Unit economics fundamentals

Unit economics is the answer to the hardest question in business: does this work?

Not “is the product good” or “do customers like us” or “are we growing.” The hardest question is: when we acquire a customer, do we make money? And if we do, do we make it fast enough to stay alive while we find more customers?

Everything that follows—scaling, hiring, which markets to pursue, whether to pivot—is noise if unit economics are broken. A company with perfect unit economics and slow growth can fix slow growth. A company with broken unit economics and fast growth is just dying faster.

What unit economics are

Unit economics are the costs and revenues associated with one customer, measured over their lifetime with you.

The math is deceptively simple:

LTV - CAC > 0?

If customer lifetime value (the total profit you make from a customer over the years they stay) exceeds customer acquisition cost (the fully loaded cost to land that customer), the business works. If not, it does not.

But “works” is a threshold, not a destination. The real question is: how much does it work, and how fast?

A customer that costs $1,000 to acquire and generates $1,001 in lifetime profit technically works, but you will run out of cash before you scale. A customer that costs $1,000 and generates $50,000 in lifetime profit works with room to invest in growth. The gap between CAC and LTV is your runway, your growth capacity, and your strategic option value.

Why unit economics matter more than growth rate

Every founder has seen a pitch deck with a hockey-stick revenue curve. Every founder has been taught to chase growth at all costs. But growth without unit economics is a accelerant on a fire you are standing in.

Here is the mechanism:

You raise money. You hire a sales team. You spend $1 million to acquire customers. This generates $800,000 in first-year revenue. Your unit economics are broken by 20%. You hide this by lowering the payback threshold, extending the contract term, or counting ARR instead of cash. You raise more money. You hire more salespeople. You spend $5 million to acquire customers. You are now losing $1 million per year, accelerating.

Growth metrics—MRR growth, ARR growth, customer count—hide unit economics failure. A company can grow 400% a year and go bankrupt because every customer it acquires loses money.

The opposite is also true: a company with boring, predictable unit economics (you spend $2,000 to acquire a customer, make $8,000 over their lifetime, have 18 months of payback) can grow slowly, survive recessions, compound into a large business, and never need another round of venture capital.

Unit economics are the engine. Growth is just the speed you run the engine. If the engine is broken, speed only matters because you will crash faster.

The unit economics diagnostic matrix: by motion type

Unit economics vary by motion. A PLG company and a sales-led company have completely different cost and revenue structures, and thus completely different unit-economics problems.

Here is how to diagnose where you stand:

Product-Led Growth (PLG)

Revenue structure: Low contract value, high customer count, viral acquisition.

CAC: $100–$1,000, or free if viral coefficient > 1.

LTV: $5,000–$50,000 (depends on retention curve).

Payback: 3–12 months (or negative if viral).

Diagnostic question: Is your viral coefficient >= 1.3?

If it is, you are compounding acquisition cost over payback into a strong LTV. If it is not, your CAC is the only way to scale, and it must be very cheap. Most PLG companies fail here: they assume virality will appear as a gift, not realizing they must engineer it into the product.

Where PLG breaks: Assuming a “freemium” funnel is automatic (it is not). Viral coefficient of 1.0 or lower (each user brings in less than one new user; you are in a slow decline). High churn in the first 30 days (payback is too long). Turning on paid to rescue unit economics (if you need paid CAC in PLG, your motion is wrong).

Sales-Led

Revenue structure: High contract value, lower customer count, direct sales effort.

CAC: $10,000–$100,000+ (fully loaded: salary, commission, marketing, onboarding).

LTV: $100,000–$500,000+ (depends on retention and expansion).

Payback: 12–24 months (sometimes longer).

Diagnostic question: Is your ACV >= 3–5x your fully loaded CAC?

Sales-led companies are profitable when LTV/CAC >= 3. At that ratio, even with 30% churn, you are cash-positive within payback and can deploy the model at scale.

Where sales-led breaks: Underestimating true CAC (forgetting sales commissions, marketing spend, onboarding, customer success). Cherry-picking your largest deals (CAC looks good on flagship accounts; it is brutal on mid-market). Selling to accounts too small for your CAC (a $50,000 CAC to close a $30,000 ACV deal destroys the model).

Land and Expand

Revenue structure: Low initial contract value, high expansion potential, long customer lifetime.

CAC: $500–$5,000 (low, because initial contract is small or land is product-led).

LTV: $50,000–$500,000 (depends entirely on expansion and retention).

