Skip to content
Back to Analytics & Experimentation

LTV Calculation

Computing customer lifetime value —the total revenue a user generates over their full relationship with the product —to set acquisition budget, prioritize segments, and measure business health.

Why

LTV is the ceiling on CAC: you can spend up to LTV to acquire a customer and still break even, and the LTV:CAC ratio is the single most important number for a subscription business. If you do not know LTV, you do not know how much you can spend to acquire a user, and you are either under-spending (leaving growth on the table) or over-spending (buying unprofitable users). The calculation seems simple —average revenue per user times average lifetime —but both inputs are distributions, not scalars, and using the average hides the skew: 20% of users generate 80% of LTV, and the median user's LTV is 1/5 of the mean. A cohort-based LTV (computed per signup cohort, tracked over time) is more honest because it shows the LTV curve and its uncertainty, and it lets you detect when a recent cohort's LTV is tracking below prior cohorts (a retention or pricing problem). LTV is also forward-looking: historical LTV is a lagging indicator, and for a fast-growing or changing product, a predictive LTV (modeled from early behavior) is more actionable because it does not require waiting 12 months to know if a cohort is healthy.

When

When setting CAC budget (LTV sets the ceiling). When prioritizing segments (high-LTV segments deserve more acquisition spend). When reporting business health to investors/board. Quarterly for subscription businesses.

How

1. Define the revenue metric: net revenue (after refunds, chargebacks, payment processing fees), gross or contribution margin —be explicit, because LTV is meaningless without knowing which revenue it is
2. Choose the time horizon: 12-month LTV (most common for SaaS, comparable across cohorts), 24-month, or lifetime (requires modeling churn to infinity —risky for young products)
3. Compute historical LTV per cohort:
   - For each signup cohort (e.g., users who signed up in Aug 2026):
   - Sum the net revenue from that cohort over the horizon (12 months)
   - Divide by the number of users in the cohort (including churned users —they count in the denominator)
   - This gives the average LTV per user for that cohort
4. Track the LTV curve over time: month 1 LTV, month 2 LTV, ..., month 12 LTV —the curve shows when LTV stabilizes and lets you compare cohorts at the same age
5. Compute the LTV distribution, not just the mean: median, P25, P75, P90 —the mean is skewed by the top 20%, and the median tells you what a typical user is worth
6. Segment LTV by dimension: by plan (Pro vs Free), by source (organic vs paid), by industry —the segments with 3x LTV deserve 3x acquisition spend
7. For a predictive LTV: model month-12 LTV from month-1 behavior (features: actions taken, sessions, plan, engagement) using a regression or survival model —validate against historical cohorts before using
8. Report LTV:CAC ratio: target 3:1 (healthy), 1:1 (break-even), < 1:1 (destroying value); also report payback period (months to recover CAC)
9. Re-compute quarterly: as pricing, retention, and mix change, LTV changes, and a stale LTV leads to stale budget

Example

# LTV calculation: Aug 2026 cohort, 12-month horizon
Cohort: 1,240 users signed up in Aug 2026
Plan mix: 78% Free, 15% Pro ($20/mo), 7% Team ($50/mo)

Net revenue by month (cohort total):
  M1:  $4,650   (148 Pro + 17 Team upgrades in month 1)
  M2:  $5,180   (retention 92%, 4 new upgrades)
  M3:  $5,020   (retention 89%, 2 churns offset by 3 upgrades)
  M4:  $4,890   (retention 87%)
  M5:  $4,720
  M6:  $4,610
  M7:  $4,540
  M8:  $4,480
  M9:  $4,420
  M10: $4,380
  M11: $4,340
  M12: $4,310
  Sum M1-M12: $55,520

Average 12-month LTV = $55,520 / 1,240 = $44.77 per user

LTV distribution (per user):
  Mean:   $44.77
  Median: $ 0.00   (78% are Free and never pay)
  P75:    $ 0.00
  P90:    $240.00  (Pro users who retained 12 months)
  P95:    $600.00  (Team users)

Segment LTV:
  Free -> Pro upgrade:  $312  (15% of cohort, healthy retention)
  Free -> Team upgrade:  $648  (7% of cohort)
  Free (never upgraded): $0    (78% of cohort)

LTV:CAC ratio:
  Blended CAC: $22  (paid + organic)
  LTV:CAC = $44.77 / $22 = 2.03:1  (below 3:1 target)
  Payback: 5.1 months

Insight: blended LTV is 2:1, but paid-only LTV is $31 (paid users
churn faster) and paid CAC is $45 -> 0.69:1 (destroying value).
Action: reduce paid spend on low-intent channels, shift to SEO
(organic LTV $58, CAC $8 -> 7.25:1).

Common Problems

  • ⚠Using mean LTV when the distribution is heavily skewed —78% of users are free (LTV $0) and 7% are Team (LTV $648); the mean $44.77 describes neither group, and budgeting on the mean over-funds channels that acquire free users.
  • ⚠Computing LTV without including churned users in the denominator —dividing revenue only by active users inflates LTV by the churn rate (e.g., 40% churn inflates LTV by 1.67x) and makes the business look healthier than it is.
  • ⚠Using lifetime LTV for a young product —a 2-year-old product has not seen the full churn curve, and lifetime LTV extrapolated from 24 months of data is a guess that can be 2x off; use 12-month LTV until the churn curve stabilizes.
  • ⚠Not segmenting LTV by acquisition source —blended LTV is $44.77 but paid LTV is $31 and organic LTV is $58; allocating budget on blended LTV over-funds paid (which has negative LTV:CAC) and under-funds organic.
  • ⚠Treating historical LTV as predictive —historical LTV is a lagging indicator; if retention dropped last month, the 12-month LTV will not reflect it for 11 more months, and the team keeps acquiring users at a CAC that is no longer profitable.

Best Practices

  • ✓Compute LTV per cohort (by signup month) and track the curve over time —the curve shows when LTV stabilizes, lets you compare cohorts at the same age, and detects when a recent cohort is tracking below prior cohorts before the full horizon elapses.
  • ✓Report the LTV distribution (median, P75, P90) alongside the mean —the mean is skewed by the top 20% and the median reveals what a typical user is worth, which prevents over-budgeting acquisition on channels that bring free users.
  • ✓Segment LTV by acquisition source and plan —the 2x LTV difference between organic and paid means the two channels deserve different CAC ceilings, and budgeting on blended LTV misallocates spend.
  • ✓Use 12-month LTV for products under 3 years old and lifetime LTV only when the churn curve has stabilized —extrapolating lifetime from partial data introduces error that compounds into budget decisions.
  • ✓Build a predictive LTV model from month-1 behavior for fast-changing products —the model gives a month-1 LTV estimate that is actionable immediately, rather than waiting 12 months to learn whether a cohort is healthy.