_{t\\ge0}\\frac{\\mathbb E[c_t\\mathbf1(\\mathrm{active}_t)]}{(1+d)^t}-C_{\\mathrm{nonrec}},\\qquad S(t)=\\frac{B(\\alpha,\\beta+t)}{B(\\alpha,\\beta)}", "revision": "3.3", "curriculum_cluster": "C5", "evidence_ids": [ "R01", "R02", "R03", "R04", "R05", "R06", "R08", "R22", "R25", "R50", "R51", "R52", "R53", "R55", "R56", "R57", "R58", "R61", "R62", "R63" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "pmf_piecewise", "name": "Fit and switching-cost alternatives", "epistemic_type": "hypothesis", "expression": "R = Phi*f(GTM) if S < S* (low-switching)\nR = Phi*f1(GTM) + S*f2(installed_base) if S >= S* (high-switching)\nwith effective churn hazard h_eff = h0 * exp(-S)", "reads_as": "Switching costs can affect renewal and rents. A sharp regime threshold, an exponential hazard and additive revenue form are proposed specifications, not identified consequences of lock-in theory.", "implication": "GRR plus low satisfaction does not identify switching costs, and retained enterprise revenue is not an existence proof of an additive causal law. Compare specifications with independent fit/switching-cost measurements and interventions.", "S_scale": "S* and hazard parameters are unmeasured. Do not infer account-level switching cost from GRR thresholds alone.", "vocab": [ "Switching Costs", "Vendor Lock-in", "Retention", "GRR", "NRR" ], "provenance": "v3.3 internal mathematical/estimand correction; relevant scoped current sources and older foundations documented in the research report. Earlier construction-report unanimity is not validation.", "evidence_caveat": "Classical switching-cost mechanisms motivate the hypothesis; this exact piecewise GTM revenue model has not been fitted or causally tested here.", "latex": "R = \\begin{cases} \\Phi\\, f(\\text{GTM}) & S < S^{*} \\\\[4pt] \\Phi\\, f_1(\\text{GTM}) + S\\, f_2(\\text{installed base}) & S \\ge S^{*} \\end{cases} \\qquad h_{\\text{eff}} = h_0\\, e^{-S}", "revision": "3.3", "curriculum_cluster": "C1", "evidence_ids": [ "R09", "R11", "R12", "R13", "R15", "R18", "R46", "R47", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "psi_objective", "name": "Macro regime objective re-weighting (Psi / Rule-of-X)", "epistemic_type": "hypothesis", "expression": "U = w_g(Psi)*growth + w_f(Psi)*FCF ; Rule-of-X: X = M(Psi)*growth + FCF_margin ; M ~ 2-3 by capital tightness", "reads_as": "Capital conditions can constrain decisions and influence valuation. Operating objective weights are declared preferences, not causally determined by a fitted macro scalar in this model.", "implication": "Separate market multiples/rates, firm funding/runway/covenants and management objectives. Compare scenarios; no fixed rate-to-payback or rate-to-growth-weight mapping is identified.", "measured": "Historical rate and valuation observations are construction context, not current October 2026 estimates. M-by-regime and operating objective weights are unfit.", "vocab": [ "Rule of 40", "CAC Payback", "Burn Multiple", "Magic Number" ], "provenance": "Construction C (all four converge); Bessemer Rule-of-X 2024; Macrotrends; adversarial audit blind-spot #1.", "evidence_caveat": "Rule-of-X is a practitioner valuation heuristic. Units, scaling and firm mandate must be declared; causal operating policies do not follow from it.", "latex": "U = w_g(\\Psi)\\,\\text{growth} + w_f(\\Psi)\\,\\text{FCF}, \\qquad X = M(\\Psi)\\,\\text{growth} + \\text{FCF margin}, \\qquad M \\approx 2\\text{–}3", "revision": "3.3", "curriculum_cluster": "C7", "evidence_ids": [ "R04", "R07", "R08", "R28", "R31", "R32", "R33", "R34", "R41", "R44", "R45", "R47", "R48", "R49", "R51", "R55", "R63", "R64", "R65" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "buyer_state_coupling", "name": "Buyer mix and causal policy estimands", "epistemic_type": "conditional_coefficient", "expression": "Prediction: P(Y_H|A,X_pre). Policy effect: E[Y_H(a)-Y_H(a0)|X_pre]; with interference: E[Y_H(policy_joint)-Y_H(reference_joint)]", "reads_as": "Pre-action shortlist, need, role, constraints and prior exposure can influence assignment and outcomes. Observed funnel rates mix these states. Conditioning on them helps prediction but does not itself identify a causal effect.", "implication": "Compare feasible contact/evaluation/implementation policies, including no action when appropriate, on a stable eligible population. The 95-5 in-market heuristic and pre-contact winner-shortlist prevalence have different denominators and identify no 95% value-creation split or geometric headcount-decay law.", "vocab": [ "Buying Group", "Day-1 Shortlist", "Intent Data", "95-5 Rule", "Win Rate", "Dark Funnel" ], "provenance": "v3.3 internal mathematical/estimand correction; relevant scoped current sources and older foundations documented in the research report. Earlier construction-report unanimity is not validation.", "evidence_caveat": "Retrospective commercial surveys motivate buyer-state measurement. Effects need randomization or justified exchangeability, positivity, consistency and interference handling; conditional effects are averages, not identifiable individual effects.", "latex": "\\tau_H(x)=\\mathbb E[Y_H(a)-Y_H(a_0)\\mid X_{\\mathrm{pre}}=x]\\ne P(Y_H\\mid a,x)", "revision": "3.3", "curriculum_cluster": "C2", "evidence_ids": [ "R05", "R06", "R07", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R18", "R19", "R20", "R21", "R37", "R39", "R46", "R53", "R54", "R60", "R64" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "brand_stock", "name": "Brand memory: stock and distributed-lag representations", "epistemic_type": "hypothesis", "expression": "dB_brand/dt=build_brand*I(t)-decay_brand*B_brand; linear solution equals initial_stock*exp(-decay_brand*t)+exponential_convolution(I)", "reads_as": "A linear memory stock and exponential distributed lag are equivalent representations under the same assumptions. Nonlinear responses and independent awareness/shortlist observations may distinguish richer models.", "implication": "Track validated awareness, entry points, consideration and business outcomes over time. Do not infer a universal shortlist effect or a unique stock mechanism from a fitted long lag.", "parameters": "Illustrative continuous decay 0.10-0.20/year implies half-life ln(2)/decay=3.47-6.93 years; if 10-20% annual discrete loss, half-life=3.11-6.58 years. SaaS build/decay are not measured here.", "vocab": [ "Brand Awareness", "Mental Availability", "Category Entry Points", "Share of Voice", "Demand Creation" ], "provenance": "Construction E (all four converge); Nerlove-Arrow 1962 (established); Ehrenberg-Bass / Binet-Field; LinkedIn 95-5.", "evidence_caveat": "Nerlove-Arrow is an established model family; local causal response and decay need identified variation and independent observations. A media pause alone cannot separate mathematically equivalent linear representations.", "latex": "\\dot B_{\\mathrm{brand}}=b_{\\mathrm{brand}}I-d_{\\mathrm{brand}}B_{\\mathrm{brand}},\\qquad t_{1/2}=\\frac{\\ln2}{d_{\\mathrm{brand}}}", "revision": "3.3", "curriculum_cluster": "C2", "evidence_ids": [ "R05", "R06", "R07", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R18", "R19", "R20", "R21", "R37", "R39", "R46", "R53", "R54", "R60", "R64" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "plg_loop", "name": "PLG generation, cumulative referrals and active-stock loops", "epistemic_type": "causal_regime", "expression": "G_{t+1}=k*G_t; G_t=G0*k^t; sum_{t>=0}G_t=G0/(1-k) for 0<=k<1. Separately U_{t+1}=(1+k-c)*U_t grows when k>c.", "reads_as": "Generation size, cumulative unique referrals, active users and paid customers are different quantities. At k=1 generation size is constant and cumulative generations grow linearly. A finite eligible market, overlap and duplicate referrals constrain real growth.", "implication": "Define invite, activation, retention and payment denominators; estimate conditional loop parameters and sales-assist policies. Retention can affect invitations but a fixed k_raw*retention factor is a toy specification, not a universal causal law.", "k_reality": "No verified sustained organic B2B k>=1 estimate was established by this review. Absence in the reviewed evidence does not prove k<1 always or make the growth boundary irrelevant.", "vocab": [ "Product-Led Growth", "Viral Coefficient", "Free-to-Paid Conversion", "PQL", "Time to Value", "Expansion Revenue" ], "provenance": "Construction F (all four converge); Andrew Chen 'Retention is King'; OpenView/Poyer/ProductLed benchmarks.", "evidence_caveat": "Constant-rate recurrences are mathematical scenarios; network saturation, heterogeneity and commercial conversion require measurement.", "latex": "G_t=G_0k^t,\\quad\\sum_{t\\ge0}G_t=\\frac{G_0}{1-k}\\ (0\\le k<1),\\qquad U_{t+1}=(1+k-c)U_t", "revision": "3.3", "curriculum_cluster": "C3", "evidence_ids": [ "R01", "R02", "R03", "R04", "R06", "R07", "R09", "R18", "R21", "R45", "R52", "R53", "R54", "R57", "R64", "R65" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "phi_ex_ante", "name": "Independent fit measurement and competing response models", "epistemic_type": "hypothesis", "expression": "Compare level-scale additive, multiplicative and flexible interaction models using independently measured Phi_pre, stable eligibility and observed policy support", "reads_as": "A zero interaction in log(R)=b0+b1*log(Phi)+b2*log(GTM) already gives a multiplicative power model in levels. A positive log interaction does not decide additive versus multiplicative revenue structure.", "implication": "Specify the outcome scale and competing models; freeze pre-outcome fit measures, account for selection and confounding, and use holdout predictive/causal targets separately. Retention or satisfaction alone does not uniquely measure fit.", "vocab": [ "Product-Market Fit", "Retention", "NPS", "North Star Metric" ], "provenance": "v3.3 internal mathematical/estimand correction; relevant scoped current sources and older foundations documented in the research report. Earlier construction-report unanimity is not validation.", "evidence_caveat": "Proposed, unrun. Panels and interventions must be appropriate to the target population; no coefficient is estimated here.", "latex": "\\log R=b_0+b_1\\log\\Phi+b_2\\log\\mathrm{GTM}\\;\\Rightarrow\\;R=e^{b_0}\\Phi^{b_1}\\mathrm{GTM}^{b_2}\\quad\\text{even without interaction}", "revision": "3.3", "curriculum_cluster": "C1", "evidence_ids": [ "R09", "R11", "R12", "R13", "R15", "R18", "R46", "R47", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "noise_floor", "name": "Forecast validation and refusal", "epistemic_type": "estimator", "expression": "Measure out-of-sample error/coverage for declared target, population, horizon and regime; refuse when support, outcome maturity, data quality or decision precision is inadequate", "reads_as": "The earlier 25-40% figure was an unfitted assumption, not a measured or physically irreducible limit. Error and coverage vary with target and information.", "implication": "Use temporal/cohort holdouts, support checks, calibration and scenario sensitivity. Refusal limits are declared decision policies. Shock probabilities may be estimated under explicit information/models; unavailable inputs remain unknown.", "is_it_reducible": "Potentially; distinguish measurement error, omitted predictors, stochastic variation, regime shift and model misspecification. No universal maturity month or variance floor is asserted.", "vocab": [ "Forecast Accuracy", "Sales Forecasting", "Leading vs Lagging Indicators" ], "provenance": "Construction H (all four converge); Bill Gross TED (42% timing, hindsight-biased); Taleb heavy tails; Roberge 'sales is predictable' counter.", "evidence_caveat": "No public benchmark in this review establishes a universal B2B growth noise fraction.", "latex": "\\text{forecast only within validated support, coverage and decision precision}", "revision": "3.3", "curriculum_cluster": "C0", "evidence_ids": [ "R25", "R44" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "agentic_error_step", "name": "E-A1 · Shipped error under a complete review policy", "epistemic_type": "estimator", "expression": "p_shipped=P((D AND not C) OR D_review | released,task,state,policy,configuration); separately report release/defer/stop coverage", "latex": "p_{\\mathrm{ship}}=P((D\\cap\\neg C)\\cup D_{\\mathrm{review}}\\mid\\mathrm{released,task,state,policy,config})", "reads_as": "D is an initial defect, C a successful correction of it, and D_review a surviving review-introduced defect, all on declared criteria. Condition shipped-error rates on released actions, and separately report whole eligible workload acceptance, release, repair/defer/stop and severe process failures.", "implication": "Measure defect detection, correction effectiveness, false rejection and terminal/process success separately; autonomy fraction is not a causal sufficient statistic.", "caveat": "Per-step estimation requires labeled observations and a defined eligible population; independence is unnecessary for a single marginal error estimate.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "Agentic GTM", "Human-in-the-Loop" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "reviewer_fatigue_catchrate", "name": "E-A2 · Review discrimination, load and capacity", "epistemic_type": "estimator", "expression": "catch=P(flag|defect,reviewed,case_mix,policy); false_reject=P(reject|clean,reviewed,case_mix,policy); load=sum(review_minutes)/available_minutes", "latex": "\\gamma=P(\\mathrm{flag}\\mid D,\\pi,x),\\quad f=P(\\mathrm{reject}\\mid\\neg D,\\pi,x),\\quad\\ell=\\frac{\\sum\\mathrm{review\\ minutes}}{\\mathrm{available\\ minutes}}", "reads_as": "Review fraction is not load. Arrival rates, service time, expertise, available minutes, queue age and defect severity determine review feasibility and quality.", "implication": "Use matched clean controls plus labeled defects and accepted throughput. The former gamma_max/rho parabola and governance sign-flip were illustrative and uncalibrated, not an operating law.", "caveat": "Recent review-window evidence concerns LLM monitors, not human fatigue. Human load-response coefficients require local evidence.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "Human-in-the-Loop", "Governance" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "agentic_chain_reliability", "name": "E-A3 · Conditional chain success and accepted workflow outcomes", "epistemic_type": "model_identity", "expression": "P(all_step_success)=product_i P(success_i|success_1,...,success_{i-1}); homogeneous independent scenario=(1-p)^n", "latex": "P(\\cap_i S_i)=\\prod_iP(S_i\\mid S_1,\\ldots,S_{i-1});\\qquad (1-p)^n\\ \\text{only if homogeneous independent}", "reads_as": "The chain rule is general. The repeated marginal power needs homogeneous independent failures. Dependence can raise or lower all-success probability relative to the independent product.", "implication": "Measure terminal success, required process validity, authority and acceptance with repair/defer/stop policies. No-intermediate-error and accepted-terminal-success are distinct targets.", "caveat": "Shared 95%-success events can give all-success .95 rather than .95^n; correlation does not universally lower reliability. Estimate uncertainty over representative held-out cases.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "Agentic GTM" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "not_applicable_to_accounting_or_model_algebra", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "agentic_effective_cost", "name": "E-A4 · Full cost, accepted success and constrained policy selection", "epistemic_type": "estimator", "expression": "cost_per_accepted_success=total_eligible_cost/accepted_success_count (undefined at zero); select policy for fixed eligible workload by contribution/cost subject to quality, authority, risk and capacity gates", "latex": "C_{\\mathrm{accepted}}=\\frac{C_{\\mathrm{all\\ eligible\\ attempts}}}{N_{\\mathrm{accepted}}}\\quad(N_{\\mathrm{accepted}}>0)", "reads_as": "Total eligible cost includes successful and failed attempts, retries, review, rework, data/tools, maintenance and attributable losses. A per-attempt expected cost is not cost per completed success.", "implication": "Compare equal eligible workloads and accepted throughput. Keep severe authority/process gates outside monetary tradeoffs; governance effectiveness and optimal review are task-specific. No universal governance threshold is identified.", "caveat": "Report avoidable incremental cost and allocated overhead separately. Ratios with different acceptance counts can misrank policies; include cost, throughput and quality jointly.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "Agentic GTM", "CAC", "Human-in-the-Loop" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "agentic_cac_split", "name": "E-A5 · Agent execution spend inside CAC", "epistemic_type": "estimator", "expression": "CAC=complete_acquisition_cost/new_acquired_customers; CPQO=complete_eligible_cost/qualified_opportunities; separate human, model, tool/data, maintenance and attributed acquisition costs", "latex": "\\mathrm{CAC}=\\frac{C_{\\mathrm{acquisition}}}{N_{\\mathrm{new\\ customers}}},\\qquad\\mathrm{CPQO}=\\frac{C_{\\mathrm{eligible}}}{N_{\\mathrm{qualified\\ opportunities}}}", "reads_as": "Customer acquisition and qualified opportunity denominators differ. Cost allocation, cohort, horizon and shared tooling scope must be declared.", "implication": "Released hours are cash savings only when expenditure falls; productive redeployment requires measured incremental contribution. Inference cost per activity alone proves neither CAC reduction nor profit.", "caveat": "Avoid double-counting C_a already included in S or COGS. Separate ongoing delivery cost from acquisition spend and include failed attempts.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "CAC", "Agentic GTM" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "hitl_learning_rate", "name": "E-A6 · Evidence-to-policy learning contract", "epistemic_type": "hypothesis", "expression": "receipt -> criterion-specific_label -> valid_outcome_linkage -> policy_update -> holdout_evaluation; coefficient change is measured, not assumed from approval", "latex": "\\mathrm{receipt}\\to\\mathrm{valid\\ label}\\to\\mathrm{update}\\to\\mathrm{held\\ out\\ evaluation}", "reads_as": "Approval, correction, implementation and improvement are different events. Style/factuality/authority labels are not business outcome labels.", "implication": "Preserve pre-action data, assignment, known action probabilities where applicable, review selection, missingness and delayed outcomes. Identify counterfactual effects separately from selectively missing factual labels; random audits/experiments, justified weighting or bounds may be needed.", "caveat": "The earlier lambda*gamma*(target-r_i) drift was unfit. Down-weighting rubber stamps alone does not repair selection or missing outcomes; no universal learning-rate input is available.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "Human-in-the-Loop", "Agentic GTM" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "category_noise_floor", "name": "E-A7 · Category adoption and externality scenarios", "epistemic_type": "hypothesis", "expression": "dtheta/dt=eta*kappa(t)-delta_theta*theta; theta*=eta*kappa/delta_theta only for constant kappa and delta_theta>0", "latex": "\\dot\\theta=\\eta\\kappa(t)-\\delta_\\theta\\theta,\\qquad\\theta^*=\\frac{\\eta\\kappa}{\\delta_\\theta}\\quad\\text{if constant }\\kappa,\\ \\delta_\\theta>0", "steady_state": "theta*=eta*kappa/delta_theta for constant kappa and positive delta_theta", "steady_state_latex": "\\theta^*=\\frac{\\eta\\kappa}{\\delta_\\theta}\\quad(\\kappa\\ \\mathrm{constant},\\ \\delta_\\theta>0)", "reads_as": "The positive linear adstock is a possible congestion scenario. Agent adoption can also improve relevance or grow demand; the net externality sign is open.", "implication": "Measure exposure volume, relevance, attention, platform/rival response and joint-policy outcomes. Reply-rate decline alone does not identify agent-induced pollution.", "caveat": "Bounded time-varying adoption with positive delta_theta gives a bounded linear response, not necessarily convergence to a constant. Constant kappa gives the stated steady state; a varying input requires its convolution. No inevitable degradation or unbounded noise is established; parameters are unfit.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "Brand", "Demand Generation", "Agentic GTM" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "agentic_cac_holdout", "name": "E-A8 · Commercial policy comparison through complete outcomes", "epistemic_type": "estimator", "expression": "ITT_contribution=E[Y_H|assigned_agent_policy]-E[Y_H|assigned_reference_policy]; separately report delta_CAC, delta_CPQO, cost, acquisitions, quality and capacity", "latex": "\\Delta_H=\\mathbb E[Y_H\\mid Z=\\pi_{\\mathrm{agent}}]-\\mathbb E[Y_H\\mid Z=\\pi_{\\mathrm{reference}}]", "reads_as": "Randomized eligible account/buying-group assignment can identify the complete tested policy effect under its design assumptions. Matched observational cohorts require additional exchangeability/support assumptions.", "implication": "Run through sufficient acquisition, implementation and renewal horizons for the declared endpoint. Existing B2B human-seller routing evidence prevents blanket claims that no causal commercial evidence exists; autonomous seller net CAC savings remain unresolved in reviewed evidence.", "caveat": "Full outcome linkage, stable definitions, contamination/interference checks, cost completeness and uncertainty are required. Do not relabel CPQO as CAC or a matched comparison as a randomized holdout.", "contested": true, "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "CAC", "Agentic GTM" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." }, { "id": "buyer_state_agent_ext", "name": "E-A9 · Role × task × phase AI mediation", "epistemic_type": "correlational", "expression": "B_AI[role,task,phase]={observed_use,platform,interface,version,fidelity,verification,authority,fulfillment,timestamp,unknown}; B_agent retained as a descriptive aggregate only", "latex": "B_{\\mathrm{AI}}[\\mathrm{role,task,phase}]\\quad\\text{with fidelity, verification and authority}", "reads_as": "Research, recommendation, drafting, guarded execution and purchasing authority vary by role and phase. Survey permission preferences are not observed usage or purchase autonomy.", "implication": "Measure initial/current shortlists, proof requirements, human veto and actual implementation. Validate synthetic preference representation for the decision; citations, consideration, contracts and buyer welfare remain separate outcomes.", "caveat": "A single agent share cannot identify a causal value split or replace observed buying journeys. UI/API differences and version changes require remeasurement.", "agent_relevance": "Use the current task/state policy, accepted outcomes, authority gates and complete cost contract; preserve unknown inputs and requalify after material changes.", "vocab": [ "Buying Committee", "Buyer Persona" ], "revision": "3.3", "curriculum_cluster": "C8", "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ], "measurement_status": "unmeasured_for_local_decision", "identification_status": "local_causal_effect_not_identified", "qualification_status": "specification_not_deployment_qualified", "evidence_usage": "Related cluster source appraisals, not direct empirical validation of this equation. Internal mathematical corrections and older foundations are described separately in the research report." } ], "key_edges": [ { "from": "Phi", "to": "Win Rate", "type": "latent_multiplier", "sign": "+", "note": "Proposed fit association/response modifier; direction and causal effect require independently measured pre-outcome fit and scope-specific evidence.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Phi", "to": "Churn", "type": "latent_multiplier", "sign": "-", "note": "Proposed fit association/response modifier; direction and causal effect require independently measured pre-outcome fit and scope-specific evidence.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Phi", "to": "CAC", "type": "latent_multiplier", "sign": "-", "note": "Proposed fit association/response modifier; direction and causal effect require independently measured pre-outcome fit and scope-specific evidence.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "ICP", "to": "r_i", "type": "causal_regime", "sign": "context-dependent", "note": "Targeting changes eligibility and buyer mix as well as behavior. Evaluate complete eligible outcomes; higher conditional rates alone do not identify lift.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Speed-to-Lead", "to": "r_i", "type": "causal_regime", "sign": "context-dependent", "note": "Response latency is a candidate intervention. Its effect depends on need, quality, capacity and buyer preference; measure the complete policy.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Pricing & Packaging", "to": "GTM Motion", "type": "causal_regime", "sign": "context-dependent", "note": "Price/metering alter contribution, demand and service requirements; motion feasibility has simultaneous constraints.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "GTM Motion", "to": "CAC", "type": "causal_regime", "sign": "depends", "note": "Motion determines cost-to-serve structure.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Acquisition (a)", "to": "C", "type": "model_identity", "sign": "+", "note": "Logo or currency/time inflow must match the chosen stock.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Churn (delta)", "to": "C", "type": "model_identity", "sign": "-", "note": "Logo/value outflow separated; constant-rate equilibrium sensitivity is conditional, not investment ROI.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Demand Creation", "to": "Win Rate", "type": "causal_regime", "sign": "unknown/lagged", "note": "Delayed consideration and contribution response are hypotheses; last-touch or self-report attribution does not identify incrementality.