Payback: 6–24 months (on initial contract, then expansion shortens it further).

Diagnostic question: Is your expansion rate (net revenue retention) > 110%?

Land and expand only works if you expand. If NRR is < 105%, you are a low-ACV, high-churn business with broken unit economics. If NRR is > 120%, your LTV is compounding and payback accelerates as you scale.

Where land and expand breaks: Confusing “land” with “get free users” (land must be profitable or at least fast payback). Assuming expansion will happen (it will not; you must engineer it into the product and sales process). Expanding within the wrong accounts (you expand where the initial customer succeeded, not everywhere).

Partner/Channel-Led

Revenue structure: Medium-to-high contract value, low direct CAC (partner covers acquisition), long sales cycle.

CAC: $2,000–$20,000 (partner takes a cut, so your all-in cost is lower, but the partner’s margin comes out of your pocket).

LTV: $50,000–$200,000 (depends on how well the partner integrates you).

Payback: 12–36 months.

Diagnostic question: Does partner margin + your CAC stay well below your LTV?

Channel economics are tricky because you pay the partner 20–30% of the first year, which is expensive acquisition upfront but predictable. The real question is whether that deal is profitable after partner payment.

Where channel-led breaks: Partner margin exceeds LTV (you are paying 40% commission and your LTV is too close to ACV to absorb it). Relying on a single partner (if one partner is 50%+ of your revenue, you have a channel concentration risk, not a motion). Partners acquiring the wrong customers (partner has incentive to book deals, not fit; you get high churn).

The diagnostic rule

For any motion: LTV / CAC >= 3 is a minimum floor for profitability and sustainable growth. Below 3, you are betting on cost reduction or LTV improvement, not on a working model. Above 5, you have strategic optionality: you can invest in growth, improve your product, or just enjoy margins.

How to calculate CAC accurately (and why founders get it wrong)

CAC looks simple: (Sales + Marketing spend) / (Customers acquired).

But every founder I know calculates it wrong, because they forget what “fully loaded” means.

The CAC formula (done right)

Fully loaded CAC = (Sales salaries + commissions + benefits + office + SaaS tools + marketing spend) / (net new customers acquired that period)

Notice what this includes:

• All sales compensation, not just commission.
• Sales operations, sales engineering, SDR team—all the people needed to close.
• Marketing spend allocated to the motion you are measuring (if you run both PLG and sales-led, split the spend).
• Onboarding and customer success spend dedicated to acquiring customers.
• If you have a demand generation team, they live in CAC; do not leave them out.

Why founders underestimate CAC

Mistake 1: Forgetting to load the salary.

A sales rep with $100,000 salary, $40,000 commission, 30% benefits, $10,000 SaaS tools, and $20,000 office overhead costs $199,000 per year. If the rep closes 20 deals per year, that is $10,000 CAC just from salary and overhead, before a single marketing dollar is spent. Many founders quote a $2,000 CAC because they only divided marketing spend by customer count.

Mistake 2: Not accounting for ramp time.

A salesperson hired today is not productive for 3–6 months. They are on salary the whole time, but generate zero revenue. When you calculate CAC, include their ramp cost. A $100,000 sales hire with 6 months of ramp is $50,000 of cost before they close their first deal. That is CAC you are paying upfront.

Mistake 3: Ignoring failed sales cycles.

You close 20 deals per year, but you touched 100 leads. The people, tools, and time to reach the other 80 are still part of your CAC. Divide total sales and marketing cost by net new customers acquired, not by deals closed this period.

Mistake 4: Attribution gaming.

You run both an inbound marketing motion and an outbound sales motion. A lead comes in from content, then a sales rep closes it. Did marketing acquire the customer, or did sales? For unit economics, it does not matter—both motions are part of your cost to acquire. Do not undercount by allocating only to one channel.

Mistake 5: Confusing blended CAC with paid CAC.

Some customers come through partnerships, product-led funnels, or direct sales with no marketing spend. When you average this with expensive paid-acquisition channels, you get a blended CAC that hides the fact that paid acquisition is actually breaking your unit economics. Segment by channel and motion, not blended.

The payback rule

Once you have true CAC, calculate CAC Payback = CAC / (monthly revenue per customer - COGS - customer success cost).

If your fully loaded CAC is $10,000, your monthly revenue is $1,000, and your COGS + support costs are $300, then payback = $10,000 / ($1,000 - $300) = $10,000 / $700 = 14.3 months.

