Customer lifetime value (LTV)

Every scaling company has a number they are terrified to calculate: How much profit do we actually make from a customer?

Not revenue. Profit. The money left after you’ve paid the cost to serve them, support them, host them, and keep them. That number is lifetime value. And it determines whether your company scales or whether you scale yourself into bankruptcy.

LTV is not an academic metric. It is the upper bound on how much you can spend to acquire a customer. If your LTV is $100,000 and you spend $200,000 to acquire a customer, you will never be profitable. If your LTV is $100,000 and you spend $30,000, you have room to scale. The gap between what you spend (CAC) and what you earn (LTV) determines your margin for error, your runway, and ultimately, whether you can build a scaling machine.

The founder mistake is simple and universal: measure blended LTV, see a number that looks okay, and assume you are fine. You are not fine. You are blind. The LTV that matters is not the average. It is the curve by cohort, by segment, and by expansion dynamic. One segment might have a healthy LTV. Another might be a black hole. Blended LTV hides both.

What LTV is — and where it comes from

LTV is the total profit a customer generates over their customer lifetime. It is not revenue. It is profit after all the costs to serve them.

The formula is simple:

LTV = (ARPU × Gross Margin) / Monthly Churn Rate

Let’s break this down:

ARPU (Average Revenue Per User per month). Not the one-time contract value. The recurring revenue per user per month. If a customer pays $10,000 per year, ARPU is $833/month. If they pay $100/month, ARPU is $100.

Gross Margin. The percentage of revenue left after cost of goods sold. For a SaaS company, this is typically 70-85% (you have hosting costs, payment processing, maybe some support). For a services company, it might be 40-60% (labor is expensive). Gross margin sets the absolute ceiling on how much profit a customer can generate. If gross margin is 50%, a customer paying $100/month will never generate more than $50/month in profit, no matter how long they stay.

Monthly Churn Rate. The percentage of customers who leave per month. If you have 100 customers and lose 2 per month, your churn rate is 2%. This is where the formula gets interesting. A company with $100 ARPU and 1% monthly churn has a fundamentally different LTV than a company with $100 ARPU and 5% monthly churn. One can build a scaling machine. The other is pouring water into a bucket with a hole in the bottom.

Using the formula:

• Company A: $100 ARPU × 75% gross margin / 1% monthly churn = $7,500 LTV
• Company B: $100 ARPU × 75% gross margin / 5% monthly churn = $1,500 LTV

Same ARPU. Same gross margin. Churn is the only difference. Company A can spend up to $7,500 to acquire a customer and still break even (though it would not be wise). Company B can spend $1,500. If both companies spend $5,000 per customer acquisition (CAC), Company A is profitable. Company B is bankrupt.

This is why churn is the hidden killer in growth companies. It is not visible until it is not. Everyone sees the new revenue coming in. No one sees the old revenue quietly draining out the back door.

The diagnostic: LTV breaks down at high churn

The formula above is naive. It assumes customers stay forever, which they do not. At high churn rates, the formula overstates LTV significantly.

The issue is that the formula calculates average customer lifetime, not actual customer lifetime. A customer at 2% monthly churn has an average lifetime of 50 months (1 / 0.02). A customer at 5% monthly churn has an average lifetime of 20 months. But customers do not stay for their “average” lifetime. They either stay or they leave. The distribution matters.

The more accurate formula is based on cohort curves. Instead of assuming a linear decay, you plot the actual retention curve for a cohort and calculate LTV from the curve:

LTV (cohort-based) = Sum of (ARPU × Gross Margin × Retention % × Month)

Example: a cohort of 100 customers with $1,000 ARPU, 70% gross margin, and the following retention curve:

Sum the contribution to get cohort LTV: $70,000 + $66,500 + $64,400 + … (all months until churn reaches zero or asymptotes) = $369,400 total margin per cohort of 100 customers = $3,694 LTV per customer.

This is more accurate than the naive formula because it accounts for the actual shape of the retention curve. Early churn hits harder (you lose customers before they pay for themselves). Late churn matters less (they have already paid back their CAC).

The takeaway: measure LTV from your cohort retention curves, not from a formula. The formula is a quick diagnostic. The curve is the truth.

The ceiling on LTV: what you cannot fix

LTV has a hard ceiling. You cannot improve it beyond what your economics allow.

LTV is constrained by three factors: ARPU, gross margin, and churn. You can only move one of three directions.

Direction 1: Raise ARPU.

Increase the price per customer, or increase their expansion within the account. If your $100/month customer becomes a $150/month customer, LTV goes up proportionally (assuming churn and margin are constant).

But there is a limit. You cannot raise price indefinitely without losing customers (see churn). And you cannot expand a customer indefinitely; they will eventually max out on your product or find an alternative.

Direction 2: Improve gross margin.

Lower your cost of goods sold. This is typically slower to move than ARPU (it requires operational efficiency or platform economics), but it has a big impact. A 5-point gross margin improvement (from 70% to 75%) directly improves LTV by 7%.