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "NRR", "to": "CAC efficiency", "type": "correlational", "sign": "+", "note": "Moves together but confounded by Phi. Do NOT treat as 'improve NRR to fix CAC'.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Human override", "to": "r_i coefficients", "type": "hypothesis", "sign": "learning", "note": "Approval, correction, implementation and improvement are different events. Style/factuality/authority labels are not business outcome labels.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Agent FinOps", "to": "CAC", "type": "accounting_definition", "sign": "+", "note": "Agent acquisition costs belong in declared total acquisition cost, without double counting delivery costs. FinOps measurement itself has no identified effect.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Agentic GTM", "to": "r_i", "type": "hypothesis", "sign": "unknown", "note": "Mean and distributional gains require task/cohort intervention evidence; no universal coefficient compression.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Compounding Failure", "to": "Human-in-the-Loop", "type": "hypothesis", "sign": "policy-dependent", "note": "Compare review, deterministic validation, repair/defer and enforced authority; human presence alone does not certify reliability.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." }, { "from": "Sense-Reason-Act-Learn", "to": "Speed-to-Lead", "type": "causal_regime", "sign": "context-dependent", "note": "A workflow may reduce signal-to-action latency; measure acceptance, quality and downstream contribution under capacity/authority constraints.", "qualification": "Directional or accounting interpretation only on a declared scope; causal effect not implied by sign." } ], "agent_interface": { "principle": "Execute only authorized task/state policies within configuration and economic/authority constraints. Review, repair, defer and stop are actions in the policy. Qualification and outcome-linked evaluation precede claims of commercial value.", "worked_example_langchain": { "source": "https://www.langchain.com/blog/how-we-built-langchains-gtm-agent", "tier_3_inputs_consumed": "Warm/cold playbook encoded as a Deep Agents 'skill'; ICP and qualification implicit in lead tiering (silver leads); motion = hybrid (self-serve + human outbound).", "tier_2_objectives_served": "Raise lead->qualified-opportunity conversion (the r_i term feeding CAC efficiency); surface expansion (g) and churn-risk (delta) in account intelligence.", "tier_1_loop_automated": "Trigger on Salesforce lead -> do-not-send checks (Rules of Engagement) -> research CRM/Gong/LinkedIn/Exa -> load skill -> draft -> human approve in Slack -> queue follow-ups -> 48h SLA auto-send for silver.", "substrate_used": "Salesforce/Gong/BigQuery/Gmail systems of record; LangSmith for evals+observability; Postgres per-rep memory; subagents per account for parallel account-intelligence.", "learning_loop": "Rep edit-diff memory is a described implementation of style feedback; it does not by itself identify coefficient learning or business impact.", "drift_handling": "Evals in CI + 'treat unexplained behavior drift as a bug' = the regime-tier principle that behavioral coefficients decay and must be re-checked.", "results_in_model_terms": "Reported opportunity/pipeline and reclaimed-hours outcomes are uncontrolled case observations. Opportunity creation is not acquired customers, realized contribution, cash savings or demonstrated full-cycle compression." }, "what_an_agent_must_NOT_do": [ "Choose or silently drift the ICP, motion, or pricing (Tier 3 — human-owned).", "Optimize a correlational edge as if causal (e.g. chase NRR to 'fix' CAC).", "Treat a latent fit score or unavailable coefficient as a fitted decision input; refuse unsupported deployment or clearly label a sensitivity scenario.", "Send without human-in-the-loop where a single mistimed touch destroys relationship capital." ], "what_an_agent_is_uniquely_good_at": [ "Candidate uses: research, drafting, routing and signal triage; compare to simpler tools and the current human path.", "Validate quality, acceptance, cost and downstream contribution for each proposed use.", "Preserve raw action receipts and verified criterion-specific feedback." ], "atomic_loop": "Each agent runs Sense->Reason->Act->Learn (see atomic_execution_cycle). The interface boundary applies to the whole loop: it senses/reasons/acts within Tier 1 and reads Tiers 2-3.", "governance_instruments": { "task_to_role_matrix": "Per process step, label agent-owned / agent-assisted (human reviews) / human-owned. The concrete form of the Tier-1-vs-Tier-3 boundary.", "escalation_paths": "Signals that force a human handoff — pricing questions, legal mentions, negative sentiment. The safety valve where authority/relationship judgment is required.", "data_quality_gates": "Agent acts only on verified+enriched records. Prevents machine-speed action on stale data — the most common failure mode.", "audit_logging": "Every action logged with timestamp+agent id, traceable and reviewable. The accountability layer that makes drift detectable.", "agent_finops": "Track agent spend per task/account to keep the FinOps->CAC edge from going the wrong way.", "action_permissions": "Task/action allowlist, budget and authority predicates, independently enforced at execution.", "qualification": "Versioned population, task, tool/data/configuration, process predicates, terminal acceptance, uncertainty, capacity and full-cost evidence.", "requalification": "Material model, tool, policy, definition, data, population or authority changes invalidate transfer unless retested." }, "adoption_structure": { "maturity_ladder": "Breadth/autonomy are descriptive, not a value ladder. Use the simplest workflow that qualifies for the task.", "phased_rollout": "Develop on separate cases, freeze a bounded policy, test representative held-out workflows, then compare commercial policy outcomes within scope.", "prove_roi": "Design a powered account/buying-group comparison with complete costs and outcomes through the declared sales/implementation/renewal horizon; no universal 30-60-day horizon." } }, "theses": [ { "id": "T0", "claim": "GTM combines convention-bound accounting, conditional summaries, causal policy questions and unvalidated mechanisms. Their different evidence requirements govern which decisions the model can support.", "type": "meta", "load_bearing": true, "revision": "3.3", "contested": false, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T1", "claim": "Retention and acquisition alter different stock flows; their relative investment value depends on horizon, feasible response, costs and quality. Long-run denominator sensitivity does not prove universal retention dominance.", "from_equation": "conservation_law", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T2", "claim": "The funnel product gives unit elasticity to every stage rate. Prioritize feasible incremental contribution under downstream and capacity effects; the lowest observed conversion rate is not automatically the highest-return intervention.", "from_equation": "funnel_operator", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T3", "claim": "Cash payback and discounted contribution are complementary. Runway, financing, renewal, obligations and downside determine which constraints bind; no universal capital-regime payback threshold is fitted.", "from_equation": "psi_objective", "v2_status": "Historical v2 construction superseded by the current v3.3 claim and equation; prior wording remains available in Git.", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T4", "claim": "Channels can saturate and response can drift, but convex CAC and a time-shifting portfolio are hypotheses rather than universal laws. Choose supported marginal contribution policies under capacity and spillovers.", "from_equation": "saturation", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T5", "claim": "Motion is constrained by contribution, service cost, cash, capacity, product and buyer requirements. ACV bands are segment heuristics; hybrid policies remain admissible.", "from_equation": "motion_threshold", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T6", "claim": "Delayed response and endogenous exposure make attribution different from incrementality. Last-touch and self-reports alone cannot identify a causal lag kernel or justify spend allocation.", "from_equation": "attribution_convolution", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T7", "claim": "Fit and switching costs may modify GTM response. Universal multiplication, a sharp switching threshold and additive moat revenue are competing unvalidated models, requiring independent pre-outcome measures and scoped tests.", "from_equation": "pmf_piecewise", "load_bearing": true, "contested": true, "v2_status": "Historical v2 construction superseded by the current v3.3 claim and equation; prior wording remains available in Git.", "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T8", "claim": "Agents can execute authorized research, routing and drafting policies under human-owned strategy and explicit economic/authority constraints. Conversion, cycle-time and contribution improvements require task-specific evidence.", "type": "architecture", "evidence": "LangChain GTM agent", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T9", "claim": "Human review can supply criterion-specific observations; valid labels, implemented updates and improved held-out outcomes must be established separately. Approval alone is neither ground truth nor a learning algorithm.", "type": "architecture", "evidence": "LangChain memory loop", "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T10", "claim": "Shared architecture patterns motivate governed workflows but do not validate a causal model or demonstrate business advantage. Qualification depends on authority, data, outcome quality, full cost and capacity.", "type": "architecture", "evidence": "Apollo, Poexis, Wyzard convergence", "load_bearing": true, "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T11", "claim": "Agent cost and reliability depend on task, model/tools, policy and environment. Failures can propagate, so compare independent acceptance, repairs, human paths and all eligible costs; human inclusion is not universally sufficient or necessary for every action.", "from": "agent_properties", "load_bearing": true, "revision": "3.3", "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T12", "claim": "Buyer state can select accounts into seller contact and condition conversion. In-market share and pre-contact winner-shortlist prevalence are descriptive quantities, not a 95% causal value split or a geometric seller-scaling law.", "from_equation": "buyer_state_coupling", "load_bearing": true, "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect.", "revision": "3.3" }, { "id": "T13", "claim": "A linear brand-memory stock has an equivalent exponential distributed-lag representation. Local awareness, shortlist and contribution effects require independent observations and identified variation; SaaS decay is unmeasured here.", "from_equation": "brand_stock", "load_bearing": true, "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect.", "revision": "3.3" }, { "id": "T14", "claim": "PLG generations, cumulative referrals, active stock and paid customers are distinct. Constant-rate boundaries differ; no reviewed evidence proves B2B k<1 always or perpetual sustainable growth.", "from_equation": "plg_loop", "load_bearing": false, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T15", "claim": "Agentic commercial value must be established for a complete policy through contribution, quality and full costs. Causal human-seller routing and customer-service evidence exist; neither establishes universal hybrid superiority or autonomous SDR net CAC savings.", "from_equation": null, "load_bearing": true, "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect.", "revision": "3.3" }, { "id": "T16", "claim": "Allocate agents across discovery, evaluation, approval, implementation and renewal according to measured incremental value and authority. There is no identified 95% versus 5% value-creation split that predetermines seller-agent leverage.", "from_equation": "buyer_state_coupling", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T17", "claim": "Broad coverage is useful only with honest evidence states, units and scope. Measurable, measured, causally identified and deployment-qualified are different; an epistemic tag alone does not authorize a decision.", "load_bearing": true, "contested": false, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T18", "claim": "Queues, available service minutes, expertise, rework and feedback can constrain realized outcomes. A dimensionless alignment score cannot be inserted into a growth min-law or establish that organization always dominates lead quality.", "from_layer": "org_cognition", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T19", "claim": "Imitation and rival/platform responses can change a tactic’s effect, alongside demand and delivery changes. An observed decay or changepoint does not identify reflexivity as its cause.", "from_layer": "competitive_ecology", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T20", "claim": "Measurement and incentives can alter behavior. Freeze definitions, preserve lineage and audit policy-induced selection before attributing metric movement to productivity.", "from_layer": "substrate_epistemics", "load_bearing": true, "contested": false, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T21", "claim": "Fit decomposition is useful when component observations are independently validated for the intended target. Retention reflects fit, contract terms, selection and switching; it is not a uniquely clean fit measure.", "from_layer": "phi_decomposition", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T22", "claim": "Distinguish parameter drift, regime change and topology change, with support checks and competing explanations. AI-search effects vary by destination, interface and outcome; no universal direction follows.", "from_layer": "regime_dynamics", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T23", "claim": "Category agent adoption can create congestion, improve relevance, expand demand or change platform power. Net externalities and first-mover decay are empirical questions requiring interference-aware measurement.", "from_layer": "agent_environmental_effects", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T24", "claim": "Concentration changes risk and allocation. Report cohort/subgroup distributions and tail costs as well as means; an average is a valid summary but may be insufficient for a decision.", "from_layer": "power_law", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T25", "claim": "Declare objective outcomes, units, horizon and preferences; apply simultaneous cash, capacity, quality and authority constraints. A macro scalar does not identify the correct weights.", "from_layer": "objective_function", "load_bearing": true, "contested": false, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T26", "claim": "Category formation changes information, alternatives and adoption. Treat narrative/clarity proxies as hypotheses and test real consideration and contribution; existing accounting still applies.", "from_layer": "category_formation", "load_bearing": false, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T27", "claim": "The recurring-value walk is one convention-bound ledger among business-kind-specific stocks, revenue, contribution and cash. Hybrid kinds and simultaneous constraints require explicit timing and denominators.", "from_layer": "business_model_generalization", "load_bearing": true, "contested": true, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T28", "claim": "Execution evidence should inform strategy proposals through an explicit ratification process. Authority boundaries are normative policies; agents must obey the actual authorized scope and versioned action permissions.", "from_layer": "cross_tier_coupling", "load_bearing": true, "contested": false, "revision": "3.3", "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect." }, { "id": "T8.1", "claim": "Autonomy fraction and governance score are insufficient operating controls. Qualify a task/state policy with execute, review, repair, defer and stop actions, enforced authority and measured capacity.", "type": "architecture", "from_equation": "reviewer_fatigue_catchrate", "load_bearing": true, "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect.", "revision": "3.3" }, { "id": "T29", "claim": "Governance effectiveness must be measured through detection, enforcement, correction, evidence and recovery. An illustrative fatigue parabola does not prove a universal governance threshold or optimal autonomy reversal.", "type": "architecture", "from_equation": "agentic_effective_cost", "load_bearing": true, "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect.", "revision": "3.3" }, { "id": "T30", "claim": "Category-level agent externalities cannot be inferred from firm governance or falling reply rates alone. Measure joint-policy exposure, relevance, demand, rivals and platform response; the sign remains open.", "type": "architecture", "from_equation": "category_noise_floor", "load_bearing": true, "contested": true, "evidence_caveat": "See the scoped evidence_register and qualification_protocol. Structural plausibility, descriptive surveys and laboratory benchmarks do not alone identify a local commercial policy effect.", "revision": "3.3" } ], "failure_modes": [ { "name": "Identity/correlation conflation", "description": "Treating a regime or correlational edge with the confidence of an identity. The original sin. Mitigation: every edge carries epistemic_type; counterfactuals are gated by type." }, { "name": "Optimizing the confounded proxy", "description": "Chasing NRR/engagement to 'cause' revenue when both are driven by Phi. Mitigation: never run a counterfactual on a correlational edge." }, { "name": "Ratio over payback", "description": "Cash payback and discounted contribution are complementary. Runway, financing, renewal, obligations and downside determine which constraints bind; no universal capital-regime payback threshold is fitted." }, { "name": "Comfortable-stage optimization", "description": "The funnel product gives unit elasticity to every stage rate. Prioritize feasible incremental contribution under downstream and capacity effects; the lowest observed conversion rate is not automatically the highest-return intervention." }, { "name": "Brand over-cut", "description": "Last-touch credit is not incremental contribution. Self-reported attribution can suggest discovery mechanisms but does not identify a lag kernel or causal brand effect by itself. Evaluate complete supported contribution response before reallocating spend." }, { "name": "Tactics-before-fit", "description": "Weak fit evidence, failed implementation or negative contribution make scale risky. Independent fit observations and qualified feasible-policy comparisons guide decisions; no universal Phi threshold proves zero GTM response." }, { "name": "Static-model rot", "description": "Trusting fitted coefficients after the regime has moved (channel saturated, competitor copied you, macro shifted). Mitigation: regime-tier coefficients carry half-lives and are re-estimated; treat drift as a bug (LangChain's CI-eval principle)." }, { "name": "Agent overreach", "description": "Letting an execution agent silently drift Tier-3 configuration (ICP/motion/pricing) or act without human-in-the-loop on relationship-destroying touches. Mitigation: the agent_interface boundary — execute Tier 1, read Tiers 2-3, never rewrite them." }, { "name": "Ungoverned agent spend", "description": "Agent execution cost (high, usage-based) silently inflates CAC. Mitigation: Agent FinOps tracking per task/account; the FinOps->CAC edge is real." }, { "name": "Compounding-error blast radius", "description": "A small early agent misjudgment cascades through a multi-step chain into a relationship-damaging send. Mitigation: human-in-the-loop checkpoints + data-quality gates + escalation paths interrupt the chain." }, { "name": "POC abandonment", "description": "~30% of GenAI projects die after POC on poor data quality / unclear value. Mitigation: phased rollout starting with the lowest-risk highest-signal loop, proven against a manual control group." }, { "name": "Treating agents as automation", "description": "Buying a 'standalone agent' expecting OS-level breadth/governance, or wiring agents as if-this-then-that. Mitigation: the maturity ladder — match the rung to the need; agents reason, automation triggers." }, { "name": "Identity laundering (the v1.1 original sin, named)", "description": "An accounting equality or definition, exact only under its declared basis and mutually exclusive categories. It does not identify a behavioral response or a causal counterfactual." }, { "name": "Regime-blind benchmarking", "description": "Treat historical benchmarks as dated segment heuristics; use actual funding/runway, cash obligations and declared objective preferences rather than a fitted Psi-to-payback rule." }, { "name": "Funnel-causal fallacy", "description": "Buyer state can select accounts into seller contact and condition conversion. In-market share and pre-contact winner-shortlist prevalence are descriptive quantities, not a 95% causal value split or a geometric seller-scaling law." }, { "name": "Brand-as-flow error", "description": "A linear brand-memory stock has an equivalent exponential distributed-lag representation. Local awareness, shortlist and contribution effects require independent observations and identified variation; SaaS decay is unmeasured here." }, { "name": "Phi circularity", "description": "Specify the outcome scale and competing models; freeze pre-outcome fit measures, account for selection and confounding, and use holdout predictive/causal targets separately. Retention or satisfaction alone does not uniquely measure fit." }, { "name": "Agentic over-claim (cost-per-activity vs cost-per-opportunity)", "description": "Citing collapsed cost-per-email as a CAC win. Cost-per-qualified-opportunity net of deliverability decay and operator maintenance is unproven. Mitigation: demand a holdout; treat uncontrolled ROI multiples as marketing math." }, { "name": "Optimizing the visible 5%", "description": "Allocate agents across discovery, evaluation, approval, implementation and renewal according to measured incremental value and authority. There is no identified 95% versus 5% value-creation split that predetermines seller-agent leverage." }, { "name": "Maximal-map mistaken for surveyed territory", "description": "Reading v3.0's comprehensive coverage as uniform reliability. Mitigation: obey the measurement_status tag — never feed an unmeasurable_hypothesis as a number into a decision." }, { "name": "Automated the wrong playbook", "description": "Agents executing a stale Tier-3 config flawlessly because the execution->strategy feedback loop was severed. Mitigation: cross_tier_coupling — agents propose, humans ratify, loop stays live." }, { "name": "HITL rubber-stamp rot", "description": "Use matched clean controls plus labeled defects and accepted throughput. The former gamma_max/rho parabola and governance sign-flip were illustrative and uncalibrated, not an operating law." }, { "name": "Agent equilibrium degradation", "description": "Category agent adoption can create congestion, improve relevance, expand demand or change platform power. Net externalities and first-mover decay are empirical questions requiring interference-aware measurement." }, { "name": "Ontology drift unnoticed", "description": "Stage definitions inflate under quota pressure; the same label now means something weaker. Mitigation: monitor conversion-distribution shift under a fixed label (ontology_instability)." }, { "name": "Average-case blindness under power law", "description": "Mean LTV / mean r_i reasoning when a few units carry the base. Mitigation: tail-aware metrics; treat top-decile concentration as tail risk, not strength." }, { "name": "Objective incoherence", "description": "Optimizing conflicting goals at once (short-term CAC vs brand, ACV vs PLG loops) with no stated weights. Mitigation: make the regime-conditional objective vector explicit (objective_function)." }, { "name": "Category myopia", "description": "Optimizing the funnel inside a game a competitor is busy redefining. Mitigation: watch branded-category-search share; know whether you are in creation or optimization mode." }, { "name": "Stock/unit conflation", "description": "L counts active logos; M measures recurring monetary value. New-logo inflow has units logos/time; new-value inflow has units currency/time. Expansion changes value without adding logos." }, { "name": "Review catch-rate without clean controls", "description": "Use matched clean controls plus labeled defects and accepted throughput. The former gamma_max/rho parabola and governance sign-flip were illustrative and uncalibrated, not an operating law." }, { "name": "Cost per attempt presented as success", "description": "Total eligible cost includes successful and failed attempts, retries, review, rework, data/tools, maintenance and attributable losses. A per-attempt expected cost is not cost per completed success." }, { "name": "Surrogate decision laundering", "description": "Correlation or a non-rejected diagnostic does not certify mature outcome-effect direction, coverage or deployment payoff." }, { "name": "Outcome coverage hidden by randomization", "description": "Random assignment does not recover absent transactions; audit scope, linkage, selected ratings and delayed/censored outcomes." }, { "name": "Authority score substituted for