Fourteen-month payback is acceptable in enterprise (where LTV is 5+ years). It is unacceptable in mid-market (where LTV might be 2–3 years). It is a death sentence in SMB (where LTV is often 18 months).

Payback governs whether you have cash to grow. If you cannot reach payback before your cash runs out, the unit economics are broken.

How to calculate LTV without overestimating lifetime

LTV is where most founders lie to themselves.

The formula looks simple: LTV = (annual profit per customer) × (customer lifetime in years).

But “lifetime” is where the fiction begins. Founders assume customers stay forever. They do not. The median B2B SaaS company has a gross churn of 3–5% per month, which means 50% of your cohort is gone within 15 months.

LTV formula (done right)

LTV = (gross margin per customer per month) / (monthly churn rate)

Or, more intuitively:

LTV = (gross margin per customer per month) × (average customer lifetime in months)

But you must know your actual churn, not your fantasy churn.

How founders overestimate LTV

Mistake 1: Using gross churn instead of net churn.

Gross churn is the percentage of customers you lose each month. Net churn accounts for expansion. If you lose 5% of customers but expand in 80% of the remaining base, your net churn is negative (you are growing). For LTV, use gross churn. If you assume the average customer stays for 4 years because your net churn is negative, but 50% of your customers actually churn out within 18 months, you are overestimating the lifetime of the customer base.

Mistake 2: Confusing cohort retention with company retention.

Your most recent cohort (customers acquired in the last 3 months) might have zero churn. Your oldest cohort (customers acquired 3 years ago) might have 80% churn. When you calculate LTV, use the churn curve of a mature cohort, not the early retention of a new cohort.

Mistake 3: Assuming margin stays constant.

As you scale, your margin per customer changes. Early customers pay high prices and have low support costs. Later customers are discounted, and support is more expensive. Your $1,000 ACV customer generates $600 gross margin today, but future cohorts at $800 ACV might only generate $400 margin. When you calculate LTV, use the margin of a mature customer, not your best customer.

Mistake 4: Not accounting for the shelf-life of the product.

A CRM software might have a 5-year customer lifetime because CRM is sticky and your switching costs are high. A trendy analytics tool might have an 18-month lifetime because it becomes commodified or the company moves to a different vendor. When you estimate lifetime, ask: where is this category in its lifecycle, and do you have defensibility?

The LTV test

Calculate LTV three ways:

1. Optimistic: Best-case churn, best-case retention, no discount.
2. Base case: Observed churn on mature cohorts, average expansion, average discount.
3. Pessimistic: Current worst-case churn in a cohort, no expansion, deep discount trend.

If your base-case LTV is < 3x your fully loaded CAC, do not rely on the optimistic case. The pessimistic case is what happens in a recession.

The three founder mistakes that destroy unit economics

Mistake 1: Ignoring unit economics entirely

The founder focuses on growth. “We will worry about unit economics later.” Later never comes. By the time you measure them, you have acquired 10,000 customers at $10,000 CAC each, and they each generate $2,000 LTV. You are in a $80 million hole.

The rule: Calculate unit economics before you scale the motion. Do not wait for 100 customers to have decent data. Calculate on your first 10. If it does not work at 10, it will not work at 1,000.

Mistake 2: Underestimating CAC

This is the most common mistake. The founder thinks CAC is just the ad spend divided by the number of signups. They forget fully loaded costs, ramp time, failed sales cycles, and the cost of the entire infrastructure behind acquisition.

The true CAC is 2–5x what founders estimate.

The rule: Calculate CAC quarterly and audit it ruthlessly. Include every cost of acquiring a customer, no exceptions. When your calculated CAC exceeds your intuition, the calculation is right and your intuition is wrong.

Mistake 3: Overestimating customer lifetime

The founder assumes customers stay for 5 years because the top 10% do. In reality, the median customer stays for 18–24 months, and cohort retention curves down sharply after month 6.

The rule: Use observed churn, not assumed churn. Run a cohort retention analysis. Find the cohort that is oldest and most mature, and use its churn curve to calculate LTV. If that LTV / CAC is < 3, your unit economics are broken and you must fix them before you scale.

Real examples: how unit economics drive decisions

Example 1: The sales-led SaaS company with broken land

Acme sells project management software to mid-market companies. Land CAC is $15,000 (fully loaded). Initial ACV is $30,000. So far, LTV / CAC = 2x on the initial deal, which is poor.

