How to Calculate LTV:CAC Properly (By Cohort, on Margin, With Payback)

Why Blended LTV:CAC Is an Average of Stories That Never Happened

The most expensive number in e-commerce is the blended LTV:CAC ratio, because it exists to justify spending and is built to flatter. Blend every cohort and every channel into one figure and the April shopping-led customers subsidize the August prospecting push, the retention flywheel subsidizes the paid cold traffic, and the resulting average describes a customer who does not exist. Budgets get scaled on that fiction, and the reckoning arrives as a cash problem months later, when the individual stories finally add up.

The fix is not a better blended number; it is refusing to compute one. To calculate LTV:CAC in a way that governs spend, you need four properties: cohort rows instead of averages, margin instead of revenue, channel-level CAC, and payback in days next to the ratio. This article builds the calculation step by step, then wires it to the zones and the monthly cadence that turn a spreadsheet into scaling rules.

Step 1: Build the Cohort Table

Start from your order export: acquisition month, channel of first order, customer ID, first-order value, and every subsequent order with dates and values. Group customers by acquisition month, then compute per cohort: customers acquired, channel mix, CAC from that month's spend by channel divided by new customers, 12-month revenue per customer, and margin LTV at your gross margin percentage. If you want the mechanical work templated, the free diagnostic prompts include the cohort-table builder that runs this on your own exports.

The data traps to flag as you build

Three traps corrupt most first attempts. Orders without customer IDs silently shrink every cohort, so reconcile the customer count against your platform's own report before trusting anything. First-order channel tags still lie when payment rails scramble sources, so spot-check five rows per channel. And partial cohorts are the quiet killer: months younger than twelve need a projected column with the method shown, or the table quietly compares full years to half-years.

What a Working Cohort Table Looks Like

The table below is the shape you are building, one row per cohort, margin-drawn, with the blended row kept visible only to show what it hides. Numbers are illustrative but the pattern is the one real stores produce: the same store, the same year, and three different businesses depending on which row you are honest enough to read.

Step 2: Compute LTV on Margin, Not Revenue

Revenue LTV is the vanity version of the metric, because revenue does not pay payroll, contribution does. Multiply each cohort's 12-month revenue per customer by your gross margin percentage, and defend the margin figure itself: blended gross margin can hide shipping surcharges, payment fees and returns, so reconcile it with finance once and freeze the definition in the sheet. The ratio drawn in revenue is the margin-drawn ratio inflated by your markup, exactly the size of the lie most stores are scaling on.

The margin redraw also changes channel verdicts, not just magnitudes. Low-AOV channels that looked marginal on revenue become clearly underwater on margin, while high-AOV considered purchases improve, which is why the redraw so often reverses the store's channel ranking. Run the redraw before the zones exist, or you will build guardrails around flattering numbers and discover the error the week the cash gets tight.

Step 3: Compute CAC by Channel and Cohort

CAC belongs in the denominator of every cohort row, not in a separate dashboard: the month's spend on each channel divided by that month's new customers from that channel. Include the costs platforms let you forget, agency fees, creative production, tooling, because finance will, and a ratio that excludes them is a ratio nobody argued with in the budget meeting. The result is a CAC per cohort per channel that usually varies two- to three-fold within the same store, which is the entire point: variance is the decision.

Reconciliation is the discipline that makes CAC numbers defensible. Platforms claim more conversions than stores process, and the gap routinely runs 30-60%, so every month's spend should be matched against the store's own new-customer count before it enters the table. Where the gap is large, use the store's count and note the platform's claim, spend arguments lose to platform panels faster than to any competitor.

Step 4: Derive Your Minimum Defendable Line

The line is business arithmetic, not benchmark copying. Three inputs set it: your contribution needs, what ratio must cover your overhead allocation and leave target profit at your current volume; your payback ceiling, how many days of cash lock-up your balance sheet and inventory cycle tolerate; and your growth stage, a store running on founder cash carries a stricter line than one with a facility and a facility's credit. Write the derivation down with its date, because a line without a written why gets negotiated away the first quarter someone wants to scale past it.

The famous industry shorthand of 3:1 can start the conversation but not end it, because the right line varies with margin structure, repeat behavior and funding reality. A replenishable brand with 65% margins and subscription share can defend a line that would starve a one-and-done gadget store, and the derivation is how you prove it to finance. Once the line exists, it needs teeth, which is what the zones provide.

Step 5: Draw the Four Zones and Wire Decision Rights

The zones carry decision rights, which is what makes the ratio govern rather than decorate. Decide in advance who can move a channel between zones, what evidence the move requires, and how long any channel may sit in repair before the stop rule executes automatically. Guardrail systems fail socially before they fail mathematically, the zones exist so the argument happens in the spreadsheet and not in the hallway.