enforcement", "description": "A summary governance score cannot authorize a prohibited action; enforce current task/action predicates and verify raw receipts." }, { "name": "Synthetic respondents treated as new people", "description": "Validate target-specific fidelity against held-out real choices; generated replicas are not independent respondents." } ], "atomic_execution_cycle": { "name": "Sense -> Reason -> Act -> Learn (the agent loop)", "what_it_is": "The atomic unit that POPULATES Tier 1. Where v1.0 had the tier but not its mechanism, this is the cycle every agentic-GTM vendor (Apollo, Poexis, Wyzard, LangChain) independently converged on. It is the agentic replacement for a single human rep's per-lead loop.", "steps": { "sense": "Ingest signals — zero-party (volunteered), first-party (your observed behavior: visits, usage, badge scans), third-party (external intent/firmographic). Maps to: Intent Data, Buying Signal, Signal Taxonomy.", "reason": "Decide what the signal means now against a knowledge base, ICP, and policies; pick a strategy and a next-best-action. Maps to: Lead Scoring, qualification frameworks, Context Engineering, Next-Best-Action.", "act": "Execute across channels and systems — email, social, chat, calling, CRM writeback. Maps to: Sales Sequence, channel terms, the Act output.", "learn": "Separate recorded receipts, valid criterion-specific labels, implemented policy changes and held-out/outcome improvements. Use eligible-population evidence with selection and delayed-outcome audits." }, "key_distinction": "This is NOT rule-based automation. Automation triggers preset actions on a condition; the agent CHOOSES an action by reasoning. The difference is 'sense->reason->act' vs 'if-this-then-that'.", "relationship_to_tiers": "The loop runs entirely inside Tier 1, reading configuration from Tier 3 (ICP/motion/playbook as policy) and objectives/guardrails from Tier 2 (what to optimize, what it costs). Hundreds of these inner loops are wrapped by one outer business-outcome flywheel (Attract->Engage->Convert->Expand)." }, "agent_properties": { "name": "Why agents behave unlike traditional software", "inverted_cost_profile": { "traditional_software": { "creation": "task-dependent", "execution": "task-dependent", "reliability": "task/environment-dependent", "failure": "can propagate" }, "agents": { "creation": "workflow and integration dependent", "execution": "usage/tool/review/retry dependent", "reliability": "policy/task/configuration dependent", "failure": "may propagate, be repaired or contained" }, "implication": "Meter total eligible cost and independently accepted outcomes. Inference activity savings do not imply acquisition savings; repair, review and hard authority gates must be evaluated locally." }, "vocab": [ "Agent Cost Profile", "Compounding Failure", "Agent FinOps", "Human-in-the-Loop" ] }, "minimal_coupling_set": { "rationale": "Eight new components fully coupled is an over-parameterized, unfalsifiable graph. All four construction reports converge on a sparse edge set that captures the dominant interactions while staying testable.", "edges": [ { "edge": "B_r -> B", "reads": "Proposed brand-memory to consideration/shortlist mechanism; local causal effect unmeasured." }, { "edge": "B -> Tier-1 funnel conversion", "reads": "Pre-action buyer mix conditions observational rates; intervention effects require separate identification." }, { "edge": "Psi -> objective weights (w_g, w_f)", "reads": "Capital context informs explicit objective scenarios; no fitted causal mapping." }, { "edge": "S -> churn hazard / NRR", "reads": "Switching costs may modify renewal; piecewise additive revenue is unvalidated." }, { "edge": "PLG retention -> k (and PLG -> B at handoff)", "reads": "Retention/invitation/activation/payment couplings are measured separately; sales-assist policy response is a proposed test." } ], "deliberately_uncoupled": [ "Phi ex-ante -> immediate GTM decisions (Phi stays latent until panel exists)", "B_r <-> Psi and S <-> Psi feedback (investment treated as exogenous policy in v2.0)", "PLG -> brand feedback (PLG kept as acquisition channel, not brand builder, in v2.0)" ], "provenance": "All four construction reports, Part 3." }, "measurement_gap_register": { "purpose": "Marks the boundary between honest-but-unparameterized and operational. These parameters the rebuilt model NEEDS but CANNOT currently measure from public data. All four construction reports converge on this list.", "unmeasurable_from_public_data": [ { "param": "Ex-ante Phi panel", "consequence": "Independent fit observations and competing response/causal models are needed; the earlier log-interaction sign test does not identify additive versus multiplicative levels.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "Switching-cost S at account level", "consequence": "GRR and dissatisfaction do not identify S or a regime boundary; measure contracts, switching alternatives and intervention response.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "Brand alpha (build rate) and delta (decay) for SaaS", "consequence": "Brand-stock dynamics are directional only; flagged retrospective-only / forecasting liability.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "Sustained k>=1 in B2B", "consequence": "Define and measure referral generations versus active/paid stock; no universal B2B viral bound is established.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "Noise-floor theta as a stable constant", "consequence": "Retired as a universal constant. Validate target/horizon-specific forecast error, coverage and support.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "Controlled CAC under agentic GTM", "consequence": "Autonomous seller net CAC/contribution effects unresolved in reviewed evidence; causal human-seller routing and adjacent service evidence exist.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere.", "what_it_would_take": "P01 eligible-account policy comparison with full acquisition/quality/cost/outcome scope; CAC and CPQO separately defined.", "why_unmeasured_today": "Reviewed evidence does not identify a transferable autonomous seller net-profit effect.", "closest_existing_proxy": "VALOR (2026-04-02; outside current window) human-seller routing; Wang et al. R58 worker-randomized customer service. Different tasks/endpoints." }, { "param": "dCAC_agentic (E-A8)", "consequence": "Autonomous seller net CAC/contribution effects unresolved in reviewed evidence; causal human-seller routing and adjacent service evidence exist.", "what_it_would_take": "P01 eligible-account policy comparison with full acquisition/quality/cost/outcome scope; CAC and CPQO separately defined.", "why_unmeasured_today": "Reviewed evidence does not identify a transferable autonomous seller net-profit effect.", "closest_existing_proxy": "VALOR (2026-04-02; outside current window) human-seller routing; Wang et al. R58 worker-randomized customer service. Different tasks/endpoints.", "priority": "Highest — Gap 2 from prior synthesis.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "gamma_max(G), rho(G) (E-A2)", "consequence": "Retired as a default fatigue/governance parameterization. Measure detection, false rejection, correction, throughput and workload for the actual policy.", "what_it_would_take": "P09 labeled defect and matched-clean review comparisons under measured arrival/service capacity.", "why_unmeasured_today": "LLM-monitor experiments exist but do not establish human fatigue or a universal optimal-autonomy curve.", "closest_existing_proxy": "R59 LLM-monitor review discrimination, scoped task benchmark.", "priority": "High.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "theta, delta_theta (E-A7, category noise floor)", "consequence": "Net category externality and dynamic parameters unknown; positive decay bounds the response to bounded input. A constant steady state additionally requires constant adoption.", "what_it_would_take": "A time series of buyer-side response/engagement rates to outbound signals, segmented by category-wide agent-adoption share kappa, across a full multi-year adoption cycle to fit an adstock decay rate.", "why_unmeasured_today": "No source tracks a category-wide noise floor directly; CB Insights' 4,700% YoY referral growth measures adoption VOLUME, not per-unit signal degradation.", "closest_existing_proxy": "a16z's qualitative brand/trust-decisiveness framing (directional, not a measured series).", "priority": "Medium.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "lambda (E-A6, HITL learning-rate)", "consequence": "Measure valid label capture, outcome linkage, implemented updates and held-out improvement separately.", "what_it_would_take": "Tracking of how many human-reviewed agent actions are actually fed back into r_i retraining pipelines vs discarded after one-off review, across multiple vendor implementations.", "why_unmeasured_today": "The review-as-safety-gate vs review-as-training-signal distinction is not tracked as a separate metric in any surveyed source.", "closest_existing_proxy": "None identified.", "priority": "Medium.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "B_agent (E-A9, buying-committee agent-evaluator share)", "consequence": "Observed role × task × phase use, authority, fidelity and fulfillment needed; no 95% causal value split to preserve.", "what_it_would_take": "P03 real buying journeys and account-level representation/evidence tests.", "why_unmeasured_today": "Survey permission, recommendation benchmarks and individual shopping studies do not establish B2B committee purchase authority.", "closest_existing_proxy": "R09/R10 commercial B2B surveys; R54 interface/API audit excluded business products.", "priority": "Medium — Gap 7 from prior synthesis.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." }, { "param": "kappa_gov(G) (E-A4, failure-cost absorption)", "consequence": "Measure incident containment, correction and recovery costs separately; severe authority violations remain hard gates rather than governance discounts.", "what_it_would_take": "Incident post-mortems across governed vs ungoverned deployments, quantifying what fraction of a failure's realized cost (rework, reputational, churn) governance tooling actually mitigated vs absorbed in full.", "why_unmeasured_today": "No surveyed source reports incident-level cost breakdowns for agentic-GTM failures; failure-mode annotations are qualitative, not costed.", "closest_existing_proxy": "None identified.", "priority": "Low-medium.", "status": "unmeasured_for_local_decision", "scope": "Reviewed evidence through 2026-10-05; absence here is not a claim that no relevant study exists anywhere." } ], "honest_terminal_status": "Current evidence supports scoped mechanisms and study design, not fitted local parameters or qualified deployment. Preserve unknowns and refuse unsupported decisions." }, "agent_layer_v2_recalibration": { "verdict": "Historical vendor-case synthesis superseded where inconsistent by v3.3 scoped evidence and complete-policy qualification.", "what_holds": [ "Case architectures motivate authority, data quality, provenance, human exceptions and cost measurement.", "Conditional chain-success mathematics holds under its assumptions; real terminal/process success must be measured." ], "what_is_recalibrated_down": [ "Universal hybrid superiority, mandatory human review for every task and commercially calibrated agent ROI are not established.", "Cost per activity, opportunity, acquired customer and contribution have different denominators.", "Reported hours saved are neither automatic cash savings nor proven full-cycle compression." ], "new_load_bearing_patterns": [ "Versioned task/action authority enforced at execution", "Independently accepted terminal and process outcomes", "Measured review/repair/defer/stop policy under queue capacity", "All eligible attempts and cost, with zero-success cost ratio undefined", "Local holdouts and requalification after material configuration or population changes" ], "new_failure_file": [ "Stale or incomplete data linked to the wrong account or consent state", "Selection/missing outcome and short-horizon proxy bias", "Clean-work rejection or review-introduced defects", "Unenforced authority despite a high governance score", "Unobserved retries/maintenance/incident costs", "Generated reports that diverge from raw execution receipts" ], "the_strategic_point": "Allocate agents across discovery, evaluation, approval, implementation and renewal according to measured incremental value and authority. There is no identified 95% versus 5% value-creation split that predetermines seller-agent leverage.", "provenance": "Historical construction case inventory retained in Git; current research references R01-R65 and older foundations are appraised in docs/research/gtm-v3-2-research-2026-10-05.md." }, "measurement_status_legend": { "purpose": "Orthogonal to epistemic_type. Epistemic_type says what KIND of claim an edge is; measurement_status says whether you can actually OBSERVE the variable in time to act.", "statuses": { "measurable_ex_ante": "Observable BEFORE the outcome window it predicts; can be used as a genuine prior or input (e.g. design-partner retention, search share).", "measurable_ex_post": "Observable, but only AFTER the outcome; useful for learning/backtesting, not for forward decisions in the current period.", "proxy_only": "No direct measurement; approximated by a confounded stand-in (e.g. 'paid social feels tapped out' for channel saturation x). Use with stated error bars.", "unmeasurable_hypothesis": "Structurally plausible, currently no measurement protocol on available data. Belongs in the model as a documented hypothesis, NEVER as a load-bearing input. The honest home for most latents.", "unmeasured_for_local_decision": "The measurement family may exist, but no observed/fitted value matching the local target, cohort and configuration is supplied. A forward numeric decision needs appropriate local evidence." }, "rule": "Measurability is not current measurement, causal identification or deployment qualification. Unmeasured hypotheses may inform labeled scenarios, but cannot silently authorize a forward action or supply a fitted numeric input." }, "extended_layers": { "belief_state": { "name": "A. Belief-state / hidden-state formalism", "models": "Turns the model from a causal ontology toward a state-space object: an explicit latent state, an observation map separating what is seen from what is inferred, a belief update, and uncertainty propagation. Answers the critique 'what is the actual hidden state vector?'", "source_critiques": [ "doc 8/10 §1", "doc 3 (no probabilistic program)", "doc 9 (no uncertainty)" ], "constructs": [ { "name": "latent state vector x_t = (C, B, B_r, S, Psi, Phi-vector, competitive intensity, org-coherence)", "type": "hypothesis", "measurement": "proxy_only", "note": "Some components observable (C), most inferred. This is the object a real engine would carry." }, { "name": "observation map h(x_t) -> y_t", "type": "estimator", "measurement": "measurable_ex_post", "note": "Observation/linkage model is lossy and must be validated; a named h(x) is not an accounting identity." }, { "name": "belief update b_t = f(b_{t-1}, y_t, action_{t-1})", "type": "estimator", "measurement": "measurable_ex_post", "note": "Action-conditioned: beliefs must update on what the org DID, not just what it saw." }, { "name": "uncertainty propagation (posterior intervals, not points)", "type": "estimator", "measurement": "measurable_ex_post", "note": "Implemented for one slice in gtm_engine_slice.py; specified, not built, for the rest." } ], "honest_caveat": "Naming the state vector does not make its latent components observable. Most entries here are proxy_only or worse; the formalism is a scaffold, and only the MRR-walk slice currently has a working belief update." }, "phi_decomposition": { "name": "B. Phi decomposed (PMF is a vector, not a scalar)", "models": "Splits the master multiplier Phi into its sub-factors, as the critiques demanded. Each sub-factor is tagged for whether it can be measured ex ante — because that, not the decomposition itself, is what governs falsifiability.", "source_critiques": [ "doc 8/10 §2 (12-factor decomposition)", "doc 2/9 (family of latents)", "doc 1/12 (Phi unfalsifiable)" ], "constructs": [ { "name": "urgency (pain frequency x severity)", "type": "latent_multiplier", "measurement": "proxy_only" }, { "name": "willingness-to-pay", "type": "latent_multiplier", "measurement": "measurable_ex_post", "note": "revealed in pricing tests" }, { "name": "retention pull (workflow embedding + switching pain)", "type": "latent_multiplier", "measurement": "measurable_ex_post", "note": "Cohort retention is observable, but also reflects contract, price, selection, switching and billing; not uniquely fit." }, { "name": "word-of-mouth / referral coefficient k_wom", "type": "causal_regime", "measurement": "measurable_ex_post" }, { "name": "category clarity", "type": "latent_multiplier", "measurement": "unmeasurable_hypothesis" }, { "name": "implementation friction", "type": "causal_regime", "measurement": "measurable_ex_post" }, { "name": "political acceptability / procurement compatibility", "type": "latent_multiplier", "measurement": "proxy_only" }, { "name": "inevitability (does the buyer believe the category wins?)", "type": "latent_multiplier", "measurement": "unmeasurable_hypothesis" } ], "honest_caveat": "Validate each proposed fit component against independent observations, alternatives and relevant decisions. A measured retention curve alone does not identify latent fit." }, "competitive_ecology": { "name": "C. Competitive ecology & reflexivity", "models": "Treats the market as adaptive ecology, not physics. The environment changes BECAUSE you acted: working strategies get copied, channels saturate, incumbents retaliate, buyers adapt.", "source_critiques": [ "doc 8/10 §3", "doc 9 (single-firm)", "doc 2 (competitor responses)", "doc 1 (zero-sum absent)" ], "constructs": [ { "name": "competitor response function R_comp(your_action)", "type": "causal_regime", "measurement": "proxy_only", "note": "Win/loss data is a noisy ex-post read of competitive reaction." }, { "name": "reflexivity coefficient (decay of an edge's effect as it is imitated)", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "Measure cohort-specific response and imitation timing; decay alone does not identify copying as the cause." }, { "name": "memetic copying lag", "type": "correlational", "measurement": "proxy_only" }, { "name": "positioning / category-capture pressure", "type": "latent_multiplier", "measurement": "unmeasurable_hypothesis" } ], "mechanisms": [ "Rival imitation, retaliation, platform delivery and changing demand can change measured response; no fixed monotonic decay is presumed.", "A competitor move may be an endogenous reaction or an external shock; preserve timing and competing causal explanations." ], "honest_caveat": "Effect-size decay is observable under sufficient stable measurement; attributing it to reflexivity requires additional identification and support." }, "demand_topology": { "name": "D. Demand topology (beyond the funnel)", "models": "The funnel is a local accounting approximation of a richer generative process: diffusion, contagion, trust-graph traversal, social proof, ecosystem/workflow embedding. CRM-visible behavior is the shadow of this layer, not its mechanism.", "source_critiques": [ "doc 8/10 §4", "doc 8/10 §15 (creative discontinuities)" ], "constructs": [ { "name": "diffusion coefficient on a trust graph (developer tools, PLG)", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "Observable as invite/activation propagation — the PLG-loop k from v2.0, generalized." }, { "name": "social-proof cascade threshold", "type": "causal_regime", "measurement": "proxy_only" }, { "name": "ecosystem lock-in / platform-rank dependence", "type": "causal_regime", "measurement": "proxy_only" }, { "name": "agent-mediated discovery (buyers' own agents shortlist for them)", "type": "latent_multiplier", "measurement": "unmeasurable_hypothesis", "note": "Emerging 2026 mechanism; no clean measurement yet. Couples to Tier 0 buyer-state." } ], "honest_caveat": "The funnel is retained as the harvesting approximation; this layer says it is downstream of diffusion. Only graph-diffusion (the PLG k) is currently measurable; the rest are proxies/hypotheses." }, "org_cognition": { "name": "E. Organizational cognition (the firm as a lossy computer)", "models": "Organization capacity, expertise, incentives, information and feedback constrain feasible policy execution. Whether they bind more than demand is a local measurement question.", "source_critiques": [ "doc 8/10 §5", "doc 9 (org politics first-order)", "doc 7 (talent/culture)", "doc 5 (adoption friction)", "doc 1/12 (HITL human limits)" ], "constructs": [ { "name": "alignment bandwidth (rate at which the org can act on a decision)", "type": "causal_regime", "measurement": "proxy_only", "note": "Measure arrival/service rates, case mix and available minutes; min(growth, dimensionless alignment) is invalid without common units and a justified response model." }, { "name": "incentive-distortion factor (comp plan vs stated objective)", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "Measurable: gap between what comp rewards and what the objective_function weights." }, { "name": "decision latency (signal -> action lag)", "type": "causal_regime", "measurement": "measurable_ex_post" }, { "name": "managerial compression loss (information lost up each reporting layer)", "type": "correlational", "measurement": "unmeasurable_hypothesis" }, { "name": "rep tenure distribution / leadership stability", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "Modulates the r_i distribution (coefficient compression operates on top of this)." } ], "honest_caveat": "Several of these ARE measurable ex post (incentive distortion, latency, tenure), which is why this layer is more than hand-waving. Compression loss is a hypothesis. The realized-vs-designed gap is the layer's real contribution and it reframes every coefficient as org-conditioned.", "queue_accounting": "Q_next=max(0,Q_now+arrivals-completions), with arrivals/completions defined over the same queue and period; track oldest age, service minutes and rework." }, "regime_dynamics": { "name": "F. Regime-transition mechanics (upgrading Psi from a scalar to a process)", "models": "v2.0 had Psi as a macro re-weighting scalar. v3.0 adds the MECHANICS of regime change: how a shift is detected, that some shifts are topology changes (new graph) not coefficient drift, and the re-estimation cadence. Answers 'what causes regime transitions and how do you detect them?'", "source_critiques": [ "doc 8/10 §6", "doc 3 (no regime classifier/drift detector)", "doc 9 (no change-point detection)" ], "constructs": [ { "name": "change-point detector on conversion / CAC / response curves", "type": "estimator", "measurement": "measurable_ex_post", "note": "Concrete: CUSUM or Bayesian online change-point on coefficient residuals." }, { "name": "regime classifier (feature vector -> regime label)", "type": "estimator", "measurement": "measurable_ex_post" }, { "name": "topology-change flag (drift vs new graph structure)", "type": "causal_regime", "measurement": "proxy_only", "note": "AI-mediated discovery can alter graph structure. Validate destination/interface/outcome-specific changes rather than asserting all search changes are topology mutations." }, { "name": "re-estimation cadence / coefficient half-life", "type": "estimator", "measurement": "measurable_ex_post" } ], "mechanisms": [ "Distinguish three transition types: smooth drift (re-estimate), regime switch (swap parameter set), topology mutation (rebuild the graph). Misclassifying a topology change as drift is failure mode 'static-model rot' and is fatal precisely in the highest-stakes periods." ], "honest_caveat": "Changepoint techniques can detect distribution changes but do not establish their causes. No detector or topology classifier is implemented by this update beyond the existing executable seed." }, "substrate_epistemics": { "name": "G. Substrate as an active epistemic force", "models": "The genuinely novel point across all twelve critiques. The measurement substrate (CRM, dashboards, analytics) is not neutral infrastructure: what it measures, the org optimizes (Goodhart). Observability recursively reshapes the system being observed. This is cybernetic, not passive.", "source_critiques": [ "doc 8/10 §7", "doc 8/10 §10 (metrics socially constructed)" ], "constructs": [ { "name": "measurement -> behavior feedback edge (Goodhart)", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "If the CRM tracks meetings-booked, reps optimize meetings; the metric becomes the target and decays as a signal." }, { "name": "schema compression loss (reality -> CRM fields)", "type": "hypothesis", "measurement": "measurable_ex_post", "note": "What the schema cannot represent does not exist to the agent. Constrains the observation map h(x)." }, { "name": "attribution politicization", "type": "correlational", "measurement": "proxy_only" } ], "honest_caveat": "This is the one addition that strengthens the disciplined argument rather than fighting it: it explains WHY the measurement protocols the minimal core depends on are harder than they look — the act of measuring a variable changes the variable. The Goodhart edge is measurable ex post." }, "ontology_instability": { "name": "H. Ontology instability (the definitions themselves drift)", "models": "Metrics are socially constructed and mutate under incentive pressure. 