The company assumes expansion will save them. They project 120% NRR and expect the customer lifetime to be 4 years. Projected LTV = $30,000 * 1.5 (expansion multiplier) * 4 (years) = $180,000. Projected LTV / CAC = 12x. They celebrate.

But the reality: actual expansion happens in 30% of customers, and only in the first year. Actual churn hits 8% per month after the second year. Actual mature-cohort LTV is $60,000, not $180,000.

Actual LTV / CAC = 4x, which is acceptable, but only on the accounts where expansion happens. On the 70% of customers with no expansion, LTV / CAC = 2x, which is a loss.

Decision driven by unit economics: Redesign the motion to focus on accounts with expansion potential (higher ACV, multi-user). Do not try to scale land to everyone; scale to where unit economics work.

Example 2: The PLG company that can’t afford marketing

Acme sells a data analytics tool to startups. Product-led funnel: 10,000 signups per month, 2% conversion to paid, 500 new customers, $500 ACV, $3,000 monthly revenue per customer. Payback is immediate (the product pays for itself in the first month).

But they have terrible retention. 20% of customers churn in month 2. 40% churn in month 3. By month 12, 90% have churned. Monthly churn rate averages 15%.

LTV = $3,000 monthly profit / 15% monthly churn = $20,000.

CAC is $0 because signups are free, and viral coefficient is 0.8 (each user brings 0.8 new users). To grow, they need to pay for traffic. CAC (paid) = $200. LTV / CAC (paid) = 100x, which looks amazing.

But wait: if they run paid ads, they are only profitable if the customer lasts 4 months (LTV) and payback is 1 month. The viral coefficient is so low that they are not compounding acquisition; they are buying every customer.

Decision driven by unit economics: Do not scale paid acquisition. Instead, fix the product to reduce churn and increase viral coefficient. At 15% monthly churn with no virality, you cannot afford paid acquisition. At 5% churn with 1.2x viral coefficient, you can.

Example 3: The sales-led company scaling into SMB too early

Acme sells compliance software to enterprise. Enterprise CAC is $40,000, enterprise ACV is $200,000, payback is 2.4 months, LTV is $1.2 million. Enterprise unit economics are great.

They want to scale, so they open up SMB as a secondary motion. SMB CAC is $8,000, SMB ACV is $30,000, payback is 3.2 months, LTV is $150,000. SMB unit economics are okay (LTV / CAC = 18.75x).

But when they actually measure, SMB churn is 12% per month (vs 1% in enterprise). Actual LTV = $30,000 * 12 months (1 / 0.12) = $30,000. LTV / CAC is now 3.75x, which is barely acceptable and below enterprise.

The SMB team needs to generate 5x more customers to hit the same revenue as the enterprise team. They need to acquire at a lower CAC, but the selling motion is the same. SMB margins are also lower because SMB customers demand discounts.

Decision driven by unit economics: Do not scale SMB with the same sales motion as enterprise. The unit economics do not support it. Either redesign the motion (product-led or inside sales), or focus entirely on enterprise and refer SMB to a partner.

Name the rules

The CAC rule: Fully loaded CAC must include salary, commission, ramp time, overhead, marketing spend, failed cycles, and all supporting functions. If you have not named every person who touches acquisition and divided their all-in cost by customers acquired, you have not measured CAC.

The payback rule: CAC payback must be short enough that you can replenish cash before you run out. For venture-backed companies, 12 months is acceptable; under 12 months is safe. For bootstrapped companies, under 6 months is required.

The LTV rule: LTV must be calculated from observed churn on mature cohorts, not from best-case assumptions or top-decile retention. If you cannot see a 3+ year cohort with 40%+ retention, your LTV estimate is fantasy.

The 3x rule: LTV / CAC >= 3 for sustainable growth. Below 3, you are losing money on acquisition. Above 5, you have room to invest in growth or improve margin. The band between 3 and 5 is the zone where you are playing the unit economics game correctly.

The motion matching rule: Different motions have different unit economics structures. PLG has low CAC but demands high virality or expansion. Sales-led has high CAC but predictable LTV. Land-and-expand has moderate CAC but requires strong expansion. Never judge a motion by its CAC alone; judge the full LTV / CAC picture for that motion.

If you know your unit economics, you know whether your business works. If you don’t, you are flying blind, and no amount of growth will save you. The next chapter explores how revenue moves over time, and how to track every dollar.