But gross margin also has a limit. You cannot get below the cost of infrastructure, payment processing, and minimal support.

Direction 3: Reduce churn.

Lower the percentage of customers who leave per month. This is often the biggest leverage, because churn appears in the denominator. Cut churn by half, and LTV doubles (all else equal).

But churn also has a floor. You cannot eliminate churn. Some customers will always leave (they go out of business, their needs change, they find a better alternative). A healthy business might achieve 1-2% monthly churn. An excellent business might achieve 0.5% monthly churn. Getting below that requires targeting only customers with zero alternatives.

The ceiling is the product of all three: (Max ARPU per segment) × (Achievable Gross Margin) / (Sustainable Churn Rate).

If you are in a market where ARPU is capped at $5,000/month per customer, gross margin is 60%, and sustainable churn is 2%, then your LTV ceiling is $150,000. You cannot exceed it no matter how hard you try. You cannot scale acquisition beyond the point where CAC exceeds that ceiling.

This is why market selection constrains economics. If you pick a market where ARPU is low (consumer, SMB), you must either achieve very low churn or very high gross margin. If you pick a market where churn is high (contract work, commodities), you must have very high ARPU or margin. There is no escape. You must understand your ceiling before you design your go-to-market.

The diagnostic: LTV by segment and cohort

Blended LTV is a lie.

It is the average of healthy and sick segments, scaling and dying segments, efficient and inefficient acquisition cohorts. It is the number that lets you sleep at night while your company is slowly dying.

The founder mistake: a founder calculates blended LTV of $80,000. CAC is $30,000. LTV:CAC is 2.7x. They think they have a problem. But digging deeper, they find:

• Enterprise segment: $150,000 LTV, $40,000 CAC, 3.75x ratio. Healthy, scaling.
• Mid-market segment: $60,000 LTV, $25,000 CAC, 2.4x ratio. Okay, growing.
• Startup segment: $15,000 LTV, $35,000 CAC, 0.43x ratio. Unprofitable, losing money per customer.

The blended number hides a catastrophe. The startup segment is destroying the company. Every new startup customer acquired costs the company more than they will ever pay back. The founder is scaling into bankruptcy.

The fix: measure LTV by segment separately, by cohort separately, by acquisition channel separately.

Create a matrix:

This matrix tells you:

1. Enterprise is your scaling engine. Double down.
2. Mid-market is showing stress: CAC is rising and LTV is falling. Investigate the cause (worse fit, worse positioning, worse conversion).
3. Startup is a black hole. Stop acquiring startups until you fix the unit economics.

Without this matrix, the founder sees the blended 2.7x, thinks everything is fine, and scales the unprofitable segment to scale the profitable segments. This destroys the whole business.

The diagnostic: expansion revenue and LTV growth

Most LTV calculations assume a customer’s ARPU is fixed. But in a healthy company, ARPU should increase over time as customers expand.

If your customer starts at $100/month ARPU and expands to $150/month by month 12, their LTV is not based on $100. It is based on the expansion curve.

This is the founder mistake: not tracking LTV by expansion dynamic.

Scenario 1: No expansion.

• Starting ARPU: $100/month
• LTV: (100 × 70%) / 2% = $3,500 per customer

Scenario 2: 15% average expansion per year.

• Starting ARPU: $100/month
• Year 1 ending ARPU: $115/month
• Year 2 ending ARPU: $132/month
• Year 3 ending ARPU: $152/month (if churn is low enough)
• LTV: much higher, because the customer base is growing in value over time.

If expansion is strong (15%+ net revenue retention), your LTV is growing. If expansion is flat, your LTV is static. If expansion is negative (customers are shrinking), your LTV is falling.

This is why net revenue retention (NRR) is often more important than logo retention. A company with 95% logo retention but 110% NRR is growing. A company with 98% logo retention but 95% NRR is shrinking.

The diagnostic: measure LTV by expansion cohort.

• High-expansion cohorts: NRR > 110%. These customers are expanding fast. LTV is growing. Scale hard.
• Medium-expansion cohorts: NRR 100-110%. These customers are stable or slightly expanding. LTV is stable or growing slowly.
• Low-expansion cohorts: NRR < 100%. These customers are shrinking. LTV is falling even if churn is constant.

If your blended LTV looks good but you have a low-expansion cohort, your LTV will fall in 12 months as that cohort ages. You are not measuring a growing asset. You are measuring a declining one.

How to predict LTV early: cohort curves

The founder mistake: waiting 24-36 months to measure LTV.

By then, you have already made all the acquisition decisions. You have already spent millions. And if LTV is lower than you thought, you are already bankrupt.

The fix: predict LTV early using cohort curves.