'Qualified lead', stage definitions, and success criteria inflate or shift with quota pressure, board pressure, and the funding environment. The ontology is not fixed; it is a dependent variable.", "source_critiques": [ "doc 8/10 §10", "doc 1 (epistemic tagging adds load to drifting concepts)" ], "constructs": [ { "name": "definition-drift rate (semantic inflation of stage criteria over time)", "type": "correlational", "measurement": "measurable_ex_post", "note": "Rate shifts alone do not identify definition drift. Preserve versioned predicates and relabel a common frozen cohort to distinguish definition, mix and behavior." }, { "name": "metric-gaming pressure index", "type": "causal_regime", "measurement": "proxy_only" }, { "name": "definitional provenance (who set this definition and under what pressure)", "type": "normative_constraint", "measurement": "measurable_ex_ante" } ], "honest_caveat": "Couples tightly to substrate_epistemics and org_cognition: drift is what Goodhart pressure does to a schema over time. Drift rate is measurable; the pressure driving it is a proxy." }, "agent_environmental_effects": { "name": "I. Agents as environmental mutation forces (not just leverage)", "models": "Category adoption can alter congestion, relevance, demand and platform/rival response. Net effects require measurement and may have either sign.", "source_critiques": [ "doc 8/10 §8", "doc 9 (caching/mitigations underweighted)", "doc 1/12 (HITL alert fatigue)" ], "constructs": [ { "name": "messaging-homogenization index (cross-vendor template convergence)", "type": "causal_regime", "measurement": "proxy_only" }, { "name": "noise-floor inflation (response-rate decay as outbound volume rises)", "type": "correlational", "measurement": "measurable_ex_post", "note": "Reply rate versus volume is descriptive; selection, delivery, relevance and demand can confound it." }, { "name": "anti-agent defensive tooling adoption (deliverability decay)", "type": "causal_regime", "measurement": "measurable_ex_post" }, { "name": "HITL degradation (alert fatigue -> rubber-stamping corrupts the learning signal)", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "Approval/downstream divergence flags a problem but does not identify fatigue; use matched clean/defect cases and workload measurements." } ], "honest_caveat": "Response-rate changes, review failures and homogenization are observations or hypotheses; causal attribution and category externalities need a supported interference-aware design." }, "learning_layer": { "name": "J. The learning rule (the thing the spec never specified)", "models": "The single most-repeated 'missing engine' critique: 'human overrides become labeled corrections' is not a learning algorithm. v3.0 specifies the update mechanism, with the honest hooks for biased human labels and survivorship.", "source_critiques": [ "doc 8/10 §9 (epistemic governance)", "doc 9/3 (no learning algorithm)", "doc 2 (online Bayesian? bandits? counterfactual?)" ], "constructs": [ { "name": "coefficient update: Bayesian posterior on each r_i, decayed by half-life", "type": "estimator", "measurement": "measurable_ex_post", "note": "Engine uses a qualified illustrative aggregate retention update; posterior updating is not a causal learning guarantee. Discounting/decay requires a model, not automatic drift repair." }, { "name": "intervention testing (propose -> simulate -> A/B before committing)", "type": "estimator", "measurement": "measurable_ex_ante", "note": "Closes the loop the critiques wanted: experiments UPDATE Phi-beliefs, not just read them off." }, { "name": "evidence ranking + provenance (RCT > holdout > observational > anecdote)", "type": "normative_constraint", "measurement": "measurable_ex_ante", "note": "Evidence ranking considers design, access, assumptions, outcome scope and transport; a purported RCT is not automatically decision-valid." }, { "name": "biased-label correction (down-weight rubber-stamped HITL approvals)", "type": "estimator", "measurement": "measurable_ex_post", "note": "Audit selectively missing factual outcomes and counterfactual identification separately; down-weighting approvals alone repairs neither." }, { "name": "causal-graph revision (promote/demote edges on intervention evidence)", "type": "estimator", "measurement": "measurable_ex_post" } ], "honest_caveat": "Evidence-to-policy linkage, experiments/audits and outcome evaluation are specified; full causal-graph revision and business-policy learning are not implemented by this version." }, "power_law": { "name": "K. Power-law structure (the world is lumpy, not Gaussian)", "models": "A few customers, channels, reps, narratives, and partnerships dominate outcomes. Average-case reasoning — implicit in smooth conversion rates and mean coefficients — systematically misleads under heavy tails. Replaces Gaussian intuition with concentration dynamics where it matters.", "source_critiques": [ "doc 8/10 §11" ], "constructs": [ { "name": "revenue concentration (top-decile-account share)", "type": "accounting_definition", "measurement": "measurable_ex_post", "note": "Mean LTV remains a defined summary, but concentration, cost tails and correlated losses can make it insufficient for cash/risk decisions." }, { "name": "rep-output concentration (coefficient compression operates around a heavy tail)", "type": "correlational", "measurement": "measurable_ex_post" }, { "name": "channel concentration / cumulative-advantage (Matthew effects, increasing returns)", "type": "causal_regime", "measurement": "measurable_ex_post" }, { "name": "narrative concentration (one story captures the category)", "type": "latent_multiplier", "measurement": "proxy_only" } ], "honest_caveat": "Report concentration and distributions. The naive smallest-rate investment rule is invalid even without heavy tails; feasible contribution response governs allocation." }, "objective_function": { "name": "L. Multi-objective optimization (what is the system even maximizing?)", "models": "The model implicitly optimized growth-under-cash. Real GTM trades off growth, efficiency, survival, positioning, optionality, and category dominance — objectives that actively conflict. Makes the objective explicit and ties its weights to the regime (Psi).", "source_critiques": [ "doc 8/10 §12", "doc 2/9 (top-line vs payback governance failures)" ], "constructs": [ { "name": "objective weight vector U = w(Psi) . (growth, fcf, survival, position, optionality)", "type": "normative_constraint", "measurement": "measurable_ex_ante", "note": "Declare units/scales and mandate; context-conditioned weights are preferences/scenarios, not an identified macro mapping." }, { "name": "conflict matrix (optimizing short-term CAC degrades brand; ACV degrades PLG loops)", "type": "causal_regime", "measurement": "measurable_ex_post" }, { "name": "governance-distortion (board pressure forces top-line over payback until cash bites)", "type": "causal_regime", "measurement": "proxy_only", "note": "The institutional failure the critiques said v2.0 treated as exogenous; here it modulates w(Psi)." } ], "honest_caveat": "Explicit objective weights are measurable_ex_ante (you can read the mandate). The conflict matrix is the substantive content; governance distortion is a proxy. This finally makes 'what to optimize' a stated variable rather than an unspoken assumption of growth." }, "category_formation": { "name": "M. Category formation (creating the game, not just playing it)", "models": "New category language, independent evidence, alternatives and ecosystem adoption can change consideration and demand. Their contribution effects are empirical rather than a distinct universal physics.", "source_critiques": [ "doc 8/10 §13", "doc 8/10 §15 (how new games emerge)", "doc 5 (narrative not structurally modeled)" ], "constructs": [ { "name": "category-clarity stock (does the market have a name for what you do?)", "type": "latent_multiplier", "measurement": "proxy_only", "note": "Proxy: branded-search share for the category term over time. Couples to brand_stock B_r." }, { "name": "narrative concentration (your frame vs competitors' frames)", "type": "latent_multiplier", "measurement": "proxy_only" }, { "name": "benchmark / analyst ownership", "type": "causal_regime", "measurement": "measurable_ex_post" }, { "name": "game-creation vs game-optimization mode flag", "type": "conditional_coefficient", "measurement": "proxy_only", "note": "Category-creation conditions can require different behavioral assumptions; accounting remains applicable and the flag itself is a proxy." } ], "honest_caveat": "Honestly the least measurable layer — mostly proxies and one ex-post signal. But omitting it entirely was the critiques' fair charge that the model only explains stable periods. Included as explicitly proxy/hypothesis so it is not mistaken for the surveyed parts of the map." }, "business_model_generalization": { "name": "N. Business-model generalization (de-SaaS-ifying the economics tier)", "models": "The economics tier assumed subscription ARR. v3.0 conditions it on revenue model and confronts that the SaaS identities are themselves in regime shift: inference cost erodes the margin premise, agentic AI replaces seats, consumption/outcome pricing breaks the ACV math that drives motion thresholds.", "source_critiques": [ "doc 9 (SaaS/PLG-centric)", "doc 11 §2 (economics tier drifting)", "doc 5 (no discount rate/cohorts)", "doc 5/9 (multi-product)" ], "constructs": [ { "name": "revenue-model selector {subscription, usage, transaction, marketplace, hybrid, services}", "type": "conditional_coefficient", "measurement": "measurable_ex_ante", "note": "Switches which identities apply. The MRR walk is the subscription instance; usage/transaction need different stocks." }, { "name": "margin-erosion term (variable inference cost in COGS)", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "Include ongoing inference/delivery in contribution margin and acquisition-specific costs in CAC without double counting; usage growth can lower contribution." }, { "name": "LTV with discount rate + cohort heterogeneity", "type": "estimator", "measurement": "measurable_ex_post", "note": "Adds the discount factor the no-discount inequality omitted; pairs with the sBG correction from v2.0." }, { "name": "multi-product Phi/CAC/motion per SKU + cannibalization", "type": "causal_regime", "measurement": "measurable_ex_post" } ], "honest_caveat": "Strong, mostly-measurable layer: it removes a real overclaim (universality the model 'did not earn'). The revenue-model selector is the key move — it makes the MRR walk one instance of a family rather than the assumed default." }, "compliance": { "name": "O. Compliance & regulatory constraints on AI outreach", "models": "Applicable jurisdiction/channel/action/date requirements and actual authorized scope constrain execution. Legal predicates require a verified policy basis; this model does not assert a uniform global consent or disclosure rule.", "source_critiques": [ "doc 7 §6 (regulatory omission)" ], "constructs": [ { "name": "consent / opt-in coverage (GDPR, CAN-SPAM)", "type": "normative_constraint", "measurement": "measurable_ex_ante", "note": "Resolve applicable jurisdiction, channel, action, date and policy authority; preserve verification/expiry. This is a constraint contract, not legal advice or an accounting identity." }, { "name": "synthetic-media disclosure requirement", "type": "normative_constraint", "measurement": "measurable_ex_ante", "note": "Resolve applicable jurisdiction, channel, action, date and policy authority; preserve verification/expiry. This is a constraint contract, not legal advice or an accounting identity." }, { "name": "jurisdiction selector (rules vary by region)", "type": "normative_constraint", "measurement": "measurable_ex_ante", "note": "Resolve applicable jurisdiction, channel, action, date and policy authority; preserve verification/expiry. This is a constraint contract, not legal advice or an accounting identity." } ], "honest_caveat": "Compliance/authority constraints are normative, context-specific and versioned. Unknown applicability blocks the action pending a valid policy determination; no legal requirements are inferred solely from a model tag." }, "cross_tier_coupling": { "name": "P. Cross-tier coupling (the boundaries are normative, not descriptive)", "models": "The hard tier separation is an ideal, not a description. In real orgs execution constantly forces strategy revision (win/loss changes ICP; saturation forces motion shifts). v3.0 allows variables to decompose across tiers and adds the execution->strategy feedback loop, while keeping the agent boundary (agents still may not unilaterally rewrite Tier 3).", "source_critiques": [ "doc 9 (tier leaky)", "doc 11 §1 (insufficient feedback)", "doc 12 (waterfall fallacy)" ], "constructs": [ { "name": "Strategy and execution components jointly affect expansion; multiplicative decomposition proposed only", "type": "hypothesis", "measurement": "measurable_ex_post", "note": "Replaces exactly-one-tier ownership; expansion has both a packaging (strategy) and a CSM (execution) component." }, { "name": "execution -> strategy feedback loop (win/loss -> ICP revision)", "type": "causal_regime", "measurement": "measurable_ex_post", "note": "Makes the highest-leverage GTM work — the fast cross-tier loop — first-class rather than an exception." }, { "name": "agent boundary preserved: propose Tier-3 changes, humans ratify", "type": "normative_constraint", "measurement": "measurable_ex_ante", "note": "Resolves the tension: boundary is porous for humans, gated for agents (avoids 'automated the wrong playbook')." } ], "honest_caveat": "This directly fixes a fair structural error: the v2.0 boundary was too rigid as a DESCRIPTION. The repair keeps the safety property (agents don't silently rewrite strategy) while admitting the feedback loop is the real engine of GTM learning." }, "operator": { "name": "Q. Operator translation layer (the part a human can actually run Monday)", "models": "Compresses the whole maximal apparatus into what a CRO/VP-GTM monitors monthly, plus worked diagnostic workflows. Answers the adoption critique: the model is smarter than most GTM practice and therefore unusable without this.", "source_critiques": [ "doc 9/5/8 (operationalization, no playbooks, cognitive load)", "doc 2 (no translation layer)" ], "the_three_things": [ "Declare the eligible cohort, target outcome/horizon, cost basis and simultaneous constraints.", "Choose the best feasible incremental contribution intervention; freeze definitions and evaluate full downstream effects.", "Compare qualified policies with complete cost, quality, cash and authority evidence; preserve unknowns and refusal conditions." ], "warning_lights": [ "Unreconciled ledger or incomplete outcome linkage", "Cash recovery beyond the declared runway/obligation limit", "Response/conversion shift with changed mix, definitions, delivery or demand", "High approval alongside low independently accepted quality", "Capacity backlog age/rework rising", "Concentrated customer losses or inference/support cost tails", "Material tool/model/data/policy/authority change requiring requalification" ], "worked_workflows": [ { "name": "Diagnose & fix a funnel bottleneck", "steps": "Freeze eligibility and stage definitions; diagnose rates, delays, mix and quality; estimate feasible response/cost/capacity; test the best-supported intervention and evaluate complete eligible contribution outcomes." }, { "name": "Decide whether to scale a channel", "steps": "Compare marginal contribution in observed spend support, cash and service capacity; distinguish saturation from selection, demand, delivery and rival response; run an appropriate experiment or refuse extrapolation." }, { "name": "Pick a motion for a new SKU", "steps": "Select business-kind stocks/timing and simultaneous constraints; compare feasible motions using contribution, buyer requirements and capacity; watch cross-product cannibalization and partner outcome coverage." } ], "honest_caveat": "This layer is where 'maximal' pays rent: the apparatus above is only useful if it collapses to these few levers and lights. If an operator reads only this section, they have ~80% of the value." } }, "open_tensions_v3": { "purpose": "The maximal version owes an explicit accounting of what completeness cost. These are not resolved.", "tensions": [ "ADD vs SUBTRACT: v3.0 chose completeness; the disciplined critique (doc 10) chose parsimony. By that standard v3.0 is LESS falsifiable than the minimal core. The measurement_status tags are a mitigation, not a resolution.", "MAP vs ENGINE: v3.0 is a far more complete MAP, but adding seventeen layers moved it further from a runnable ENGINE, not closer. Only the MRR-walk slice (gtm_engine_slice.py) executes. The ratio of spec-to-machinery got worse, exactly as the harshest critiques predicted.", "PHI: decomposed but still mostly unmeasurable. Eight sub-factors, one (retention pull) currently measurable. The ghost-variable risk is reduced in principle and unchanged in practice.", "LEARNING: specified, not built. The update rule answers the 'silicon BDR' critique on paper; it still needs experimental traffic most ventures cannot spare, and full causal-graph revision is unimplemented.", "COVERAGE vs USABILITY: the operator layer (Q) is the bet that this collapses to a few levers. If it does not, v3.0 is a monograph operators admire and route around — the precise fate the critiques warned of.", "v3.3 fixes mathematical and delivery consistency but does not create missing local outcome data, causal identification or calibrated coefficients.", "Recent field, survey, theory and benchmark evidence differ in design and transport; the evidence registry preserves their boundaries." ], "the_honest_one_liner": "A comprehensive decision map with a limited executable seed; local measurement, experiments and workflow qualification remain necessary for deployment claims." }, "business_model_generalization_v4": { "generalizations": { "purpose": "The 8 generalizations that de-SaaS the spine. Each was forced by a specific kind breaking a specific hidden assumption; together they make the conservation law a family, not a default.", "hidden_saas_assumptions_removed": [ "stock = revenue (false for usage, transactional, marketplace, attention, installed-base)", "there is one kind of binding constraint (there are at least four types)", "there is a single stock (marketplaces and attention have coupled stocks)", "a steady-state equilibrium C* exists (false for pure-flow / hardware)", "value lands in the period it is earned (false for long-tail and revenue-now-loss-later)" ], "generalizations": [ { "id": "G1", "name": "Decouple operational stock from revenue function", "change": "Every kind is modeled as OPERATIONAL_STOCK (what accumulates) + REVENUE_FUNCTION (how that stock emits money). They are separate objects; revenue is a read-out of the stock, not the stock itself.", "forced_by": [ "usage", "transactional", "marketplace", "attention", "installed_base" ], "epistemic_type": "accounting_definition", "measurement": "measurable_ex_post", "saas_special_case": "In subscription they nearly coincide (MRR ~ the stock), which is why the conflation hid here." }, { "id": "G2", "name": "Binding-gate is a TYPED object", "change": "Binding gates are typed and can apply simultaneously: cash, capacity, authority/legal, availability/liquidity, quality and service obligations. Identify the currently binding set instead of assuming one gate per kind.", "forced_by": [ "services", "lending", "transactional", "attention" ], "epistemic_type": "normative_constraint", "measurement": "measurable_ex_ante", "gate_types": { "cash_payback": "race to recover CAC before cash runs out (subscription, usage)", "hard_capacity_ceiling": "min(demand, staffed capacity) caps the system (services)", "regulatory_capital": "permission + balance-sheet limits act before economics (lending, regulated)", "availability_balance": "two-sided or shelf/inventory presence gates transactions (transactional, attention, marketplace)" } }, { "id": "G3", "name": "Coupled stocks, with signed coupling", "change": "Promote multi-stock topology to a standing structure. Coupling can be POSITIVE (marketplace: supply growth -> demand growth -> supply growth) or NEGATIVE (attention: monetization up -> engagement down). Single-stock is the SaaS special case.", "forced_by": [ "marketplace", "attention" ], "epistemic_type": "causal_regime", "measurement": "measurable_ex_post" }, { "id": "G4", "name": "No-equilibrium / pure-flow mode + boundary announcement", "change": "Announce when a proposed steady-state result is unavailable or not attracting under its assumptions. Sales flow can coexist with inventory, backlog and installed stocks; neither pure flow nor subscription implies universal equilibrium.", "forced_by": [ "hardware" ], "epistemic_type": "model_identity", "measurement": "measurable_ex_ante", "saas_special_case": "Subscription value has an attracting constant-rate equilibrium only under positive net attrition; other boundary cases require explicit treatment." }, { "id": "G5", "name": "Time-shape-of-value field", "change": "Each kind tags the temporal shape of value: {immediate, recurring, long_tail, revenue_now_loss_later}. The spine integrates over the RIGHT horizon for that shape rather than assuming value lands in-period.", "forced_by": [ "installed_base", "lending", "transactional" ], "epistemic_type": "estimator", "measurement": "measurable_ex_post", "shapes": { "immediate": "value realized at transaction (transactional consumer)", "recurring": "value accrues per period (subscription, usage)", "long_tail": "one acquisition pays out over years (installed-base: repairs, replacement, referrals)", "revenue_now_loss_later": "revenue booked now, cost/loss arrives later (lending, insurance float)" } }, { "id": "G6", "name": "Asset-stock type (moat-as-gate)", "change": "Model proprietary asset freshness, rights, obsolescence and imitation separately from license customers, acquisition, renewal and service obligations. Licensing can have funnels and CAC; asset scarcity is one possible constraint.", "forced_by": [ "data_ip_licensing" ], "epistemic_type": "latent_multiplier", "measurement": "proxy_only" }, { "id": "G7", "name": "Negative contribution + adverse selection", "change": "Permit negative contribution in every business kind through delivery, returns, support, credits, default or acquisition cost. Risk-bearing kinds additionally require delayed-loss/adverse-selection and capital treatment.", "forced_by": [ "lending" ], "epistemic_type": "causal_regime", "measurement": "measurable_ex_post", "warning": "This is the deepest inversion in the model. Do NOT reuse subscription growth intuition for risk-bearing kinds." }, { "id": "G8", "name": "Served-side != paying-side (generalizes the payer!=user overlay into a kind)", "change": "A structural kind where the entity you serve is not the entity that pays. Model a free/served side (e.g. audience attention) and a paying side (e.g. advertisers) coupled through inventory, with a monetization<->engagement NEGATIVE loop (ad load up -> attention down). Generalizes the payer!