Channel CAC & LTV calculator

Showing the latest cohort per channel.

Case loss % = 100% − Onboard % − Refund %. Avg upfront cost feeds case-loss and onboard payout calculations only. Avg profit per done feeds LTV only — these are separate inputs so each can move independently. Fraud-adjusted LTV: effective dones = 1 + (LTV input − 1) × (1 − Fraud %), then × Avg profit per done.

Enter an ad-spend budget for the currently-selected channel. Working forward through the funnel rates, channel cost, fraud, and refund rate — here's what that budget produces. Total cash needed includes onboard payments (product cost) on top of the ad spend; profit/loss is gross LTV minus ad spend minus case loss cost.

Working backward from each LTV/CAC target ratio: given the current channel's funnel rates, profit, fraud, refund, and avg upfront cost, here's the maximum cost per acquired user at which you'd hit that ratio. The unit (signup / joined / done) matches how this channel is priced. Updates live as inputs change. Green means the current cost sits at or under the ceiling (you have headroom). Red means the current cost exceeds it (target is missed).

How LTV/CAC ratios and total revenue shift across different fraud rates, holding everything else constant. The shaded row is your current fraud %. The "Total 1yr LTV" column shows gross 1-year revenue across all dones at that fraud rate; "vs current" shows the dollar gain (green) or loss (red) compared to your current fraud rate. Below the table: per-done LTV recovered if fraud were eliminated entirely.

–

Funnel data sourced from Metabase exports. Use the channel preset and cohort year selectors above to filter the views: pick a channel to isolate trend charts and table rows; pick a year to filter strategy rankings and benchmarks. Refund %, fraud %, profit per done, and avg upfront cost are held constant — only funnel rates and channel costs are cohort-specific. Partial-year cohorts (Jan–March 2026) are flagged but otherwise treated like full-year data.

Showing latest cohort per channel.

All 17 cohorts. Latest cohort per channel is shaded.

Charts always span the full timeline regardless of year filter. Pick a channel above to isolate that channel across all six charts; pick "Custom" to show all channels together. The legend below also lets you toggle individual channel visibility.

Build a custom improvement scenario and see the combined impact on profit and LTV/CAC. Each lever has its own % improvement input — set them to model "what if we improve O→D by 15% AND CPL by 5%" combined plans. The scenario uses the channel + cohort year selected above as the baseline. Refund recovery converts would-be refund users into dones (case loss % stays constant). Profit improvement assumes profit per done can be lifted directly (e.g., better monetization).

Each input is the % improvement on its lever (positive numbers always mean better — CPL improvement means cost goes down, refund improvement means refund rate drops with retention). Defaults to 10%. Set to 0 to exclude a lever.

Applying ALL the improvements above simultaneously, here's what happens to the channel's economics at the test budget. Compare current state vs improved scenario.

If you applied this same set of lever improvements to every channel's latest cohort at the same test budget, here's the cross-channel total. Useful for evaluating whether a company-wide initiative (e.g., "improve onboarding completion 15% across the board") has enough aggregate leverage to justify the cost.

Every formula on this page resolves against the input values above. Edit any input and watch the math walk through live. The chain reads top-to-bottom: funnel volume feeds cost feeds CAC, then LTV with fraud adjustment, then the final ratios.

Every user does at least 1 done by definition. Of the users who reach a first done, the Fraud % gets banned and contributes nothing further. So the effective dones-per-user is 1 (guaranteed) plus the rest of the trajectory weighted by the non-fraud rate.

For each target ratio × LTV horizon cell on the Calculator tab, the same four-step solve runs. Below is the walkthrough for 1yr LTV at the 2.0× target — same logic applies to all 9 cells.