What is LTV?
LTV (Customer Lifetime Value) estimates the net revenue a user will generate over the entire relationship with your product or site. Analysts use it to judge if acquisition costs (e.g., CPA, CPC, CPL) are sustainable and to prioritize retention work. In web analytics, LTV connects marketing inputs—Campaign, UTM, traffic Source/Referral—to business outcomes, not just surface metrics like Pageview or Session.
How LTV is calculated (practical models)
There’s no single “correct” formula; you pick a model that matches your business mechanics and data quality.
| Context | Formula (simplified) | Notes |
|---|---|---|
| E-commerce | LTV = AOV × Purchase Frequency × Avg. Lifetime × Margin | Use your AOV and orders/user. Estimate lifetime from retention curves via Cohort Analysis. |
| Subscription | LTV = ARPU × Avg. Lifetime × Margin | If churn is stable, Avg. Lifetime ≈ 1 / churn rate. |
| Lead-gen | LTV = Lead-to-Sale Rate × Avg. Deal Size × Repeat Rate × Margin | Tie back to channel via UTMs and form events. |
Workflow in analytics tools
- GA4 / alternatives (Matomo, Plausible, Simple Analytics): Track revenue or goal values, user identifiers (privacy-safe), and acquisition dimensions. Build cohorts to estimate average lifetime and repeat frequency.
- Attribution: Align your LTV model with the chosen Attribution Model so channel ROI comparisons are fair.
- Engagement signals: Higher Engaged Sessions and longer Engagement Time often correlate with higher LTV, but validate with cohorts.
- Quality gates: Guard against inflated values by excluding refunds, taxes, and promo abuse; use net margin if possible.
Why LTV matters
- Budgeting: Compare LTV to acquisition cost to set profitable bidding/targets and to interpret ROI.
- Segmentation: Different segments (channel, geo, device) can have the same Conversion Rate but radically different LTV.
- Roadmapping: LTV surfaces the impact of retention features, lifecycle messaging, and product improvements—beyond vanity metrics.
Implementation tips
Start simple, document assumptions, and iterate. Validate your model quarterly: rerun Cohort reports, reconcile finance vs. analytics, and stress-test sensitivity (e.g., lifetime ±20%). The goal isn’t a perfect number—it’s a consistent yardstick for decisions.