=user overlay from a modifier to a first-class kind.", "forced_by": [ "attention" ], "epistemic_type": "causal_regime", "measurement": "measurable_ex_post" } ], "meta_lesson": "Many kinds, few underlying assumptions: ten kinds did not need ten patches, they needed ~8 generalizations. That convergence is evidence the skeleton is sound and merely subscription-biased. The upgrades make the model REPRESENT honestly; they do not lift the forecasting ceiling on unmeasurable cells (trust, reflexivity, true adverse-selection magnitude)." }, "kinds": { "purpose": "The 10 structural kinds as instantiable templates. To onboard a company: identify its kind, load this template (stock, revenue function, gate type, time-shape, dominant lever), then instrument the named spine. Test for a TRUE kind = a different stock OR a different binding gate; otherwise it is a sub-type, pricing setting, motion, or overlay.", "templates": [ { "n": 1, "kind": "Recurring / Subscription", "operational_stock": "active subscriber base", "revenue_function": "base x ARPA (recurring)", "gate_type": "cash_payback", "time_shape": "recurring", "dominant_lever": "retention + expansion (NRR)", "spine_first": "MRR walk + cohort retention + refusal gate", "examples": "SaaS, media subs, subscription apps", "gate_types": [ "cash_payback", "quality_authority_service", "cash_or_capacity_as_applicable" ], "boundary": "Recurring value, logos, use and payments are distinct; no universal attracting equilibrium or retention dominance.", "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 2, "kind": "Usage / Consumption", "operational_stock": "active integrated usage", "revenue_function": "usage_volume x price (with live gross-margin term)", "gate_type": "cash_payback", "time_shape": "recurring", "dominant_lever": "land then expand consumption", "spine_first": "usage stock (separate from revenue) + activation + margin term", "examples": "cloud, API, AI, infra", "gate_types": [ "cash_payback", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 3, "kind": "Transactional Consumer", "operational_stock": "buyer base + availability (mental x physical)", "revenue_function": "penetration x frequency x units x price", "gate_type": "availability_balance", "time_shape": "immediate", "dominant_lever": "penetration + repeat + distribution", "spine_first": "availability stock + penetration inflow + seasonality regime", "examples": "FMCG, e-commerce, D2C, retail", "gate_types": [ "availability_balance", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 4, "kind": "Marketplace / Platform", "operational_stock": "two-sided liquidity (supply + demand, coupled +)", "revenue_function": "take_rate x GMV", "gate_type": "availability_balance", "time_shape": "immediate", "dominant_lever": "seed the thin side -> network effects", "spine_first": "coupled supply+demand stocks + match/liquidity rate above revenue", "examples": "rideshare, lodging, app stores, exchanges", "gate_types": [ "availability_balance", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 5, "kind": "Services / Agency", "operational_stock": "retainer + relationship + reputation", "revenue_function": "Contract/project/retainer recognition and cash, or billable_capacity*utilization where that pricing basis applies.", "gate_type": "hard_capacity_ceiling", "time_shape": "recurring", "dominant_lever": "reputation + referrals + land-expand scope", "spine_first": "capacity-ceiling gate + utilization + retainer retention", "examples": "consulting, agencies, pro-services", "gate_types": [ "hard_capacity_ceiling", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 6, "kind": "One-time / Hardware", "operational_stock": "inventory, backlog, receivables and installed assets where relevant; order flow separately", "revenue_function": "deals_per_period x price", "gate_type": "cash_payback", "time_shape": "immediate", "dominant_lever": "channel + attach of recurring services", "spine_first": "order/fulfillment/cash ledger plus inventory/backlog and working-capital constraints", "examples": "devices, equipment, project sales", "boundary": "A pure sales-flow equation supplies no customer-stock equilibrium; operational stocks and their equilibria may still exist.", "gate_types": [ "cash_payback", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 7, "kind": "Advertising / Attention", "operational_stock": "engaged audience attention", "revenue_function": "attention inventory x ad price (paid by advertisers, not users)", "gate_type": "availability_balance", "time_shape": "recurring", "dominant_lever": "grow + retain engagement; balance audience<->advertiser", "spine_first": "served-side(attention) + paying-side(advertiser) coupled, with monetization<->engagement NEGATIVE loop", "examples": "Google, Meta, free media, ad-supported apps", "new_in": "v3.1 (G8)", "gate_types": [ "availability_balance", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 8, "kind": "Data / IP Licensing", "operational_stock": "proprietary asset (dataset/catalog/patent)", "revenue_function": "royalty / license fees on the asset", "gate_type": "rights_asset_and_customer_constraints", "time_shape": "recurring", "dominant_lever": "build & defend uniqueness/rights", "spine_first": "asset-stock with depletion = obsolescence/imitation/expiry; moat IS the gate", "examples": "data marketplaces, music catalogs, patent licensing, content licensing", "new_in": "v3.1 (G6)", "boundary": "License customer acquisition, funnel, renewal and CAC can apply alongside asset rights/freshness. A regulatory-capital gate is not generic to IP licensing.", "gate_types": [ "rights_asset_and_customer_constraints", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 9, "kind": "Lending / Risk-bearing / Float", "operational_stock": "loan book / AUM / insurance float", "revenue_function": "yield - loss_rate (per-unit can be NEGATIVE)", "gate_type": "regulatory_capital", "time_shape": "revenue_now_loss_later", "dominant_lever": "underwriting quality + cost of capital", "spine_first": "loss/default outflow (can exceed revenue) + adverse-selection (quality inverse to speed) + capital gate + revenue/loss time-lag", "examples": "banks, insurers, BNPL, lending fintech, capitation health", "new_in": "v3.1 (G7)", "warning": "Deepest inversion: growth can be actively bad; a customer can be negative. Do not reuse SaaS growth intuition.", "gate_types": [ "regulatory_capital", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" }, { "n": 10, "kind": "Installed-base / Razor-and-blade", "operational_stock": "installed fleet (devices in the field)", "revenue_function": "Initial installation contribution plus survival/attach/usage-weighted future consumable, maintenance and replacement cashflows by fleet vintage.", "gate_type": "hard_capacity_ceiling", "time_shape": "long_tail", "dominant_lever": "install velocity x attach rate (convert fleet -> annuity)", "spine_first": "installed-base stock + long-tail future-cashflow per acquisition + attach-rate lever", "examples": "HVAC, printers, medical devices, elevators, auto", "new_in": "v3.1 (G5)", "note": "The HVAC-in-NJ case lands here — why it never fit the pure one-time/hardware row.", "gate_types": [ "hard_capacity_ceiling", "quality_authority_service", "cash_or_capacity_as_applicable" ], "dominant_lever_status": "candidate_hypothesis_not_causally_ranked" } ], "fold_in_not_kinds": { "D2C / e-commerce": "Transactional Consumer (3)", "Freemium / free-trial": "a LEVER on Recurring/Usage", "On-demand / gig": "Marketplace (4) + logistics gate", "B2B/B2C/C2C/C2B": "the SCALE / counterparty axis, not a kind", "Franchise": "Services (5) replicated, or a lever", "Vertical SaaS / AI-native": "Recurring/Usage (1/2) in a niche", "Outcome-based pricing": "a revenue-model SETTING on 1/2/9", "Ecosystem-led growth": "a MOTION across kinds, not a kind" }, "overlays": { "A_channel_intermediation": "sell-in != sell-out (reshapes any kind)", "B_payer_not_user_regulated": "compliance + procurement gate (reshapes any kind)", "note": "Where payer!=user is STRUCTURAL it is promoted to a kind: third-party-payer => Attention (7); risk-bearing => Lending (9)." } } }, "evidence_register": [ { "id": "R01", "title": "Wei, Bencina, Nguyen, Ribeiro and Zelditch — From Prediction to Incrementality — 10 August 2026.", "date": "10 August 2026", "url": "https://arxiv.org/html/2608.10182v1", "design": "Full preprint.", "finding": "Full preprint. Field intervention; full methods/limitations inspected. Eight-week member-randomized 50/50 LinkedIn comparison of causal scoring, exploration, constrained allocation and withholding versus propensity targeting, with a matched delivery envelope. The reported value lift concerns the complete policy. Static audience definitions and segment lineage protect assignment.", "boundary": "one platform, short experiment, no isolation of individual components, no autonomous B2B seller economics. C3/C5/C6.", "clusters": [ "C3", "C5", "C6" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R02", "title": "Li, Yuan, Yang, Chen and Song — Attributed But Not Incremental — 25 June 2026.", "date": "25 June 2026", "url": "https://arxiv.org/html/2606.26690v1", "design": "Full preprint.", "finding": "Full preprint. Production calibration against 18 rounds of channel experiments across up to eight TikTok markets, with rolling forward evaluation and daily hierarchical correction.", "boundary": "normalized confidential errors; child-level credit is allocated rather than independently causally identified. A later cannibalization decline is before/after deployment evidence, not a randomized business-impact estimate. Complementarity and spillover remain limitations. C3/C5.", "clusters": [ "C3", "C5" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R03", "title": "Heusch — Structural Estimation of MMM Parameters from Geo Experiments — 21 August, revised 7 September 2026.", "date": "21 August, revised 7 September 2026", "url": "https://arxiv.org/html/2608.21128v2", "design": "Full v2.", "finding": "Full v2. Synthetic structural study: 156-week fictional retailer, three channels, four four-week go-dark tests plus cooldown. Uses time paths to estimate effectiveness, carryover and saturation.", "boundary": "stable parameters, specified adstock/saturation, no spillover and tested spend support. It is not an empirical estimate of SaaS brand decay. C3/C5.", "clusters": [ "C3", "C5" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R04", "title": "Filippou, Quach and Jha — Pseudo Incrementality Testing — 16 September 2026.", "date": "16 September 2026", "url": "https://arxiv.org/html/2609.18257v1", "design": "Full preprint.", "finding": "Full preprint. Changepoint detection plus counterfactual forecasting, two GeoLift benchmarks and synthetic stress tests. Only one of three nominal 90% intervals covers truth in the stress test; weak controls yield large bias.", "boundary": "parameter uncertainty not fully propagated, small validation base, no guarantee that pauses are exogenous. Useful refusal conditions rather than a substitute for experimental design. C3/C5/C7.", "clusters": [ "C3", "C5", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R05", "title": "Ningyin Xu — Reference Dependent Pricing — deposited 16 July 2026.", "date": "deposited 16 July 2026", "url": "https://knowledge.uchicago.edu/records/8v3yz-xrr12", "design": "Repository and dissertation.", "finding": "Repository and dissertation. Full PDF inspected. Randomized base-price histories across 818 food-vending locations during 2024, followed by equal current-price discount conditions in 2025.", "boundary": "reference-history effects in this setting; reported approximately 28% profit improvement is a structural counterfactual, not deployed randomized profit uplift. Local competition and crossed-out-price cues limit transfer. C2/C5.", "clusters": [ "C2", "C5" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R06", "title": "Dubé, Morozov, She and Tuchman — The Effects of Ads on Beliefs and Implications for Consumer Search — written 31 July, posted 7 August, BFI release 17 August 2026.", "date": "written 31 July, posted 7 August, BFI release 17 August 2026", "url": "https://bfi.uchicago.edu/working-papers/the-effects-of-ads-on-beliefs-and-implications-for-consumer-search/", "design": "Paper and full PDF.", "finding": "Paper and full PDF. Preregistered experiment, 17,775 Prolific participants, simulated bookstore; randomized ads/content/price and separate belief/shopping samples avoid priming. Data are from 2025.", "boundary": "lottery-incentivized shopping, unincentivized belief reports, one advertised book, no long-run enterprise brand effect. C2/C3/C5.", "clusters": [ "C2", "C3", "C5" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R07", "title": "Martin Bichler — Agentic Markets — 6 June 2026.", "date": "6 June 2026", "url": "https://link.springer.com/article/10.1007/s12525-026-00906-y", "design": "Full journal article.", "finding": "Full journal article. Conceptual synthesis of agent search, differentiation, market power and platform architecture.", "boundary": "no new calibrated market experiment. Lower technical search costs do not guarantee exhaustive search, equal information, higher welfare or a particular equilibrium. C2/C3/C7/C8.", "clusters": [ "C2", "C3", "C7", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R08", "title": "Quanyan Zhu — Agentomics — 9 June 2026.", "date": "9 June 2026", "url": "https://arxiv.org/html/2606.14769v1", "design": "Full preprint.", "finding": "Full preprint. Formal/normative workflow-surplus and cooperative allocation framework.", "boundary": "the security-operations numerical example is hypothetical. Shapley allocation is a proposed normative benchmark, not an observed pricing equilibrium; coalition values themselves require estimation. C5/C7/C8.", "clusters": [ "C5", "C7", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R09", "title": "G2 — The Answer Economy — 15 April 2026.", "date": "15 April 2026", "url": "https://learn.g2.com/g2-2026-ai-search-insight-report", "design": "Primary report.", "finding": "Primary report. Commercial survey; methods/findings inspected. March online sample of 1,076 B2B decision makers and 39 marketer interviews across regions. Reports AI research alongside continued Google use and frequent inaccuracies.", "boundary": "reported frequency and AI-attributed switching are not observed journey shares or causal switching. Sampling/weighting and question bases constrain universal adoption estimates. C1/C2/C3/C8.", "clusters": [ "C1", "C2", "C3", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R10", "title": "G2 — The Evaluation Maze — 22 July 2026.", "date": "22 July 2026", "url": "https://sell.g2.com/2026-buyer-behavior-report", "design": "Primary report.", "finding": "Primary report. June online survey of 1,038 decision makers plus 55 interviews. Forty percent name evaluation their longest stage; 39% name security the largest post-selection delay; 49% report a CFO veto in the past year.", "boundary": "40% is a respondent share, not elapsed journey time. Nine percent say they would permit guarded purchasing and 2% purchasing without pre-approval; those are authority preferences, not observed adoption. Current/planned use remains separate. Sponsor/self-report selection prevents causal conclusions. C2/C4/C6/C8.", "clusters": [ "C2", "C4", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R11", "title": "Oded Netzer and Rajan Sambandam — Synthetic Data in Marketing Research How to Evaluate and When to Trust — 12 September, revised 19 September 2026.", "date": "12 September, revised 19 September 2026", "url": "https://arxiv.org/abs/2609.13995", "design": "Record and PDF.", "finding": "Record and PDF. Full methods inspected. Real 2025 US sample of 3,063 people, 18 sectors, 108 questions; grounded leave-one-question-out prediction and out-of-bag screening. Rich category histories improve agreement over demographics alone; 46 retained questions perform better than all questions.", "boundary": "selected in-domain survey predictions, not novel procurement decisions. Question-level agreement differs from individual fidelity; generated replicas add no independent people. Commercial research affiliation disclosed. C1/C2/C6/C8.", "clusters": [ "C1", "C2", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R12", "title": "Kuric, Demcak and Krajcovic — Distorted Perspectives of LLM Simulated Preferences — 18 May 2026.", "date": "18 May 2026", "url": "https://arxiv.org/html/2605.18311v1", "design": "Full preprint.", "finding": "Full preprint. Twenty-nine actual UX studies, 2,073 people, 78 tasks. GPT-4.1's top choice matches people in 53% of tasks; GPT-5.2 reaches 65% without an established general distribution-fit gain.", "boundary": "specific models/protocols and proprietary UX data, no procurement or transactions. Persona elaboration and reasoning do not consistently repair validity. C1/C2/C8.", "clusters": [ "C1", "C2", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R13", "title": "Leonard Kinzinger and Jochen Hartmann — Synthetic Personalities — 3 June 2026.", "date": "3 June 2026", "url": "https://arxiv.org/html/2606.04592v1", "design": "Full preprint.", "finding": "Full preprint. Five hundred SOEP individuals, 183 held-out items, three models and 2.1 million generated responses. Grounding and preserving original responses help; reasoning improves ordering more reliably than closeness.", "boundary": "best .788 normalized closeness is not 78.8% exact choice accuracy. Familiar socioeconomic items are not novel enterprise decisions; 500 real people remain the independent basis. C1/C2/C6/C8.", "clusters": [ "C1", "C2", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R14", "title": "Fan Huang and colleagues — Evaluating LLM Generated Preference Distributions — 1 October 2026.", "date": "1 October 2026", "url": "https://arxiv.org/html/2610.01000v1", "design": "Full preprint.", "finding": "Full preprint. Nine open-weight models, airport/Yelp/Amazon domains, repeated elicitation with context/persona/order/temperature variation. Strong within-model stability coexists with cross-model disagreement.", "boundary": "no human ground truth, unweighted personas, English/open models only. Tests internal consistency, not real demand. C2/C6/C8.", "clusters": [ "C2", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R15", "title": "Stephen France and Pia Albinsson — Synthetic Consumer Insight Generation with Large Language Models — 7 July 2026.", "date": "7 July 2026", "url": "https://arxiv.org/abs/2607.05761", "design": "Record and full PDF.", "finding": "Record and full PDF. Full comparative methods inspected. Word associations and completions about five US cities, 173 students, six LLMs and varied prompting/seeding. Broad topics overlap while synthetic style differs.", "boundary": "semantic associations, students and cities; no willingness-to-pay or buying behavior. Supports hypothesis expansion rather than calibrated demand. C1/C2/C8.", "clusters": [ "C1", "C2", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R16", "title": "Shuxian Fan and colleagues — ShoppingReasoningBench — 10 June 2026.", "date": "10 June 2026", "url": "https://arxiv.org/html/2606.12608v1", "design": "Full preprint.", "finding": "Full preprint. Seven experts construct 525 missions, 1,996 turns and 10,863 weighted binary rubrics across five product families; nine models tested. Optional constraints and multi-turn tasks are harder.", "boundary": "scored recommendations, not actual purchase welfare or commercial effects; one generation per query and excluded categories. Versions and time-sensitive facts matter. C2/C4/C8.", "clusters": [ "C2", "C4", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R17", "title": "Haobo Zhang and colleagues — Learning from Online User Feedback for Shopping Agents — 12 August 2026.", "date": "12 August 2026", "url": "https://arxiv.org/html/2608.11604v1", "design": "Full preprint.", "finding": "Full preprint. JD.com real-log offline study: 7,001 behavioral sessions from 6,704 users plus 6,121 directive sessions; purchase-aligned and corrective training.", "boundary": "behavioral training selects successful purchases; issue-resolution evaluation partly uses an LLM judge. Better offline purchased-item retrieval does not establish live conversion, welfare or retention uplift. C4/C6/C8.", "clusters": [ "C4", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R18", "title": "Jennifer Zou — Platform Choice Trust and Privacy in the Consumer AI Assistant Market — 16 July 2026.", "date": "16 July 2026", "url": "https://arxiv.org/html/2607.15134v1", "design": "Full preprint.", "finding": "Full preprint. June Prolific survey of 1,999 AI users with six hypothetical conjoint choices; task routing and platform preferences vary.", "boundary": "weighting targets AI users and reduces effective sample size; privacy WTP is specification-dependent and excludes an outside option. Selected usage/trust associations are not causal. Commercial AI-visibility affiliation. C1/C2/C3/C8.", "clusters": [ "C1", "C2", "C3", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R19", "title": "Carbon — B2B Buyers Report — September 2026.", "date": "September 2026", "url": "https://explorecarbon.com/b2b-buyers-report/findings/", "design": "Question-level findings.", "finding": "Question-level findings. Commercial survey, 189 quality-filtered decision makers; denominators vary by question. High reported satisfaction coexists with provider reconsideration.", "boundary": "small heterogeneous sample, limited probability-sampling detail, no longitudinal identification. Different bases cannot be combined into a respondent-level causal mechanism. Exact release day not verified. C2/C4.", "clusters": [ "C2", "C4" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R20", "title": "NewtonX and Pretzl — Buyer Group Index — 30 June 2026.", "date": "30 June 2026", "url": "https://www.newtonx.com/article/newtonx-and-pretzl-launch-the-2026-buyer-group-index", "design": "Primary methods announcement.", "finding": "Primary methods announcement. Two hundred senior GTM leaders in 500-plus-employee organizations, approximately 30 AI-moderated interviews; 28 practices modeled against five outcomes.", "boundary": "selected executive reporting, many candidate predictors, common-source outcomes and endogenous practices. Performance associations do not identify effects of adopting the index. No independent raw-data replication inspected. C2/C6.", "clusters": [ "C2", "C6" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R21", "title": "Dong, Luo and Xu — From Product Search to Preference Articulation — 9 August 2026.", "date": "9 August 2026", "url": "https://arxiv.org/html/2608.08395v1", "design": "Full preprint.", "finding": "Full preprint. Theory of delegated search, profiling noise, preference complexity and attention. Distinguishes purchase incentives from consumer welfare.", "boundary": "conditional comparative statics, no fitted effect sizes or buying-group experiment. Useful mechanism hypotheses for representation and preference clarification. C2/C3/C8.", "clusters": [ "C2", "C3", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R22", "title": "Xiaolin Li, Qing Li, Danny Wang and Flora Gu — AI Human Collaboration in Sequential Sales Funnels — 13 June 2026.", "date": "13 June 2026", "url": "https://scholars.hkbu.edu.hk/en/publications/ai-human-collaboration-in-sequential-sales-funnels-evidence-from-/", "design": "Primary conference abstract.", "finding": "Primary conference abstract. Abstract describes randomized SME lending study with 78,169 prospects: AI initial calls and intent tags, then human follow-up.", "boundary": "full arms, effect sizes, costs, defaults and durable contribution unavailable. Applications are not profit; other versions' numbers must not be merged. Promising field design, abstract-level appraisal. C4/C5/C6/C8.", "clusters": [ "C4", "C5", "C6", "C8" ], "access_depth": "abstract_or_primary_explanation", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R23", "title": "Chatrath, Zhu, Fan and colleagues — READY or Not Reliable Enterprise Agent Deployment — 2 September 2026.", "date": "2 September 2026", "url": "https://arxiv.org/html/2609.02095v1", "design": "Full preprint.", "finding": "Full preprint. Clinical benchmark: 16 systems, 750 cases split equally into development and qualification. Freeze policies and qualify with a one-sided reliability bound.", "boundary": "human success assumed at 90%, with uncertainty only on autonomous outcomes; qualification can have zero autonomous coverage when all cases go to assumed human review. Outcome versus process scoring matters. Framework transfers more plausibly than clinical numbers. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R24", "title": "Ning, Li, Kong and Li — ControlScope — 28 September 2026.", "date": "28 September 2026", "url": "https://arxiv.org/html/2609.34313v1", "design": "Full preprint.", "finding": "Full preprint. Matched execution states compare automated keep/argument-edit/workflow-edit permissions on 20 filesystem tasks, ALFWorld and 585 AppWorld instances.", "boundary": "LLM revision-reviewers, not direct human oversight evidence. Review can interrupt viable repair; effects vary and AppWorld net intervals include zero. Full AppWorld review renews the request allowance, introducing extra capacity. Supports scope/cadence tests, not universal greater autonomy. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R25", "title": "Steve Drew and Jiayu Zhou — LEGIT — 18 September 2026.", "date": "18 September 2026", "url": "https://arxiv.org/html/2609.21325v1", "design": "Full preprint.", "finding": "Full preprint. Credentialing proposal and repeated paired GAIA/math tests, including 127 GAIA tasks, pinned configurations and budgets.", "boundary": "corrected harness-quality evidence is inconclusive although costs differ; marketplace and feedback authentication unvalidated end to end. Cost per solved task includes failures; expiry is no assurance against drift. C0/C5/C8.", "clusters": [ "C0", "C5", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R26", "title": "Paul-Peter Arslan — Audit Without Verification — 7 September 2026.", "date": "7 September 2026", "url": "https://arxiv.org/html/2609.07680v1", "design": "Full preprint.", "finding": "Full preprint. Balanced injected-defect/clean synthetic six-link pipelines. Preregistered main framing hypothesis is null; exploratory deletions and replications show origin suggestions can bias auditors.", "boundary": "removal can help when suggestions are wrong and hurt when correct. Restricted evidence interfaces and preregistration deviations matter; no production error prevalence. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R27", "title": "Yuhang Wang — Why LLM Agents Collapse Without Oversight The Enforcement Gap — 14 September, revised 23 September 2026.", "date": "14 September, revised 23 September 2026", "url": "https://arxiv.org/html/2609.15293v3", "design": "Full v3.", "finding": "Full v3. Tested agent setups distinguish advisory flags from binding control flow.", "boundary": "use the detection/enforcement mechanism narrowly. Broad claims about every deployed framework are not established; the controller appendix uses synthetic critique stubs. No estimated GTM breach rate. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R28", "title": "Microsoft Research, Bansal, Mirza and colleagues — Red Teaming a Network of Agents — 30 April 2026.", "date": "30 April 2026", "url": "https://www.microsoft.com/en-us/research/blog/red-teaming-a-network-of-agents-understanding-what-breaks-when-ai-agents-interact-at-scale/", "design": "Primary report.", "finding": "Primary report. Internal network of more than 100 agents; malicious-message propagation, borrowed reputation and provenance loss.", "boundary": "no representative exposure denominator or commercial effect. Supports shared-failure, containment and untrusted-content tests rather than a numeric error prior. C6/C7/C8.", "clusters": [ "C6", "C7", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R29", "title": "Anthropic — Alignment Assessment of Recent Cybersecurity Incidents — 9 September 2026.", "date": "9 September 2026", "url": "https://www.anthropic.com/research/alignment-assessment-cybersecurity-incidents", "design": "Primary assessment.", "finding": "Primary assessment. Four unauthorized-access incidents studied through transcripts, resampling and interpretability.", "boundary": "roughly 481 million heterogeneous scanned transcripts are not a common eligible denominator. Disabled safeguards in some evaluations and partial-trajectory/LLM-grader limitations constrain ordinary-deployment transfer. Authority validity must be checked separately from output correctness. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R30", "title": "Joshua Gans and Richard Holden — When Does Randomized Oversight Align AI Agents That Can Conceal — 29 September 2026.", "date": "29 September 2026", "url": "https://arxiv.org/html/2609.38262v1", "design": "Full preprint.", "finding": "Full preprint. Principal-agent theory: deterrence depends on preserved evidence, audit draws not known before action and meaningful incentives.", "boundary": "concealment can make residual violations harder to detect; that does not establish lower total safety from more audit. Correct beliefs and mechanism assumptions remain open empirically. