The numbers, computed the honest way.
CAC and LTV are usually quoted in ways that flatter, LTV on revenue that ignores refunds and margin, CAC blended so one channel hides another, retention averaged into noise. Decifer computes them honestly, shows the assumptions, and says plainly when the data can’t settle a question.
A flattering number is worse than no number.
The metrics everyone repeats are often built to look good. LTV quoted on revenue ignores the refunds and margin that decide whether a customer was actually profitable. A blended CAC averages a cheap channel and an expensive one into a number that looks fine while one subsidises the other.
Decisions made on flattering numbers are confidently wrong. Decifer’s job here is the opposite of a prettier dashboard: compute the figures the honest way, show what they rest on, and be candid about what the data can and cannot tell you.
What we won't do
- Quote LTV on revenue while ignoring refunds and margin.
- Hand you one blended CAC that hides which channel is expensive.
- Average every cohort into a single retention rate that buries the detail.
What the module computes.
The core measurement layer, calculated from your own data, with its limits stated, not hidden.
Cohort retention
Repeat behaviour over time
Customers grouped by the month and source they arrived, then tracked as repeat-rate curves. Averaging everyone into a single retention number hides which cohorts actually come back; cohorts show it.
What we measure
How each intake of customers repeats, and which sources bring the loyal ones.
LTV on contribution margin
Net of refunds, with assumptions shown
Lifetime value built on contribution margin and net of your refund rate, not inflated revenue. The assumptions behind every LTV figure are shown, because an LTV you can't inspect is a number you can't trust.
What we measure
What a customer is really worth after refunds and cost, and on what assumptions.
CPA by channel
Not a blended average
Cost per acquisition worked out per channel from each platform's own reported spend and conversions, not one blended number where a cheap channel quietly subsidises an expensive one. It is a true CPA, the denominator is the platform's attributed purchases, so it is labelled as such, not dressed up as new-customer CAC.
What we measure
Which channels clear your cost bar, and which are dragging the blend up.
How it works, step by step.
The measurement an analyst should do, and the honesty about uncertainty most reporting skips.
Assemble the data
Orders, refunds, spend and sessions, read-only from Google Analytics or entered by hand, into your isolated workspace.
Build cohorts and retention curves
Customers are grouped by acquisition month and source, then their repeat behaviour is tracked over time as real curves, not a single blended rate.
Compute LTV on margin, net of refunds
Lifetime value is calculated on contribution margin after refunds, with every assumption stated on the figure so you can challenge it.
Work out CPA per channel
Cost per acquisition is computed per channel from each platform's own spend and attributed conversions, instead of blended into one flattering average.
Issue the decision
What the honest numbers say to do, and, where the data genuinely can't settle a question, a plain statement that it can't, rather than a confident guess.
One decision, reasoned end to end.
A measured signal, what it really means, and the move. The numbers below are illustrative.
Margin-true LTV reranks your channels.
Signal
On revenue, Channel A’s customers look like your best. Net of a 12% refund rate and on contribution margin, their LTV falls 22%, and the channel drops from #1 to #3. A cohort that looked healthy was propped up by refunded orders.
What changed
Nothing broke, revenue-based LTV flattered the channel all along. Refunds and thin margin on its orders mean it earns far less than the headline suggests, so ranking on revenue would keep funding it ahead of channels that actually pay back.
The move
- Rank channels on margin-true LTV, net of refunds, not on revenue.
- Watch this cohort's refund rate; it is the cause, not the symptom.
- Shift acquisition budget toward the channels that clear your margin bar.
Computing LTV on margin instead of revenue is the difference between funding a channel that looks profitable and one that is.
What the module reads.
Read-only from Google Analytics and Shopify, or enter any source by hand, both work today. Every figure is computed, never invented.
Forecasting is live: explainable projections, trend, seasonal and pace-to-target, each with a prediction interval that widens with the horizon, and an honest “not enough history to forecast” when the data is too thin. No black box; every figure traces to a formula you can read.
See your numbers, honestly.
Connect your data, and the cohorts and margin-true LTV appear the same day, assumptions and all.