• Scale, comfortably above the line: the channel grows under written rules, typically up to 20% per month, with the ceiling recorded in advance
• Hold, above the line but not comfortably: maintain and optimize, re-read next cohort, no growth this month
• Repair, below the line: a named fix with an owner and a date, budget frozen until the fix ships
• Stop, well below the line: spend pauses within the week, no exceptions and no brand-awareness clause

Step 6: Compute Payback and the Cash It Locks

Payback converts the LTV:CAC discussion from accounting to operations. A channel at 1.8:1 with 60-day payback is a growth asset; the same ratio at 120-day payback is a working-capital problem wearing a healthy ratio. Stores that manage both numbers scale on cash they actually have, which is why the payback audit, not the ratio, is the number finance brings to the growth meeting. When payback is the constraint, retention work is acquisition work: the friction-surgery method and the flow map both shorten the clock without touching the ad account.

• Median days from first order to gross-margin break-even, per cohort, the cash lock-up clock for each acquisition source
• Cash currently locked: monthly acquisition spend multiplied by median payback divided by 30, the working capital your growth is holding hostage
• The scaling check: what a 20% spend increase does to locked cash at current payback, the constraint the ratio alone never shows
• The two levers that shorten payback fastest: better repeat behavior from flows and timing, and cheaper acquisition from structure and channel mix

The Retention Economics Behind Every Ratio

The numerator of LTV is a retention outcome wearing a marketing costume, and the published numbers make that concrete. Marketing Metrics' widely cited odds put the chance of selling to an existing customer at up to 70%, against 20% at best for a new prospect, and BIA/Kelsey's research found repeat customers spend 67% more per order than first-time buyers. Every point of repeat behavior compounds directly into the cohort rows you built in Step 1, which is why the cohort table and the retention program are one system, not two dashboards.

When the ratio disappoints, check the numerator before the denominator: the fix is often the flow map, not the bid cap. The cheapest margin improvement available to most stores is the 90-day repeat rate, it responds to post-purchase flows, timing, and the product's consumption cycle, none of which live in the ad account. That is also why the zone rules above read the numerator first: a channel's ratio can heal from the top down.

The Profit and Cost Forces Around Your Line

Two external forces keep moving the goalposts while you manage the ratio. Acquisition costs have inflated sharply over the half-decade, ProfitWell's analysis put the five-year rise at roughly 60%, and most operators still allocate more attention to acquisition than retention: Invesp found 44% of companies prioritize acquisition against 18% on retention, the exact inversion the retention economics above punish. Bain's long-cited finding completes the case: a 5% retention gain lifts profits by at least 25%. The figures below are worth pinning next to the derivation.

• Re-derive the line annually, and date the derivation, a stale line is a subsidy to whoever wants to spend past it
• The CAC inflation trend is the argument for payback discipline: inflation does not have to beat you if your clock is short

Step 7: Wire the Model to Monthly Decisions

The model governs only if it has a cadence, and the cadence is deliberately boring. Refresh the cohort table with the newest complete month, redraw the zone chart, read every channel's zone out loud in a twenty-minute monthly review with finance in the room, and log the decisions, scale moves, repairs with dates, stops executed. The shared-number design is the point: marketing and finance reading the same margin-drawn table is how the spend debate ends, because both sides are now arguing with the cohort rows instead of with each other.

The monthly review agenda

Twenty minutes, five beats: the refreshed table on screen, each channel's zone read aloud with its numbers, every repair checked against its promised date, the locked-cash figure stated once, and the decisions logged with owners. No new dashboards are allowed in the meeting, the cohort table is the only source of truth, which is precisely what makes it govern. If a channel's zone reading starts an argument, the zone rules from Step 5 end it.

The written rules that stop the negotiation

Three sentences in the company wiki do most of the governance work: the line and its derivation date, the zone transitions and who owns them, and the stop rule with its deadline. When someone wants to spend past the line, the conversation is with the document, not with the analyst, and the document wins because it was written on a calm day. If you want the full system with the derivation worksheet, the zone rules and the guardrail chart template, it is Day 11 of the E-Commerce Growth Engine program, built on your own store's exports.

Your 90-Day Implementation Checklist

Here is the quarter that turns this method into the store's operating habit. Days 1-30: build the cohort table from clean exports, run the margin redraw, and reconcile CAC against the store's own new-customer counts. Days 31-60: derive and sign the line, draw the zones, and run the first monthly review with finance in the room. Days 61-90: execute one repair and one scale decision by the written rules, and put payback on the finance one-pager. The free e-commerce course day walks the first artifacts, and the AI tools shortlist for e-commerce covers the automation layer if the exports fight back.

For teams that run the model in planning workshops, the ready-made e-commerce case studies books and coaching toolkits carry the cohort template and the zone rules into the room without screens. And the closing discipline is the one this article opened with: never compute a blended ratio again, the cohorts are the only rows that ever paid for the ads.

Frequently asked questions