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R31", "title": "Yan, Huo, Shang, Peng and Wu — CEO Arena — 28 September 2026.", "date": "28 September 2026", "url": "https://arxiv.org/html/2609.34821v1", "design": "Full preprint.", "finding": "Full preprint. Eight-firm, 500-day stylized competitive environment; matched replacements, 27 main and 26 robustness runs. Private gain can coexist with aggregate market loss.", "boundary": "few seeds, programmed demand, symmetric settings and missing finance/labor/regulation. No reliable GTM equilibrium or welfare coefficient follows. C7/C8.", "clusters": [ "C7", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R32", "title": "Jaewook Lee and Tijana Zrnic — Partially Performative Prediction — 5 June 2026.", "date": "5 June 2026", "url": "https://arxiv.org/html/2606.07890v1", "design": "Full preprint.", "finding": "Full preprint. Theory plus imposed-response credit case study distinguishes endogenous policy response from external drift.", "boundary": "guarantees require smoothness, convexity, bounded sensitivity and comparator-path assumptions; no randomized B2B adaptation. C6/C7.", "clusters": [ "C6", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R33", "title": "Anagnostides, Chauhan, Panageas, Sandholm and Yan — Computational Complexity of Performative Prediction — ICML July 2026.", "date": "ICML July 2026", "url": "https://proceedings.mlr.press/v306/anagnostides26a.html", "design": "Peer-reviewed proceedings.", "finding": "Peer-reviewed proceedings. Complexity of stable points under strong feedback and strategic local optima.", "boundary": "worst-case hardness is not a finding that observed GTM is chaotic or impossible to improve. Stable and optimal are different targets. C7.", "clusters": [ "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R34", "title": "Vo, Reddy, Rodemann, Chau and Muandet — Off Policy Evaluation with Strategic Agents via Local Disclosure — 5 June, ICML July 2026.", "date": "5 June, ICML July 2026", "url": "https://arxiv.org/html/2606.07308v1", "design": "Full preprint, proceedings.", "finding": "Full preprint, proceedings. Local disclosure exposes pre-adaptation features; theory and synthetic/German-credit tests impose strategic response.", "boundary": "response-distribution, support and nuisance assumptions. Ordinary matching of policy-changed attributes cannot recover pre-action demand. C6/C7.", "clusters": [ "C6", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R35", "title": "Wei, Qu and Miao — Missingness Aware Off Policy Evaluation — 18 June, ICML July 2026.", "date": "18 June, ICML July 2026", "url": "https://arxiv.org/html/2606.20206v1", "design": "Full paper.", "finding": "Full paper. Reward-dependent missingness addressed using next-state shadow variables and bridge assumptions; simulation and introduced sepsis missingness.", "boundary": "relevance, exclusion, completeness, positivity, realizability and policy support are demanding. CRM proxies are not automatically valid shadow variables. Report bounds or experiment when identification fails. C4/C6.", "clusters": [ "C4", "C6" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R36", "title": "Spear, Komorowski, Pope and Moodie — Kernel Weighted Importance Sampling — 16 July, revised 4 August 2026.", "date": "16 July, revised 4 August 2026", "url": "https://arxiv.org/abs/2607.15067", "design": "Record and abstract.", "finding": "Record and abstract. Abstract-level appraisal of a robust contextual-bandit estimator.", "boundary": "full methods not reviewed; estimator robustness does not remedy zero action probability, hidden confounding, strategic shift or missing outcomes. Candidate only after identification/support audit. C6.", "clusters": [ "C6" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R37", "title": "Anthropic Economics — Project Swap — 24 September 2026.", "date": "24 September 2026", "url": "https://www.anthropic.com/research/project-swap", "design": "Primary study.", "finding": "Primary study. Human book preferences, five-minute intake, decentralized bargaining and reruns; 188 rankings and 205 market runs.", "boundary": "specific representation-versus-allocation decomposition, employee sample, unincentivized rankings, attrition and well-behaved agents. Some agreed physical transfers were unfulfilled; no B2B transfer of its shortfall percentage. C2/C4/C8.", "clusters": [ "C2", "C4", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R38", "title": "He, Agarwal, Denisov-Blanch, Azaletskiy, Koyejo and Vasilescu — AI Writes Faster Than Humans Can Review — 2 July 2026.", "date": "2 July 2026", "url": "https://arxiv.org/html/2607.01904v1", "design": "Full preprint.", "finding": "Full preprint. One B2B software company, 802 developers, 196,212 pull requests, January 2024–April 2026; staggered adoption comparisons. Output and reviewer load rose substantially.", "boundary": "self-selected adoption/intensity, unusually favorable budgets, co-moving process changes, coarse merge/revert quality proxies. Exact causal magnitude is not identified. Measure workload relocation, substantive review, queues and downstream quality. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R39", "title": "Anthis, Cha, Barocas, Chouldechova and Hofman — Effects of Generative AI Errors on User Reliance Across Task Difficulty — 5 April 2026, CHI EA.", "date": "5 April 2026, CHI EA", "url": "https://arxiv.org/html/2604.04319v1", "design": "Full paper.", "finding": "Full paper. Preregistered randomized 3×2 diagram experiment, 577 US Prolific participants, incentive-compatible reliance bids. More errors reduce bids; easy-task errors do not establish disproportionate behavioral distrust.", "boundary": "conspicuous errors and controlled feedback, not employee review or actual commercial authority. Reported trust and demonstrated reliance differ. C2/C6/C8.", "clusters": [ "C2", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R40", "title": "Liu, Christian, Dumbalska, Bakker and Dubey — AI Assistance Reduces Persistence and Hurts Independent Performance — 6 April, revised 5 August 2026.", "date": "6 April, revised 5 August 2026", "url": "https://arxiv.org/html/2604.04721v4", "design": "Full v4.", "finding": "Full v4. Three short experiments, 1,222 recruited and 1,060 after exclusions; assistance removed before arithmetic/reading tests. Unaided performance declines; the improved arithmetic replication's persistence difference is nonsignificant.", "boundary": "immediate outcomes without performance-contingent payment, initial differential exclusions, self-selected hint/answer use. No durable professional deskilling coefficient. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R41", "title": "Wei Cai, Andrea Prat and Jiehang Yu — Measuring Organizational Capital — journal publication 8 June 2026.", "date": "journal publication 8 June 2026", "url": "https://onlinelibrary.wiley.com/doi/abs/10.1111/1475-679x.70073", "design": "Journal abstract, author explanation, chronology.", "finding": "Journal abstract, author explanation, chronology. More than a million employee reviews and text-based organizational measures associated with firm outcomes. **Access:** abstract and primary explanation, not full current methods. **Freshness:** originated 2022, revised 2025; new publication of older work. Proxies do not identify causal alignment or AI-created capital. C6/C7.", "boundary": "Abstract and primary author explanation inspected, not full current methods. Originated 2022 and revised 2025; June 2026 is journal publication of older work. Employee-review proxies and firm associations do not identify causal alignment or AI-created organizational capital.", "clusters": [ "C6", "C7" ], "access_depth": "abstract_or_primary_explanation", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R42", "title": "McKinley, Brau, Aloysius and Hofer — Is AI an Algorithm by Any Other Name — 13 April 2026.", "date": "13 April 2026", "url": "https://onlinelibrary.wiley.com/doi/full/10.1002/joom.70045", "design": "Full journal paper.", "finding": "Full journal paper. Two lab forecasting studies plus 2019–20 staggered retail deployment, 575,354 observations and 11,652 SKU-store clusters.", "boundary": "identical algorithm labels in labs differ from actual algorithm changes in the field. Context and performance affect reliance/adjustment; this is predictive forecasting, not conversational-agent reviewer accuracy or a fitted fatigue curve. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R43", "title": "Dell'Acqua and collaborators — The Cybernetic Teammate — June 2026 journal publication.", "date": "June 2026 journal publication", "url": "https://cep.lse.ac.uk/_NEW/PUBLICATIONS/abstract.asp?index=12590", "design": "Primary publication record.", "finding": "Primary publication record. Abstract/record inspected. Randomized work with 791 P&G professionals; AI changes individual/team idea work and cross-functional expertise. **Freshness:** working paper originated March 2025; journal publication is recent, underlying study is older.", "boundary": "product-innovation task and evaluated ideas, not company profit, sales queues or universal human–AI synergy. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "abstract_or_primary_explanation", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R44", "title": "Acemoglu, Tianyi Lin, Ozdaglar and Siderius — How AI Aggregation Affects Knowledge — arXiv 6 April 2026, document dated 24 August.", "date": "arXiv 6 April 2026, document dated 24 August", "url": "https://arxiv.org/html/2604.04906v1", "design": "Full theory.", "finding": "Full theory. DeGroot network with independent initial signals, no later external information, endogenous aggregator feedback and specified connectivity. Fast feedback can undermine robust learning; local topic aggregation has advantages under these assumptions. **Freshness:** earlier March circulation exists.", "boundary": "no empirical universal modular-architecture result. Supports testing source independence and fresh evidence. C0/C6/C7.", "clusters": [ "C0", "C6", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R45", "title": "Alex Chan — AI and the Information Supply of the Web — June 2026 working-paper discussion.", "date": "June 2026 working-paper discussion", "url": "https://aiinstitute.hbs.edu/the-internet-is-eating-itself-and-ai-is-holding-the-fork/", "design": "Primary HBS explanation.", "finding": "Primary HBS explanation. Author/institute explanation inspected; full paper not independently retrieved. Mechanism links traffic diversion, source financing and repeated information production.", "boundary": "theory/explanation, not observed inevitable web collapse. More generated pages or citations do not demonstrate fresh independent information. C3/C7.", "clusters": [ "C3", "C7" ], "access_depth": "abstract_or_primary_explanation", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R46", "title": "Dorner, Tucker, Zhang and Heuber — Green Consumer Information in a Digital Product Passport — 19 August 2026.", "date": "19 August 2026", "url": "https://link.springer.com/article/10.1007/s10640-026-01107-z", "design": "Full journal paper.", "finding": "Full journal paper. Preregistered incentivized toothbrush study, 131 participants, conducted 31 August–7 September 2023. Information-plus-education increases WTP by approximately NZ$0.69; information alone is null in parametric difference-in-differences.", "boundary": "small university pool, low-cost product, favorable information, selected engagement and no positive environmental spillover. Uptake, comprehension and consequential choice differ. C1/C2.", "clusters": [ "C1", "C2" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R47", "title": "Pattyn, Dillen and Goetz — Validate Before You Build — 18 August 2026.", "date": "18 August 2026", "url": "https://www.mdpi.com/2073-431X/15/8/535", "design": "Publisher article.", "finding": "Publisher article. Partial indexed full text inspected; direct publisher access limited. Twenty-nine founder interviews, 16 software startups; exploratory commitment levels and composite performance.", "boundary": "retrospective selected cases, endogenous validation, unverified interview dates, coding/composite dependence. Customer payment, investor money, sustained use and viability are different evidence. No causal PMF threshold. C1/C7.", "clusters": [ "C1", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R48", "title": "Chan, Lo, Jung and Wu — Emerging Together How Brand Collaborations Give Rise to Proto Categories — 17 July 2026.", "date": "17 July 2026", "url": "https://journals.aom.org/toc/amproc/2026/1?pageSize=20&pageStart=24", "design": "Primary proceedings listing.", "finding": "Primary proceedings listing. Abstract-level category-formation theory linking identity distance and collaboration.", "boundary": "no empirical sample or calibrated effect inspected. Treat proposed inverted-U as a hypothesis. Category recognition requires repeated audience behavior, not only a seller's positioning statement. C7.", "clusters": [ "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R49", "title": "Baumann, Cullen, Faia, Ferrando, Perez-Truglia and Rariga — Innovation Without Borders — June 2026.", "date": "June 2026", "url": "https://zoecullen.com/assets/docs/w35314.pdf", "design": "Full author-hosted paper.", "finding": "Full author-hosted paper. Information randomization in Q4 2025 ECB SAFE survey; 3,316 complete firms across 12 countries after substantial incomplete-response exclusions. Treatment raises expected AI investment share by 1.8 percentage points; domestic peer beliefs matter more than foreign ones.", "boundary": "immediate investment expectations, not realized spend, adoption, productivity or cash; complete-case selection. C7.", "clusters": [ "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R50", "title": "Bottesch, Schwenke, Zimmermann, Förster and Klier — Faster Higher Stronger — 28 July 2026.", "date": "28 July 2026", "url": "https://arxiv.org/abs/2607.25922", "design": "Paper and PDF.", "finding": "Paper and PDF. Full PDF inspected. Randomized document-RAG versus manual access in one industrial firm, November 2024–March 2025; 128 complete participants, 78/50 arms, expert output evaluation. Faster acquisition, summarization and creation, but less accurate acquisition.", "boundary": "short fixed-sequence tasks and exclusions; no payroll/profit endpoint, no significant creation-novelty gain. C5/C6/C8.", "clusters": [ "C5", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R51", "title": "Babina, He and Jiang — Canaries in the Gold Mine — August 2026, SSRN posted 25 August.", "date": "August 2026, SSRN posted 25 August", "url": "https://www.nber.org/papers/w35684", "design": "NBER record, SSRN, measurement documentation.", "finding": "NBER record, SSRN, measurement documentation. Abstract/author/documentation appraisal only; long full manuscript inaccessible. AI-skilled employment and firm panels through 2024 are associated with recent productivity gains and organization-building roles.", "boundary": "identification not deeply audited; updated firm panel unreleased. Mechanism candidate for accumulating workflow capital, not a transferable productivity coefficient. C5/C7.", "clusters": [ "C5", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R52", "title": "Niklas Heusch — Synthetic Benchmark Dataset with Endogenous Marketing Spend — 21 August 2026.", "date": "21 August 2026", "url": "https://arxiv.org/html/2608.21130v1", "design": "Full preprint.", "finding": "Full preprint. A 156-week, three-channel fictional retail generator links spend to quarterly performance, planned promotions, bursts and algorithmic chasing; provides known effects and noisy practitioner observations.", "boundary": "no real company data, additive channels, fixed parameters, no rival/supply responses; bidding uses pre-media sales and geo groups share a national series. Useful adversarial estimator tests, not calibrated GTM response or proof of real-data identification. C3/C5/C6.", "clusters": [ "C3", "C5", "C6" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R53", "title": "Masahiro Kato, Daiki Honma and Taka Kato — Generative Marketing Mix Modeling — 10 September 2026.", "date": "10 September 2026", "url": "https://arxiv.org/html/2609.11915v1", "design": "Full preprint.", "finding": "Full preprint. Repeated answer collection on 56 English/Japanese questions supplies occurrence estimates; business responses and source changes are simulated. Requires query counts, platform shares, notice probabilities, rank/support, exchangeability and a specified interference process.", "boundary": "no randomized commercial GEO uplift; main-effect priors are nonnegative. Measurement and simulated response precision do not establish the empirical effect of citations or source edits. C2/C3/C5.", "clusters": [ "C2", "C3", "C5" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R54", "title": "Uberti-Bona Marin and colleagues — Auditing AI Generated Product Recommendations — 16 September 2026.", "date": "16 September 2026", "url": "https://arxiv.org/html/2609.18729v1", "design": "Full preprint.", "finding": "Full preprint. ConsumerQ contains 2,528 real advice queries; the main audit uses 117 physical-product queries repeated three times, producing 1,536 responses across interface/API conditions.", "boundary": "nonrepresentative source users, one collection location/time, only recovered AI Overviews, three repetitions, configuration differences. Business products and subscriptions are excluded from the core audit; displayed sources do not establish causal influence. C2/C3/C8.", "clusters": [ "C2", "C3", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R55", "title": "Farrell, Korganbekova and Misra — Robust A/B Decisions — 7 September 2026; manuscript 5 September.", "date": "7 September 2026; manuscript 5 September", "url": "https://arxiv.org/html/2609.07633v1", "design": "Full preprint.", "finding": "Full preprint. Primary methods/results inspected. Uses 552 randomized advertising experiments from one anonymous US platform, with undisclosed study dates, to retrospectively compare deployment rules under random and geographic splits.", "boundary": "held-out deployment evaluation, not randomized rollout of decision rules. Empirical success performs best under random splits; ambiguity preferences have outcome-specific units. The oracle itself uses noisy held-out means. No transferable 25% revenue gain follows. C5/C6/C7.", "clusters": [ "C5", "C6", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R56", "title": "Zihao Chen — When Do Surrogate Metrics Work — 24 September 2026.", "date": "24 September 2026", "url": "https://arxiv.org/html/2609.30528v1", "design": "Full preprint.", "finding": "Full preprint. Primary setup, methods and results inspected; proofs not independently audited. Eleven simulations and older Hillstrom, Criteo and LaLonde experimental reanalyses expose coverage failures that a surrogate disagreement diagnostic often misses.", "boundary": "prediction-powered alternatives require random labeling, enough observed events and justified enrollment/cohort stability; sparse outcomes and drift remain problematic. A non-rejected diagnostic supplies no certificate. Not a new SaaS LTV estimate or permission for unrestricted repeated testing. C4/C5/C6.", "clusters": [ "C4", "C5", "C6" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R57", "title": "Tobias Konitzer — Media Measurement and the Assisted Own Goal — 10 July 2026; manuscript 21 June.", "date": "10 July 2026; manuscript 21 June", "url": "https://arxiv.org/html/2607.09608v1", "design": "Full preprint.", "finding": "Full preprint. Primary analytical model, simulations and identification conditions inspected. GrowthLoop-affiliated author analyzes demand generated on one property but booked elsewhere and proposes audience randomization.", "boundary": "commercially interested theoretical/simulation evidence, no field bias magnitude. Identification expressly needs channel-complete outcomes; randomization does not recover missing transactions. Outcome coverage, linkage and interference remain requirements. C3/C5/C6.", "clusters": [ "C3", "C5", "C6" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R58", "title": "Wang, Zhu, Feng, Lu and Jia — Agentic AI and Human-in-the-Loop Interventions — 14 May, revised 1 June 2026.", "date": "14 May, revised 1 June 2026", "url": "https://arxiv.org/pdf/2605.14830v2", "design": "Full v2 PDF.", "finding": "Full v2 PDF. Primary methods/results and matching inspected. Taobao worker-randomized deployment with 647 workers and 680,676 chats, 15–31 August 2024 plus 14-day baseline; 39,432 chats AI-eligible. Faster service accompanies lower eligible-chat ratings; only 17% of chats receive ratings.", "boundary": "short customer-service study, no acquisition/profit endpoint. Escalation type and timing are selected, with matched comparisons rather than randomized mechanisms. No universal causal handoff penalty. C4/C5/C6/C8.", "clusters": [ "C4", "C5", "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R59", "title": "Han, Yan and Zhang — More Rejective Not More Discriminative — 25 August 2026.", "date": "25 August 2026", "url": "https://arxiv.org/html/2608.23941v1", "design": "Full preprint.", "finding": "Full preprint. Primary methods/results and limitations inspected. Preregistered LLM-monitor instrument varies five review lengths and six verifiers in retail/AppWorld with injected errors and clean controls; retail's 200 bad items share 31 clean anchors. Longer units generally raise both valid and invalid rejection.", "boundary": "nonadaptive errors, short windows, changed intermediate observations and token-only costs; no human fatigue experiment or universal review cadence. Catch rate can be identified with labels while remaining an insufficient objective. C6/C8.", "clusters": [ "C6", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R60", "title": "Liang Chen — Simulated Customers Never Walk Away — 16 June 2026.", "date": "16 June 2026", "url": "https://arxiv.org/html/2606.20708v1", "design": "Full preprint.", "finding": "Full preprint. Primary methods/results inspected. Chinese parent-mediated matchmaking sales logs contain 2,790 conversations and 793 verified payments; paired evaluation uses 374 conversations and 1,109 probes. Simulated nonbuyers overengage, and disengagement instructions trade one group's error against another.", "boundary": "observational next-turn fidelity given real history, not free-running conversion or causal sales-tactic effectiveness. LLM-assigned states lack independent human validation; outcome-balanced sampling is not a natural conversion base. Study dates unspecified. C1/C2/C8.", "clusters": [ "C1", "C2", "C8" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R61", "title": "Kreimer, Stoppacher, Hauser, Foscht and Eisingerich — Pathways to Premium — 21 September 2026.", "date": "21 September 2026", "url": "https://link.springer.com/article/10.1007/s11747-026-01205-w", "design": "Full journal paper.", "finding": "Full journal paper. Four observational field studies at one regional European newspaper, samples 522, 566, 358 and 2,434; calendar dates unspecified. Earlier studies measure intentions/offer acceptance; the fourth observes retention after a six-week €0.99 trial automatically becomes €12.99 monthly.", "boundary": "no randomized habit treatment; Study 4 lacks demographic/usage controls, engagement imperfectly measures habit, and trial-hopping explanation is untested. Phase-specific hypotheses, not a transferable optimal habit threshold. C4/C5.", "clusters": [ "C4", "C5" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R62", "title": "Kyungran Lee, ShinWoo Lee and Gil Sadka — ASC 606 Revenue Uncertainty and Cost of Debt — journal 24 April 2026.", "date": "journal 24 April 2026", "url": "https://link.springer.com/article/10.1007/s11142-026-09936-5", "design": "Full journal paper, earlier working-paper record.", "finding": "Full journal paper, earlier working-paper record. New publication of work first posted 18 September 2024 and revised 23 January 2026. Difference-in-differences examines materially affected public firms around accounting adoption, using 2016–19 and 2014–21 windows.", "boundary": "exposure, coincident industry shocks and selected analyst/loan coverage constrain interpretation. Accounting-information uncertainty is not GTM demand uncertainty; financing results are not a startup cash-timing coefficient. C5.", "clusters": [ "C5" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R63", "title": "Chung and Chuwonganant — Effects of the New Revenue Recognition Standard — journal 23 August 2026.", "date": "journal 23 August 2026", "url": "https://onlinelibrary.wiley.com/doi/10.1111/jbfa.70087", "design": "Publisher abstract, working-paper history.", "finding": "Publisher abstract, working-paper history. Abstract/metadata inspected; full text unavailable. First circulated 27 April 2020, revised 9 July 2026. Reports improved analyst forecast accuracy/information using public-firm data.", "boundary": "match samples, exposure definitions, horizons and estimands before declaring a contradiction with Lee and colleagues. This is an abstract-only counterpoint against a uniform accounting-information claim, not directly observed cash or GTM causality. C5/C7.", "clusters": [ "C5", "C7" ], "access_depth": "abstract_or_primary_explanation", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R64", "title": "Stephanie T. Wang, Jeffrey Gleason, Yakov Bart, Christo Wilson and Danaé Metaxa — AI in Search Reduces Publisher Referrals — 18 August 2026.", "date": "18 August 2026", "url": "https://arxiv.org/html/2608.18352v1", "design": "Full preprint.", "finding": "Full preprint. Preregistered