A cohort curve shows the retention percentage by month for a single cohort of customers. Plot this starting from month 1:

• Month 1: 100% (all customers here)
• Month 2: 95% (5% churned)
• Month 3: 92% (3% of remaining churned)
• Month 6: 82% (retention is flattening)
• Month 12: 70% (long-tail churn is slow)

The shape of this curve tells you what LTV will be at maturity. If the curve is steep (lots of churn early), LTV will be low. If the curve is shallow (customers stick around), LTV will be high.

By month 3-6, you can already predict month 24+ retention with reasonable accuracy. If you see a steep curve, you know LTV will be low. Fix the product, the motion, or the ICP before you scale acquisition.

Use this formula to project LTV early:

Projected LTV = (ARPU × Gross Margin) × Σ(Month retention % × Month)

If your month-3 retention is 92% and you extrapolate that curve to month 36, you can estimate LTV with a 20-30% margin of error. That is good enough to decide whether to scale.

Founder mistakes: the LTV blindness patterns

Mistake 1: Not measuring LTV by cohort.

A founder calculates blended LTV and assumes it applies to all acquisition. It does not. Each acquisition channel, each segment, each time period has different cohort behavior. One cohort might have 70% month-12 retention. Another might have 40%.

Fix: measure LTV separately for each cohort from the beginning. Do not blend. Do not average. Measure the curve for each cohort and hold each accountable.

Mistake 2: Not tracking expansion in LTV.

A founder calculates LTV based on starting ARPU and ignores expansion. The customer who starts at $100/month and ends at $150/month has higher LTV than the customer who stays at $100. But if you measure both as $100 ARPU LTV, you underestimate the expanding customer and overestimate the flat one.

Fix: measure LTV by expansion cohort. Track NRR by cohort. If NRR is 110%, your LTV is growing. If it is 95%, your LTV is falling.

Mistake 3: Ignoring churn trends.

A founder looks at the current 2% monthly churn and calculates LTV based on it. But churn is rising (2.0% last month, 2.3% this month, 2.6% next month). LTV is falling even though the current number looks okay.

Fix: measure churn month-over-month and predict forward. If churn is rising, flag it. Do not calculate “what if churn stays constant.” Calculate “what if the trend continues.”

Mistake 4: Treating LTV as a static target.

A founder calculates LTV once per year in the annual planning cycle. It is useless. LTV moves with every change in churn, ARPU, or margin. A competitor launches a feature and churn rises. LTV falls. Your support quality improves and churn drops. LTV rises. LTV should be measured monthly and tracked like revenue.

Fix: measure LTV monthly by cohort. Track the trend. Alert when LTV is falling.

Mistake 5: Not understanding the LTV ceiling.

A founder wants to scale acquisition but does not understand that LTV has a hard ceiling based on ARPU, margin, and churn. They think they can lower CAC by improving funnel conversion, but they cannot exceed the LTV ceiling. If they do, they are unprofitable.

Fix: calculate your LTV ceiling first. Then design your acquisition plan to stay below it. If the ceiling is too low to be sustainable, change the market, the motion, or the pricing.

Rules: the LTV checklist

1. Measure LTV from cohort retention curves, not from the formula. The formula is a quick diagnostic. The curve is truth.

2. Measure LTV by segment separately. Blended LTV hides disasters. Enterprise might be 4x, mid-market might be 2x, SMB might be 0.8x. Do not average them. Measure each.

3. Measure LTV by cohort. Each acquisition cohort has different retention. Q1 2024 might be 70% month-12 retention. Q2 2024 might be 60%. Track the trend.

4. Track expansion in LTV. If NRR is growing, LTV is growing. If NRR is falling, LTV is falling. Do not ignore expansion.

5. Predict LTV early from month 3-6 cohort curves. Do not wait 24 months to learn your LTV is terrible. Extrapolate early and fix before you scale.

6. Alert on churn trends, not absolute churn. A rising churn trend is more dangerous than a high static churn. If churn is 2% and steady, that is okay. If churn is 1.5% and rising, that is a warning sign.

7. Measure LTV monthly. LTV should be a monthly dashboard metric, not an annual calculation. It moves with your business.

8. Understand your LTV ceiling. Calculate the theoretical maximum LTV based on ARPU, gross margin, and sustainable churn. Do not expect to exceed it. Design your CAC and acquisition plan around this ceiling.

9. LTV:CAC determines your scaling ceiling. Below 1.5x: unprofitable, do not scale. 1.5-2x: breakeven + some growth, careful. 2-3x: healthy, scale. 3x+: very healthy, scale aggressively.

10. Use LTV as a diagnostic tool, not a vanity metric. When LTV is falling, something is broken: product, ICP, positioning, or pricing. Find the cause before it becomes a catastrophe.

With LTV measured and understood, you now have the two halves of unit economics: what you spend to acquire (CAC) and what you earn over their lifetime (LTV). The ratio between them—LTV:CAC—determines whether you can scale sustainably or whether scaling destroys profitability. That is the next node.

Coming next: LTV:CAC Ratio — why 3x is the magic number, why anything below 1.5x is broken, and how to diagnose which lever to pull when the ratio falls out of range.