Why it matters
Platform dashboards show attributed ROAS inside a short optimization window. Cohort LTV shows what actually happened to customers after the click. A Meta campaign can look strong on D7 purchase ROAS while the January cohort never repurchases and the February cohort expands.
Cohort LTV is how analytics and finance answer questions performance marketing cannot see in real time: Did this channel buy one-and-done buyers? Do trial starts from TikTok renew at the same rate as organic? When does refund pain show up relative to first order?
The failure mode is reading cohorts too early. Cutting budget at D14 because LTV looks flat misses repeat revenue that arrives at D45. Agreeing on cohort maturity before judging channels prevents false negatives and false scale decisions.
Cohort LTV
Cohort LTV is the readout layer that validates whether early signals match long-term economics:
- Inputs: User-level revenue, refunds, and subscription events from your data warehouse, grouped by acquisition date and source.
- Cohort curves: Track average or median LTV by days-since-acquisition (D0, D7, D30, D90, etc.).
- Model calibration: Compare user-level pLTV scores at anchor events to realized cohort LTV at maturity.
- Activation: When pLTV is calibrated, Churney sends predicted values directly to ad networks so platforms learn toward the same LTV definition you track in cohort reports.
- Experiment readout: Holdout or BAU pilots use cohort LTV by test cell to judge incremental customer quality, not just platform metrics.
Cohort LTV tells you whether acquisition was good. pLTV activation tries to make the next cohort better before maturity.
Average cohort LTV at day D:
Cohort LTV(D) = Total net revenue from cohort through day D / Number of customers in cohortCumulative curve: Plot Cohort LTV(D) for D = 0, 7, 14, 30, 60, 90, …
Interpretation guardrails:
Use the same net revenue definition as finance (refunds, discounts).
Report cohort size alongside the curve.
For paid channels, align cohort grain to how you attribute acquisition in your data warehouse.
Category variants
| Model | How cohort LTV shows up |
|---|---|
| Ecommerce / DTC | Curves by first-order month and channel; repurchase and refund slopes separate strong from weak acquisition. |
| Subscription app | Curves by install or trial-start cohort; trial-to-paid and renewal steps visible before full subscription LTV. |
| SaaS / PLG | Curves by signup month and product tier; expansion revenue may lag activation by quarters. |
Common mistakes
- Comparing cohorts at different ages. A 90-day-old cohort vs a 14-day-old cohort is not a fair readout.
- Small sample sizes. Thin cohorts produce noisy curves that drive wrong budget moves.
- Treating cohort LTV as a platform signal. Cohort charts are analytics; platforms need per-user value events with identifiers and freshness.
- Wrong acquisition grain. Campaign-level cohorts need consistent UTM and attribution alignment to your first-party data source of truth.
Advertiser lens
| Role | What they ask | What good looks like |
|---|---|---|
| Head of Performance / UA | Which campaigns produce high-LTV cohorts? | Channel and campaign cohorts with agreed maturity gate before budget changes. |
| VP Growth / CMO | When can we trust the readout? | Documented cohort maturity window and sample-size thresholds. |
| Marketing Analytics / Data Science | Do pLTV scores match realized cohort LTV? | Calibration plots by decile and acquisition source at maturity. |
| Data Engineering | Can we rebuild cohorts reliably? | Stable user ID, daily append-only feeds, and attribution joins in the data warehouse. |
| Finance / Procurement | Does cohort LTV support scale? | LTV:CAC by cohort with margin definition and audit trail for pilot success. |
FAQ
What is cohort LTV?
Cohort LTV is lifetime value measured for a group of customers acquired in the same period, tracked over days or months since acquisition. It shows how average customer value accumulates over time for that group.
How is cohort LTV different from customer LTV?
Customer LTV can refer to a single user's total value or a general LTV definition. Cohort LTV always groups users by acquisition time (and often by channel or campaign) to compare acquisition quality.
How is cohort LTV different from pLTV?
Cohort LTV is retrospective: you observe outcomes after customers age. pLTV is a forward-looking score sent early to ad platforms. Teams use cohort LTV to validate whether pLTV predictions matched reality.
When is a cohort mature enough to evaluate?
When key behaviors (repeat, renewal, refunds) have mostly occurred for that business. Ecommerce often needs 60–90+ days; subscription and SaaS may need longer. Define the window before running channel comparisons.
Can cohort LTV replace incrementality testing?
No. Cohort LTV describes who you acquired and how they behaved. Holdout or structured experiments measure whether a change caused better outcomes vs BAU.
What breaks cohort LTV analysis?
Inconsistent IDs, attribution gaps, mix shifts, and reading immature cohorts. Data readiness and aligned attribution data are prerequisites.
How does Churney use cohort LTV?
Cohort LTV curves calibrate pLTV models and judge pilot readout. Activation still requires sending user-level values to ad networks, not exporting cohort charts to platforms.
Not the same as
| Term | Difference |
|---|---|
| Predicted lifetime value (pLTV) | pLTV is an early per-user prediction for platforms; cohort LTV is observed after acquisition. |
| Cohort-based LTV model | A modeling approach using segment history; cohort LTV is often a reporting output. |
| Platform ROAS | Short-window attributed value, not full cohort lifetime curves. |
| Retention rate | Retention counts active users; cohort LTV sums economic value including spend per retained user. |