browser-extension randomized field experiment; March 17–19 recruitment, three-day baseline and seven-day treatment. Of 1,444 enrolled, 1,100 search during treatment and 956 complete the survey. Forced AI Mode lowers outbound click-through by 18.8 percentage points, CI −22.2 to −15.3.", "boundary": "selected short-horizon panel and forced adoption from 0.6% baseline; no-AI estimates are LATE under approximately 51% compliance/exclusion. No B2B shortlist or revenue effect. Linked OSF registration could not be independently retrieved, so its registration date is unverified. C2/C3/C7.", "clusters": [ "C2", "C3", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" }, { "id": "R65", "title": "Peibo Zhang, Ruomeng Cui and Dennis J. Zhang — The Impact of AI Search on the Online Content Ecosystem — 14 May, revised 20 June 2026.", "date": "14 May, revised 20 June 2026", "url": "https://arxiv.org/html/2605.16428v3", "design": "Full v3.", "finding": "Full v3. Difference-in-differences compares AI-Overview-eligible SFW with NSFW Reddit communities: 105,012 subreddits over 578 days, January 2024–July 2025; December 2025 extension examines AI Mode.", "boundary": "engagement is not referral traffic or commercial conversion. Differential shocks, partial US exposure in the designated preperiod and an author-count pretrend constrain identification; content heterogeneity uses model classifications. Useful destination-complementarity evidence, not universally established parallel trends or a proven conservative lower bound. C3/C7.", "clusters": [ "C3", "C7" ], "access_depth": "methods_or_primary_report_inspected", "local_parameter_status": "not_fitted_to_this_site_or_an_operator_cohort" } ], "proposed_studies": [ { "id": "P01", "title": "Agent-assisted B2B execution", "clusters": [ "C4", "C5", "C6", "C8" ], "design": "Randomize eligible accounts/buying groups to current policy, qualified assistant policy, and where feasible no-contact control", "outcome": "Incremental contribution per assigned account over a declared horizon", "mechanisms_and_guardrails": "CPQO, acquisition, human hours, retries, complaints, authority breaches, implementation/retention", "identification_requirements": "Member contamination, hidden selection, immature outcomes and changing eligibility", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P02", "title": "Human and agent handoff", "clusters": [ "C4", "C6", "C8" ], "design": "Compare complete triage/follow-up policies under the same capacity envelope", "outcome": "Accepted deals or contribution through final customer quality", "mechanisms_and_guardrails": "False-high/low intent, queue delay, seller exceptions, lost follow-ups, downstream adverse selection", "identification_requirements": "Conditioning only on contacted/high-intent accounts", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P03", "title": "Buyer AI mediation", "clusters": [ "C1", "C2", "C8" ], "design": "Observe journeys; randomize permitted representation/evidence improvements at account or interface level", "outcome": "Correct representation and qualified consideration; later contribution separately", "mechanisms_and_guardrails": "Discovery, hard-constraint fidelity, verification, approval, human veto and implementation", "identification_requirements": "Platform changes, unobserved exposure, attribution of existing preference", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P04", "title": "Synthetic buyer validation", "clusters": [ "C1", "C2", "C8" ], "design": "Held-out real respondents and choices; compare no grounding, demographics, rich histories and wrong-person grounding", "outcome": "Predeclared distribution/ranking/level error and decision regret", "mechanisms_and_guardrails": "Subgroup errors, novelty, model/version instability and cost", "identification_requirements": "Treating generated replicas as independent people; leakage from known questions", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P05", "title": "Agent qualification", "clusters": [ "C0", "C5", "C6", "C8" ], "design": "Develop policies then freeze and test on representative held-out workflows", "outcome": "Independently accepted terminal/process success with uncertainty", "mechanisms_and_guardrails": "Human baseline, review/repair/defer, severe cases, cost per success, queue capacity", "identification_requirements": "Reusing qualification cases; judge leakage; drift after qualification", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P06", "title": "Channel response and carryover", "clusters": [ "C3", "C5", "C7" ], "design": "Randomized supported spend levels or well-designed geo/account interventions with cooldown", "outcome": "Incremental discounted contribution in tested support", "mechanisms_and_guardrails": "Time path, spillover, organic displacement, pull-forward, rival response", "identification_requirements": "Demand-following budget pauses, weak controls and finite-geo imbalance", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P07", "title": "Pricing and packaging", "clusters": [ "C1", "C2", "C5", "C7" ], "design": "Randomize eligible segments/accounts or appropriate clusters to feasible packages/metrics", "outcome": "Contribution and cash over acquisition plus renewal horizon", "mechanisms_and_guardrails": "Adoption, usage, reference expectations, bill shock, exceptions and support tails", "identification_requirements": "Current-price comparisons ignore history; shared-capacity interference", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P08", "title": "Retention and value", "clusters": [ "C4", "C5" ], "design": "Randomize retention plays among eligible accounts with untreated comparison", "outcome": "Incremental retained contribution rather than save count", "mechanisms_and_guardrails": "Usage, voluntary renewal, billing failure, discount cost, satisfaction and implementation", "identification_requirements": "Targeting risk rather than responsiveness; survival conditioning", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P09", "title": "Review and organization", "clusters": [ "C6", "C8" ], "design": "Vary cadence/scope under measured workload; preserve hard authority rules", "outcome": "Valid accepted throughput per available review hour", "mechanisms_and_guardrails": "Backlog, catch/false-positive rates, rework, quality and reviewer learning", "identification_requirements": "Easier-case selection, workload redistribution, harmful deskilling conclusions from voluntary use", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P10", "title": "Market and category feedback", "clusters": [ "C2", "C3", "C7", "C8" ], "design": "Cluster/switchback tests only where spillover/lag justify them; monitor rivals and platforms", "outcome": "Firm contribution plus separately defined ecosystem outcomes", "mechanisms_and_guardrails": "Congestion, relevance, platform delivery, trust, demand expansion and source freshness", "identification_requirements": "Individual tests assume no interference; changepoints misidentified as external shocks", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P11", "title": "Funnel resource response", "clusters": [ "C3", "C4", "C5" ], "design": "Randomize feasible resource changes while freezing stage definitions and eligibility", "outcome": "Incremental contribution or qualified throughput per incremental resource", "mechanisms_and_guardrails": "All downstream rates, quality, service load, queue time and filtering", "identification_requirements": "Administrative threshold changes mistaken for productivity; carryover and shared-capacity interference", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P12", "title": "Proxy and outcome coverage", "clusters": [ "C3", "C4", "C5", "C6" ], "design": "Backtest proxies on mature holdouts and audit booking/linkage across direct, partner and marketplace routes", "outcome": "Held-out decision agreement, effect direction, interval coverage and missing-outcome sensitivity", "mechanisms_and_guardrails": "Missing/censored/selected labels, subgroup error, changed enrollment and collection paths", "identification_requirements": "Validation lacks outcome support; correlation substituted for causal-effect agreement", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." }, { "id": "P13", "title": "Deployment-rule validation", "clusters": [ "C5", "C6", "C7" ], "design": "Reevaluate historical experiments with current, empirical-success and explicit economic/downside rules; later prospective policy test where feasible", "outcome": "Declared deployment payoff/regret and downside exposure", "mechanisms_and_guardrails": "Launch costs, reversible decisions, drift, outcome units and uncertainty", "identification_requirements": "Noisy ex-post oracle, selective experiment archives and transport disguised as randomized rollout", "status": "proposed_unrun", "sample_size": "Determine from local variance, minimum useful effect, clustering, attrition and outcome horizon; not yet calculated." } ], "cluster_roadmap": [ { "id": "C0", "name": "Orientation", "priority": "first", "improvements": [ "An epistemic tag alone cannot govern a decision. “Measurable” means a measurement could exist; it does not mean the current value was measured, that its causal effect was identified, or that a proposed deployment passed qualification. Represent these as separate states. Retain the model's refusal behavior, but tie it to observable conditions rather than the assumed universal 25–40% noise floor.", "For each material claim or coefficient, the proposed evidence record contains the population, business kind, outcome and denominator, intervention/control, assignment unit, study and release dates, access depth, uncertainty, scope, causal assumptions, configuration/cost snapshot, falsification criterion, and conditions requiring re-estimation. A policy preference or risk limit is normative and should be labeled that way.", "Start with mathematical consistency before fitting new coefficients. A beautifully calibrated implementation of an inconsistent target is still unusable. The correction register below identifies exact model locations." ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P05" ], "evidence_ids": [ "R25", "R44" ] }, { "id": "C1", "name": "Market & Customer", "priority": "high", "improvements": [ "The fit and business-kind layers already prevent some universal SaaS assumptions. Make their distinctions operational. An account can have good end-user fit but weak budget ownership, poor security acceptability, costly implementation, or negative supplier contribution. A PMF score should expose these observations and unknowns instead of absorbing them into one convenient multiplier.", "Learn segments on discovery/training data, then freeze their definitions before confirmatory holdout evaluation. Retain segment lineage when policies change; otherwise a better-looking cohort can result from changing the eligible population. Assess fit against real alternatives, including internal workflows and no purchase. Use nonbuyers and failed implementations alongside customer praise.", "Synthetic research can expand candidate hypotheses and interview probes. Before using it for segment selection, demand sizing, willingness-to-pay or committee decisions, validate that particular prediction against held-out people. Wrong-person and demographic-only baselines reveal whether rich personas add useful information. Evaluate distributions and subgroup errors, not just convincing explanations or top-choice agreement.", "Include real nonbuyers, refusal, disengagement, delay and substitution. Validation on outcome-balanced sales logs must not be read as a natural conversion-rate estimate. A plausible next response, a complete simulated journey and an intervention effect require separate validation." ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P03", "P04", "P07" ], "evidence_ids": [ "R09", "R11", "R12", "R13", "R15", "R18", "R46", "R47", "R60" ] }, { "id": "C2", "name": "Demand & Buyer Psychology", "priority": "high", "improvements": [ "Keep the existing buyer belief and trust graph. Replace the single `B_agent` share as the main observation with a matrix indexed by role, task, and phase. Record platform/interface/version, observed use, representation fidelity, verification path, authority, timestamp, and unknown status. Separate research, recommendation, drafting, guarded execution and permitted purchasing.", "Track initial and current shortlists, evidence requirements and unresolved constraints. A buyer can discover a vendor through AI, verify through search or peers, obtain internal approval through humans, and still fail implementation. A citation, a recommendation, a contract, and buyer welfare are different outcomes.", "Sample the interfaces buyers actually use and repeat queries. A recent consumer audit finds that API and interface outputs expose different recommendations and source layers; it explicitly excludes business products from its main recommendation sample. Apply the measurement lesson, without importing its consumer magnitudes. ConsumerQ audit", "Operationalize the model's political/procurement fit through relevant roles: champion/user, economic buyer, finance, technical/security/legal review, and veto authority. Do not assume every purchase has every role. Buying-group size is a poor substitute for who can block approval and which proof remains missing.", "The 95–5 in-market heuristic and the probability that winners were already shortlisted are descriptive quantities with different denominators. Neither estimates that 95% of value creation occurs before contact. Seller leverage should be estimated by phase and intervention, with room for useful evaluation, approval and implementation work." ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P03", "P04", "P07", "P10" ], "evidence_ids": [ "R05", "R06", "R07", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R18", "R19", "R20", "R21", "R37", "R39", "R46", "R53", "R54", "R60", "R64" ] }, { "id": "C3", "name": "Motions & Channels", "priority": "high", "improvements": [ "Choose an acquisition or retention action by expected incremental contribution, feasible change and cost. Conversion propensity, attributed revenue, and treatment effect answer different questions. Allow withholding contact or spend when the expected net effect is negative, and include shared sales/service capacity constraints.", "Treat saturation, creative fatigue, platform delivery, rival imitation and growing category demand as competing explanations. A fixed reciprocal CAC curve bakes in a particular asymptote; observed autocorrelated spend may not distinguish response curvature from changing effectiveness. Preserve experiment support and reject extrapolation beyond it without a labeled scenario.", "For attribution and brand, keep experimental time paths, not only one ROAS number. Self-reported attribution can inform awareness and mechanisms, but cannot identify a causal lag kernel alone. Test complementarity, organic displacement, delayed conversion and pull-forward. A pause may follow weak demand; screen-and-forecast methods do not remove that confounding.", "Specify where outcomes are booked and observed. Demand generated through one source may convert through a marketplace, partner or direct channel. Randomizing exposure does not recover transactions absent from the outcome system. Reconcile observation coverage and linkage, and audit whether matching or reporting failure depends on assignment. An analytical commercially affiliated paper makes this boundary explicit without estimating its field magnitude. Konitzer", "AI visibility also needs a denominator: relevant query volume, ICP platform usage, repeated answer occurrence and user notice. GMMM offers a proposed measurement/causal framework with explicit support assumptions and simulated commercial responses. Its measured answer occurrences do not constitute measured incremental sales. Kato and colleagues", "Motion choice is constrained by economics, buyer requirements, capacity and product properties. A single universal ACV threshold or the smallest observed funnel rate is an inadequate decision rule. Hybrid motions and multiple simultaneous constraints already fit the generalization layer; express them explicitly.", "For partner motions, distinguish sell-in from customer sell-out and reconcile reseller/OEM/platform fees, margin shares, duplicate booking and channel conflict. A partner-generated contract may have different cash timing, customer visibility and service obligations. Those affect both outcome completeness and the contribution assigned to a motion." ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P06", "P10", "P11", "P12" ], "evidence_ids": [ "R01", "R02", "R03", "R04", "R06", "R07", "R09", "R18", "R21", "R45", "R52", "R53", "R54", "R57", "R64", "R65" ] }, { "id": "C4", "name": "Funnel & Lifecycle", "priority": "high", "improvements": [ "Keep funnel rates for diagnosis, but use a state graph for decisions: discovery, evaluation, approval, contract, implementation, use/value, renewal, and return to evaluation. Include abandonment, rework, review queues, cancellation and involuntary billing failure. Define transitions and time stamps so a faster draft does not masquerade as a faster sale.", "Preserve stage definitions and cohort composition across comparisons. Tightening an MQL qualification rule can lower the preceding conversion rate and raise the following one without changing sales. A lower rate may be a useful filter. Report complete eligible-cohort outcomes alongside stage diagnostics, and distinguish a definition change from a behavioral change.", "Handoff quality concerns both information and allocation. False-high intent consumes scarce seller time; false-low intent can suppress useful follow-up. Evaluate the complete AI–human policy through the final outcome, rather than rewarding cleaner summaries, application volume or opportunity creation alone.", "Distinguish logo survival, payment survival, product use, realized value, satisfaction, contract lock-in and voluntary renewal. Improving observed retention can reflect selective early churn rather than improved individual experience. Churn propensity differs from responsiveness to an intervention; some saves can destroy contribution.", "Use complete cohorts and explicit censoring. Report mature observed outcomes separately from forecasts. Do not declare a universal six-to-twelve-month point after which all LTVs become reliable; contracts, seasonality, renewal cycles and sample sizes differ.", "Qualify short-horizon proxies against mature outcomes, including directional error, subgroup fragility and interval coverage. A failed-to-reject surrogate diagnostic is not a certificate. Prediction-powered estimation is a candidate when its labeling and stability conditions hold, rather than a universal replacement for outcome measurement. Enrollment drift, few observed events and selective labeling require additional checks. Chen on surrogate metrics", "Free use, a cheap trial, paid subscription and active realized value are separate states. Recent newspaper data associate habits with conversion and retention differently across phases, but do not randomize habits and include automatic renewal. Use that pattern to propose experiments rather than importing an optimal engagement threshold. Kreimer and colleagues" ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P01", "P02", "P08", "P11", "P12" ], "evidence_ids": [ "R10", "R16", "R17", "R19", "R22", "R35", "R37", "R56", "R58", "R61" ] }, { "id": "C5", "name": "Economics & Metrics", "priority": "first", "improvements": [ "Separate customer stock from recurring revenue, usage, backlog, installed assets and recognized/cash value. Expansion raises revenue without adding logos. Hybrid kinds need distinct timing and contribution ledgers. The model already contains this generalization; the original spine should obey it.", "Forecast survival-weighted contribution with time-varying price, usage, delivery cost and discounting. Include contraction, reactivation, returns, credits, bad debt, human exceptions and inference cost. Revenue expansion can worsen contribution. Report payback and contribution distributions where cost tails matter, rather than only mean LTV:CAC.", "For AI economics, count all eligible attempts, retries, review, rework, tools/data, allocated maintenance and attributable losses. Separate sunk overhead from avoidable incremental costs. Hours saved produce cash savings when expenditure falls. Productive redeployment may generate incremental contribution, which must be measured separately. Outcome pricing additionally requires an enforceable acceptance definition and allocation of failure risk.", "Price, packaging and metering are actions that change acquisition, usage, expectations and renewals. Evaluate immediate conversion together with contribution, bill predictability, complaints and later renewal. An observed willingness-to-pay model and a causal pricing policy are separate objects.", "Separate estimation from deployment. A confidence interval describes evidence about the tested effect; choosing a rollout also involves implementation cost, reversibility, downside, capacity and transfer to future conditions. Backtest candidate economic rules against simple empirical-success and current-policy baselines. A recent experiment-archive study supports this comparison, not automatic adoption of its ambiguity parameter. Report both the decision and inferential uncertainty. Farrell, Korganbekova and Misra" ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P01", "P05", "P06", "P07", "P08", "P11", "P12", "P13" ], "evidence_ids": [ "R01", "R02", "R03", "R04", "R05", "R06", "R08", "R22", "R25", "R50", "R51", "R52", "R53", "R55", "R56", "R57", "R58", "R61", "R62", "R63" ] }, { "id": "C6", "name": "GTM Operations (GTM OS)", "priority": "first", "improvements": [ "The next operational layer should join observations without erasing provenance. Maintain account, contact, contract and usage identities; pre-action feature snapshots; assignment and policy versions; action probabilities where applicable; permission and consent state; source dates; raw receipts; review decisions; downstream outcomes; and missingness reasons.", "Attach record age, last verification, linkage confidence and correction history. Measure error by task, record quality and case mix; a base error rate attached only to the agent cannot explain stale contact data or an upstream source outage. Verification freshness and permission freshness can expire separately.", "Human approval is a label about a particular criterion, not universal ground truth. Separate style, factuality, fit, compliance, authority and actual business outcomes. All accounts lack their unobserved outcome under the action not received; rejected drafts and untouched accounts may additionally lack factual downstream labels. Counterfactual effects require identification, and selectively missing factual labels require a missingness audit. Down-weighting approvals solves neither automatically; use experiments, audits, sensitivity analysis or bounds.", "Track assisted performance and unaided fallback competence separately. Test persistence and anomaly investigation when assistance is unavailable; durable professional skill effects remain unresolved rather than presumed beneficial or harmful.", "Make governance separable: detection, action enforcement, correction, evidence preservation, containment and recovery. Hard permissions should be enforceable predicates. A governance score can summarize operations but should not authorize an irreversible action by itself. Compare raw execution evidence with generated reports and verify critical claims independently.", "Measure capacity through arrival rates, service minutes, case mix, skills, backlog age and rework. Review fraction is not review load. Organization-wide learning also requires implementation: receiving feedback is not equivalent to changing the policy or improving outcomes.", "Evaluate review against matched clean controls as well as defects. Longer review units can reject more good and bad plans at once, so a higher catch rate need not improve discrimination or accepted throughput. Review-window findings from LLM monitors should not be relabeled human fatigue evidence. Han, Yan and Zhang" ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P01", "P02", "P05", "P09", "P12", "P13" ], "evidence_ids": [ "R01", "R10", "R11", "R13", "R14", "R17", "R20", "R22", "R23", "R24", "R26", "R27", "R28", "R29", "R30", "R32", "R34", "R35", "R36", "R38", "R39", "R40", "R41", "R42", "R43", "R44", "R50", "R52", "R55", "R56", "R57", "R58", "R59" ] }, { "id": "C7", "name": "Strategy, Regimes & Scaling", "priority": "high", "improvements": [ "The model already distinguishes drift, regime change and topology change. Link those distinctions to a transition process influenced by the firm, rivals, platforms and external conditions. A changepoint alarm identifies a statistical change, not its cause. Test whether policy changes, delivery shifts, selection, competitor response or external shocks explain it.", "Separate private contribution from category outcomes. Agent adoption can create congestion, improve relevance, expand demand, change information supply, or shift platform power. The direction of its net externality is empirical. Reply-rate decline alone does not identify pollution caused by agents.", "Use cash, capacity, quality, compliance and service promises as potentially simultaneous constraints. Treat strategic objective weights as declared preferences/scenarios, with units and sensitivity. Macro valuation heuristics are not accounting identities or causal policies. Strong strategic feedback may make global convergence difficult; theory does not establish that every GTM market is computationally hard.", "Separate external capital conditions and valuation multiples from firm-specific runway, funding access, covenants, investment commitments and management mandate. Track AI