Why it matters
Ad platforms optimize on short-window events. Your finance team cares about what customers are worth over months or years. When those two views diverge, campaigns can look efficient on platform ROAS while the business acquires one-and-done buyers, high-refund segments, or trial users who never renew.
LTV is the language that connects marketing spend to business outcomes. A Head of Performance might hit CPA targets on first purchase while cohort LTV curves show the channel bought customers who never repurchase. A VP Growth needs LTV to justify scaling spend. Finance uses LTV and payback to decide whether acquisition is accretive.
The measurement challenge is timing. True LTV is often unknown at conversion time. Repeat purchases, subscription renewals, returns, and upsells reveal value weeks or months later. That gap is why teams pair retrospective LTV reporting with early predicted value (pLTV) for platform activation.
Customer lifetime value (LTV)
LTV defines the outcome you want platforms to learn toward. pLTV activation estimates that outcome early and sends it as a value signal:
- Inputs: Revenue, refund, subscription, and behavioral history from your data warehouse and first-party data.
- Measurement layer: Cohort LTV curves, unit economics, and maturity windows show realized value after customers age.
- Model layer: User-level pLTV predicts expected LTV before outcomes fully mature.
- Activation: Churney sends predicted values directly to ad networks so value-based bidding optimizes on lifetime economics, not first-order proxies.
- Readout: Holdout or BAU tests compare incremental ROAS and customer quality against the LTV definition you agreed upfront.
Reporting LTV alone does not change live bidding. Connecting LTV definitions to platform-ready signals is the activation step.
Simplified revenue LTV (per customer):
LTV = Sum of net revenue from customer over horizon HMargin-based LTV:
LTV = Sum of (revenue − variable costs − allocated refunds) over horizon HInterpretation guardrails:
State horizon H (D90, D180, 12 months, etc.) whenever you cite LTV.
Separate acquisition cohort LTV from blended active-customer metrics.
Compare LTV to CAC or payback period for spend decisions; platform ROAS alone uses a shorter attribution window.
Category variants
| Model | How LTV shows up |
|---|---|
| Ecommerce / DTC | Cumulative net revenue per customer after returns and discounts; repurchase rate and AOV expansion drive curve shape. |
| Subscription app | Subscription LTV from trial-to-paid, renewal rate, and early churn; often modeled over 6–12 month horizons. |
| SaaS / PLG | Expansion and retention revenue per account; sales-assisted deals may need longer maturity than self-serve signup. |
Common mistakes
- Using gross revenue without refunds or discounts. Inflates LTV and misaligns with margin.
- Mixing new and repeat customers in one LTV number. Acquisition LTV and blended customer LTV answer different questions.
- Judging channels before cohort maturity. Reading LTV at D7 when repeat peaks at D60 leads to wrong budget cuts.
- Equating LTV dashboards with platform optimization. BI reports do not send value events to Meta, Google, or TikTok.
- Ignoring contribution margin. Revenue LTV without COGS or variable costs overstates what acquisition can afford.
- Changing LTV definition mid-experiment. Shifting from revenue to margin mid-pilot confounds readout.
Advertiser lens
| Role | What they ask | What good looks like |
|---|---|---|
| Head of Performance / UA | Which channels deliver high-LTV customers? | Cohort LTV by source with agreed maturity window and sample size checks. |
| VP Growth / CMO | Can we scale spend and stay profitable? | LTV:CAC or payback framing tied to margin, not revenue alone. |
| Marketing Analytics / Data Science | Is our LTV model stable? | Clear grain (user vs account), leakage checks, and refresh cadence documented. |
| Data Engineering | Can we trust the revenue feed? | Append-only event history, refund timing, and ID resolution in the data warehouse. |
| Finance / Procurement | What LTV threshold unlocks budget? | Pre-agreed definition, horizon, and success criteria for pilot approval. |
FAQ
What is customer lifetime value (LTV)?
LTV is the total value a customer generates over their relationship with your brand, typically measured as cumulative revenue or margin after discounts, refunds, and churn over an agreed time horizon.
How is LTV different from predicted lifetime value (pLTV)?
LTV usually describes realized or cohort-level value after customers mature. pLTV is an early per-user prediction designed to influence ad platform optimization before full LTV is observable.
What time horizon should LTV use?
It depends on your business model. Ecommerce often uses 6–12 month horizons; subscription and SaaS may extend further. The horizon should match when retention and expansion patterns stabilize for your category.
Should LTV be revenue or margin?
Finance decisions should use contribution margin when possible. Revenue LTV is common in marketing reporting but can overstate what acquisition can afford if COGS and fulfillment vary by customer.
Does improving LTV reporting improve ad performance?
Not automatically. LTV reporting helps you diagnose acquisition quality. Platform performance improves when calibrated value signals (often pLTV) reach ad networks with strong match rates and freshness.
How do refunds and returns affect LTV?
They reduce realized LTV. Models and platform signals should account for refund timing; otherwise first-order value overstates long-term worth.
What data do you need to measure LTV reliably?
Consistent user IDs, complete purchase and refund history, subscription events if applicable, and daily append-only updates in your data warehouse. See Churney's data guide.
Not the same as
| Term | Difference |
|---|---|
| Predicted lifetime value (pLTV) | pLTV forecasts LTV early for platform signals; LTV measures realized outcomes. |
| Cohort LTV | Cohort LTV is an aggregate curve by acquisition group; customer LTV can be per-user or averaged within a cohort. |
| Average order value (AOV) | AOV is one order; LTV sums value across the full relationship. |
| Platform ROAS | ROAS uses attributed conversion value in a short window, not full customer LTV. |