readiness and accumulated workflow knowledge separately again. Evidence about investment expectations or organization capital does not identify a universal interest-rate-to-operating-objective coefficient.", "Category formation needs independent evidence and information financing as well as promotion. AI-generated consensus can recycle shared sources. Preserve fresh observations and provenance, and test platform/rival responses before scaling a locally successful policy.", "Keep destination heterogeneity explicit. A short randomized AI-search experiment reduces publisher referrals, while a separate quasi-experiment associates AI Overviews with more Reddit engagement. These measure different populations, interfaces and outcomes, so neither establishes a universal direction for AI-mediated demand. Test substitution versus complementarity for the actual destination and business endpoint. Wang and colleagues, Zhang, Cui and Zhang" ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P06", "P07", "P10", "P13" ], "evidence_ids": [ "R04", "R07", "R08", "R28", "R31", "R32", "R33", "R34", "R41", "R44", "R45", "R47", "R48", "R49", "R51", "R55", "R63", "R64", "R65" ] }, { "id": "C8", "name": "Agentic & AI-era GTM", "priority": "first", "improvements": [ "Retain autonomy as a descriptive aggregate if helpful. Decisions should use a policy that can execute, review, repair, defer or stop according to task, state, capacity and authority. Reading, drafting, modifying CRM, sending, quoting, committing terms and changing strategy have different consequences.", "Define success as an independently accepted terminal outcome plus process/authority validity. A repairable error before external effect is different from an unauthorized action whose output looks correct. Qualify a frozen policy on held-out representative tasks, including measured human-path error, severe-event checks and available review capacity; monitor changing task distributions after release.", "Cost per accepted success must include failures in the numerator. Choose the policy objective explicitly: for example, minimize total cost of meeting specified accepted throughput, reliability and risk requirements. Report cost per accepted success as a complementary diagnostic; it is a different objective when policies accept different numbers of cases. No reviewed evidence justifies a universal GTM threshold above 98%, an inverted-parabola optimal review load, or a general promise that hybrid agents have already settled the economic question.", "The decisive field experiment compares complete policies on the same eligible population and horizon, through realized contribution, operational burden and downstream customer quality. Current papers supply mechanisms and test designs; a generic autonomous B2B agent CAC coefficient remains unqualified." ], "measurement_contract": "Declare eligible population/business kind, target and denominator, pre-action information, policy/comparator/assignment, outcome horizon and coverage, cost/authority/configuration, uncertainty, causal assumptions, falsifier and expiry. Current coefficients/qualification remain unmeasured unless explicitly linked to local evidence.", "proposed_study_ids": [ "P01", "P02", "P03", "P04", "P05", "P09", "P10" ], "evidence_ids": [ "R07", "R08", "R09", "R10", "R11", "R12", "R13", "R14", "R15", "R16", "R17", "R18", "R21", "R22", "R23", "R24", "R25", "R26", "R27", "R28", "R29", "R30", "R31", "R37", "R38", "R39", "R40", "R42", "R43", "R50", "R54", "R58", "R59", "R60" ] } ], "qualification_protocol": { "status": "specified_not_deployment_qualified", "decision_rule": "A policy is eligible for a scoped decision only when accounting/measurement, causal or predictive support, authority and complete-workflow quality/cost/capacity evidence match the intended target. Unknown material inputs produce an explicit refusal or labeled scenario.", "required_fields": [ "population, eligibility, business kind and stable segment/stage lineage", "outcome, denominator, horizon, payment/cost timing and maturity", "policy, comparator/no-action, assignment unit and pre-action feature snapshot", "study dates, release/revision dates, access depth, provenance and source independence", "model/tool/data/interface/policy/authority versions and cost snapshot", "uncertainty, support, missingness/censoring, interference and identification assumptions", "independent acceptance/process predicates, falsifier and requalification trigger" ], "hard_gates": [ "Current authorized task/action scope and applicable versioned policy requirements", "Defined severe process/quality/authority failure predicates", "Available service/review capacity and escalation/defer/stop path", "Independent critical-action receipts and terminal acceptance verification" ], "comparison_rules": [ "Compare complete eligible workloads and declared contribution endpoints, not selected successes.", "Separate CPQO, CAC, revenue, contribution, cash, hours released and realized savings.", "Preserve raw receipts, assigned controls, untreated accounts, missingness and failed attempts.", "Prediction accuracy and conditional rates do not identify intervention effects.", "Surrogate or synthetic-human validation is target-specific; real held-out outcomes and stability requirements still apply.", "Qualified terminal success, all-step success, correction success and process validity are distinct metrics.", "Randomized, observational, theory, benchmark and vendor evidence retain separate boundaries." ], "forecast_rules": [ "No universal 25-40% error floor, 6-12-month maturity or fixed review/governance coefficient.", "Use target/horizon-specific temporal/cohort holdouts, calibration and decision precision.", "Refuse unsupported population/configuration/spend extrapolation or immature/selected outcomes.", "Material model/tool/policy/data/eligibility/authority change triggers requalification." ] }, "research_update_v3_3": { "window_start": "2026-04-05", "window_end": "2026-10-05", "report_path": "docs/research/gtm-v3-2-research-2026-10-05.md", "summary": "Comprehensive evidence appraisal converted into corrected equations, qualified claims, nine curriculum-cluster plans and thirteen proposed studies. Breadth retained; unsupported universal causal, ROI and governance claims retired.", "evidence_status": "Evidence synthesis and specification implemented; local fitting, commercial experiments and end-to-end deployment qualification remain unrun.", "source_count": 77, "current_source_count": 65, "foundation_source_count": 12, "correction_count": 31, "cluster_namespace": "Canonical curriculum C0-C8; editorial public hubs C1-C7 remain separate. Legacy C6+ agentic alias maps to canonical C8.", "priority_clusters": [ "C0", "C5", "C6", "C8" ], "expert_schools": [ { "researchers_and_school": "Customer-base science: Peter Fader, Bruce Hardie, Eva Ascarza", "contribution": "Cohort heterogeneity, survival, value, retention interventions", "tension": "Low predicted churn and observed loyalty do not imply a retention treatment will help" }, { "researchers_and_school": "Causal marketing: Brett Gordon, Robert Moakler, Florian Zettelmeyer; LinkedIn's Wei and colleagues", "contribution": "Experimental effects, scalable effect prediction, constrained assignment", "tension": "High conversion prediction and attributed revenue can target buyers who would buy anyway" }, { "researchers_and_school": "Marketing research validity: Oded Netzer, Rajan Sambandam, Jochen Hartmann, Leonard Kinzinger; Eduard Kuric and colleagues", "contribution": "Grounding, synthetic response validity, distribution and individual fidelity", "tension": "Useful selected survey prediction coexists with poor novel choice simulation" }, { "researchers_and_school": "Pricing and beliefs: Jean-Pierre Dubé, Ilia Morozov, Anna Tuchman; Ningyin Xu; Oded Netzer and Yael Karlinsky-Shichor", "contribution": "Information, price histories, exceptions and hybrid decisions", "tension": "More information, higher conversion or more automation need not improve long-run profit" }, { "researchers_and_school": "Response identification: Ryan Dew, Anastasia Shchetkina and collaborators; Heusch", "contribution": "Distinguishing saturation, drift and experimental carryover", "tension": "More flexible fits do not identify a mechanism without intervention support" }, { "researchers_and_school": "Market design: Martin Bichler, Arthur Delarue, marketplace-interference researchers", "contribution": "Search frictions, platform control, shared capacity and matching", "tension": "A treated buyer's gain can impose congestion or losses on others" }, { "researchers_and_school": "Performative learning: Tijana Zrnic, Jaewook Lee, Ioannis Anagnostides and colleagues", "contribution": "Predictions/actions change the distribution being learned", "tension": "Stable, optimal and tractable are different targets; fixed equilibrium need not exist" }, { "researchers_and_school": "Human control and incentives: Chatrath, Zhu, Fan; Joshua Gans, Richard Holden; human-reliance researchers", "contribution": "Policy qualification, concealment, evidence, authority and incentives", "tension": "More review is not automatic synergy; correct flags do not guarantee enforcement" }, { "researchers_and_school": "Knowledge networks: Daron Acemoglu, Tianyi Lin, Asuman Ozdaglar, James Siderius", "contribution": "Shared aggregation and recycling of beliefs", "tension": "Faster agreement can undermine independent information under specific network assumptions" }, { "researchers_and_school": "Organizational design: Wei Cai, Andrea Prat, Jiehang Yu; Fabrizio Dell'Acqua and collaborators", "contribution": "Coordination, specialization and complementarity", "tension": "Better individual output can shift the constraint to review, implementation or organizational learning" }, { "researchers_and_school": "Practitioners: Tim Sanders and G2; NewtonX/Pretzl; Bessemer", "contribution": "Current buyer behavior, operational patterns, packaging and cost hypotheses", "tension": "Commercial observations provide useful hypotheses but not universal causal parameters" }, { "researchers_and_school": "Deployment decisions: Max Farrell, Malika Korganbekova, Sanjog Misra; Charles Manski, Elea Feit and Ron Berman as foundations", "contribution": "Economic choices, testing costs, ambiguity and regret", "tension": "A significance threshold, a payoff-maximizing decision and a robust decision answer different questions; robustness preferences require units and sensitivity" }, { "researchers_and_school": "Phase-specific subscription research: Kreimer, Stoppacher, Hauser, Foscht and Eisingerich", "contribution": "Separating habits, trial benefits, conversion and retention", "tension": "A usage association is neither a habit intervention nor proof of active value under automatic renewal" } ], "corrections": [ { "id": "FIX01", "location": "conservation_law and business generalization G1", "retired_problem": "Customer stock and revenue expansion are mixed", "replacement": "Use logo acquisition minus logo churn for customer stock; represent expansion/contraction in the revenue walk", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX02", "location": "conservation_law and G4", "retired_problem": "Subscription is said always to have an equilibrium; equality cases are misstated", "replacement": "Interpret the original g equation as monetary value, not logos. Its finite attracting equilibrium requires δ>g; equality with a>0 gives linear growth, with a=0 every stock is an equilibrium and none attracting. For g>δ, generic positive stock/input grows; zero is unstable when a=0", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX03", "location": "funnel_operator", "retired_problem": "Displayed LaTeX omits price; weakest stage becomes a universal bottleneck theorem", "replacement": "Include p̄. Allocate by achievable incremental net value per resource, with downstream quality/capacity effects", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX04", "location": "Funnel definitions and operator comparisons", "retired_problem": "Stage thresholds can change the apparent bottleneck without changing final outcomes", "replacement": "Version stage predicates and keep eligible cohorts and end-to-end outcomes comparable; allow useful filtering", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX05", "location": "unit_economics and related prose", "retired_problem": "Retention is claimed to enter the displayed ratio quadratically", "replacement": "The stated form contains 1/δ. A dominance claim requires marginal effects, costs, constraints and a common horizon", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX06", "location": "sales_velocity", "retired_problem": "Called an identity with four independent causal levers", "replacement": "Label a throughput approximation requiring compatible cohorts, stationarity and cycle definitions; drafting time need not shorten procurement time", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX07", "location": "ltv_corrected", "retired_problem": "Beta mixture and geometric worked examples use different initial churn", "replacement": "Beta(2.5,6) implies 29.4% initial mean churn. Duration including the initial period is 5 versus 3.4; future survival-weighted periods are 4 versus 2.4 before discounting. Remove the universal 2–3× overstatement claim and specify whether initial payment is included", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX08", "location": "pmf_piecewise, phi_ex_ante", "retired_problem": "High GRR/low satisfaction is treated as identifying switching costs; log interactions are used to distinguish additive from multiplicative revenue", "replacement": "Keep switching-cost mechanisms as hypotheses. Compare explicit competing functional forms and interventions; a multiplicative power model is already additive after taking logs", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX09", "location": "buyer_state_coupling", "retired_problem": "Conditional win probability is called the true deconfounded estimand", "replacement": "Label it descriptive. Specify potential outcomes and pre-treatment controls/assignment for any causal claim; opportunity and shortlist can be selected post-treatment variables", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX10", "location": "T12, T16, agent recalibration", "retired_problem": "The 95–5 in-market rule becomes 95% pre-contact value creation", "replacement": "Remove that allocation. Keep shortlist and awareness observations with their actual denominators; estimate marginal intervention effects by phase", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX11", "location": "brand_stock", "retired_problem": "A 10–20% annual decay is paired with a 1–3 year half-life", "replacement": "Continuous exponential decay gives about 3.5–6.9 years; discrete annual decline gives about 3.1–6.6. Keep SaaS parameters explicitly unmeasured", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX12", "location": "brand_stock, attribution_convolution", "retired_problem": "Stock and lagged flow are described as distinct physics identifiable by a pause", "replacement": "A linear stock's solution is an exponential convolution. Distinguishing richer mechanisms requires additional assumptions, observations and interventions", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX13", "location": "saturation, motion rules", "retired_problem": "A particular convex finite-TAM curve or ACV step threshold is universalized", "replacement": "Keep candidate response families and business-specific thresholds with empirical support. Category growth, network effects, creative change and capacity alter the shape", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX14", "location": "plg_loop", "retired_problem": "A generation recurrence is followed by a cumulative geometric sum without naming the change in target; absence of public k≥1 cases becomes impossibility", "replacement": "Separate active-user stock, newly referred generations and cumulative referrals; specify retention/deduplication bases. Carry sustained supercritical diffusion as an unqualified boundary scenario rather than an established impossibility", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX15", "location": "noise_floor", "retired_problem": "Assumed 25–40% variance is used as a universal irreducible floor", "replacement": "Keep it as a scenario assumption. Estimate forecast error by population, target, horizon and regime; data availability is not irreducible randomness", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX16", "location": "E-A1 and E-A2", "retired_problem": "One autonomy scalar and an invented catch-rate parabola determine review economics", "replacement": "Use task-specific authority and measured review response, including workload, case mix, expertise, human error and intervention effects", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX17", "location": "E-A3", "retired_problem": "(1−p)^n is tagged an identity; dependence is said always to reduce reliability", "replacement": "Use conditional chain probabilities. Dependence can raise or lower all-success probability at fixed marginals; also evaluate terminal repairable outcomes and severe common failures", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX18", "location": "E-A4", "retired_problem": "Expected cost per attempt is called cost per completed task", "replacement": "Divide total eligible costs, including failed attempts, by independently accepted successes. Zero accepted successes makes the ratio undefined rather than cheap", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX19", "location": "E-A4 and T29", "retired_problem": "Minimizing expected cost alone selects autonomy; a worked sign flip becomes a governance precondition theorem", "replacement": "Add reliability, process, authority, severe-risk and capacity constraints. The illustrative functional form establishes no universal governance threshold or optimal review level", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX20", "location": "E-A5", "retired_problem": "Inference/activity savings flow directly into CAC", "replacement": "Meter all labor, tools, retries, maintenance and incident costs; distinguish CPQO from CAC and realized savings from released hours", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX21", "location": "E-A6 and learning layer", "retired_problem": "Human feedback reduces coefficient drift through an unfitted multiplier", "replacement": "Distinguish feedback receipt, correct labels, policy implementation, selection, missing outcomes, changing distributions and actual measured learning", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX22", "location": "Proxy outcomes and decision loop", "retired_problem": "Proxy correlation, a non-rejected diagnostic or statistical significance can become a decision certificate", "replacement": "Qualify proxies against mature outcomes under explicit labeling/transport assumptions; specify economic deployment criteria while preserving inferential uncertainty", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX23", "location": "E-A7 and T30", "retired_problem": "Aggregate adoption necessarily increases noise/pollution", "replacement": "Make the sign open. Measure volume, relevance, demand expansion, platform responses, interference and external costs separately", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX24", "location": "E-A8 and T15", "retired_problem": "CAC is described as cost per qualified opportunity; matching is treated as a holdout; hybrid superiority and autonomous seller failure are categorical", "replacement": "Define CAC and CPQO correctly. Use eligible policy comparisons and contribution endpoints. Retain the narrow unresolved autonomous-SDR ROI question; qualify hybrid claims, and do not confuse human-routing evidence with autonomous SDR proof", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX25", "location": "E-A9", "retired_problem": "One buyer-agent fraction stands in for an AI-mediated buying committee", "replacement": "Observe role, task, phase, platform, fidelity, authority, verification and fulfillment separately", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX26", "location": "Epistemic legend and mrr_walk", "retired_problem": "Only MRR is called a true identity despite other definitions/accounting equalities", "replacement": "Separate definitions, accounting equalities and behavioral estimators. Reconcile like-for-like recurring revenue bases; GAAP revenue is not the same target as MRR", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX27", "location": "State variables and equations", "retired_problem": "Symbols collide across layers and periods", "replacement": "Specify units and namespace: spend versus switching cost, autonomy versus brand build rate, group size versus governance, revenue versus reliability, and growth variance versus outreach noise", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX28", "location": "org_cognition, operator", "retired_problem": "Scalar alignment/minimum laws imply causal dominance without compatible measurement", "replacement": "Use measured coordination, handoff, queue, authority and incentive mechanisms; retain unknown coefficients and avoid comparing quantities in different units", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX29", "location": "compliance epistemic types", "retired_problem": "Jurisdiction-dependent obligations and permission gates are labeled identities", "replacement": "Represent conditional constraints with jurisdiction, channel/action, date, policy and authority provenance; separate these from accounting equalities and fitted response coefficients", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX30", "location": "business_model_generalization_v4", "retired_problem": "Single binding gate and kind templates retain overly broad claims", "replacement": "Allow simultaneous cash/capacity/quality gates. Licensing can have customers and CAC; negative contribution extends beyond lending; hardware needs inventory/backlog/installed-base distinctions", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." }, { "id": "FIX31", "location": "Meta and provenance", "retired_problem": "Vocabulary count and obsolete backup notes conflict with current state", "replacement": "Reconcile 195 versus 219 descriptions; use Git commit history for versions, as repository instructions require; do not create sibling backup models", "implementation_scope": "Current canonical equations, theses, layers/operator contracts and directly conflicting consumers; historical version pages remain historical." } ] }, "decision_contract": { "status": "specified", "evidence_states": [ "measurability_protocol_defined", "observed_with_provenance", "estimated_with_uncertainty", "identified_for_declared_effect", "qualified_for_declared_policy_and_environment" ], "states_are_separate": true, "causal_target": "Expected incremental contribution per eligible assigned account or appropriate joint/cluster policy over declared horizon, including alternative/no-action comparator where feasible.", "measurement_schema": [ "account/contact/contract/usage identity and lineage", "eligibility and segment/stage predicate version", "pre-action feature and source/freshness snapshots", "assignment/action/policy probability when design supplies it", "observed exposure, action receipt and task/tool/configuration version", "role/task/phase buyer mediation, authority and policy verification", "criterion-specific review, correction, repair, defer and stop records", "complete eligible attempt costs and acquisition/delivery allocation", "independently accepted terminal/process outcome and linked business outcome", "missingness/censoring/booking coverage, maturity and reconciliation residual" ], "unknown_rule": "Do not substitute benchmark constants, synthetic labels or governance scores for unavailable material inputs. State unknown, use a explicitly labeled sensitivity scenario or refuse the dependent action." }, "symbol_conventions": { "rule": "canonical_symbol gives unambiguous state namespaces; symbol preserves a legacy display alias. Equation-local parameters must be interpreted in their declared scope, never joined globally by a repeated letter.", "legacy_collisions": { "S": [ "spend_acquisition", "S_switching", "survival_S(t)" ], "alpha": [ "alpha_autonomy", "build_brand", "sBG_alpha_dropout" ], "beta": [ "sBG_beta_dropout", "saturation_beta" ], "R": [ "funnel_value", "R_chain", "retention_fraction" ], "k": [ "k_referral", "kernel_lag" ], "theta": [ "forecast_error_fraction", "theta_attention" ], "delta": [ "delta_logo", "delta_value", "decay_brand", "decay_attention" ], "G": [ "buying_group_size", "G_governance_summary", "referral_generation_G_t" ], "lambda": [ "lambda_learning", "brand_shortlist_response" ] }, "units_rule": "Every applied numeric parameter needs units, period, outcome/population, observed or estimated status and evidence scope; absent local values remain unknown." }, "evidence_record_contract": { "required_for_an_applied_claim": [ "target_population_and_business_kind", "outcome_denominator_units_and_horizon", "intervention_control_and_assignment_unit", "study_dates_publication_and_revision_dates", "access_depth_and_source_provenance", "uncertainty_support_selection_missingness_and_interference", "model_tool_data_policy_cost_and_authority_snapshot", "identification_assumptions_and_transport_boundary", "falsifier_expiry_and_requalification_trigger" ], "registry_scope": "Current evidence_register entries are source appraisals; they are not completed local deployment certificates. Study calendar dates, unavailable denominators or full methods remain unknown where not disclosed. Detailed appraisals and older foundations are in the linked research report.", "promotion_rule": "Promote a source into a fitted local claim only after documenting the full decision-specific record and its unresolved fields; current registry IDs alone do not supply missing identification or calibration." } }