Unit Economics Model Template
Settles what counts as acquisition cost before a number exists, then recovers each cohort's own cost out of its own contribution.
Free download · No account needed
Definitions Note
[Company] — one quarter of spend, seven defensible answers
There is no accounting standard behind customer acquisition cost and no convention narrow enough to settle an argument. Sign the line before anyone computes anything.
| Definition | Numerator | Denominator | CAC |
|---|---|---|---|
| Incremental cost of obtaining a contract | New-business commission | New logos | — |
| Spend intensity against the base | Sales and marketing | Active customers | — |
| What the marketing dashboard reports | Paid media | New logos | — |
| Counting an upsell as a win | Agreed acquisition | Logos plus expansions | — |
| The agreed line | Agreed acquisition | New logos | — |
| Income statement line, undivided | Sales and marketing | New logos | — |
| Everything it took to land them | Fully loaded | New logos | — |
What the note has to settle
| Renewal commissions | In or out. Out charges retention to retention |
| Brand and content | Split by who reads it, or excluded entirely |
| Onboarding and trial cost | Out, if they already sit inside gross margin |
| Lifetime value window | A stated month count, never a perpetuity |
| Cohort rule | Quarter of first paid subscription, never re-based |
Fill it from your own ledger, or open the pack and let the agent classify every spend line and show you the spread before you pick.
Two people at the same company can compute customer acquisition cost from the same books and land thirteen times apart, and neither of them is wrong. Pellmore Analytics spent 3,771,000 landing 61 new customers in one quarter. Commission only over new logos gives 4,672. Paid media over new logos gives 13,770. The whole sales and marketing line over new logos gives 57,689. Fully loaded gives 61,820. Four of those have appeared on somebody's slide.
So this pack settles the definition first and writes it down. Every spend line goes into acquisition, existing customers, or already inside cost of revenue, with a reason on each row. The note bridges the agreed numerator back to the income statement, fixes the lifetime value window at a stated month count rather than a perpetuity, and gets signed before a single ratio is computed. Every sheet afterwards names the line it runs on, so a series that shifted because the denominator changed never reads as a series that shifted because the business did.
The second move is refusing to blend. Pellmore's blended 2026 payback is 21.7 months, which is real arithmetic on two unrelated populations: this year's acquisition cost, recovered out of a base mostly acquired in earlier years at a third of the price. Run it by acquisition quarter and payback goes 16, 16, 18, 20, 24, 27, 32, 37. The newest cohort has not recovered its cost inside three years, and the 13-week cash forecast pack is where that shows up as cash.
What's in the pack
Definitions Note
Signed before the model exists. Carries the seven-way table, every excluded line with its reason, the bridge back to the sales and marketing line on the income statement, the lifetime value window and the cohort rule, plus a dated change log.
Unit Model sheet
Every spend line in the quarter classified into acquisition, existing customers, or already inside cost of revenue, with the reason on the row and the three candidate denominators underneath. This is the sheet the argument actually happens on.
Acquisition Cost and Lifetime Value sheet
All seven definitions computed side by side on the same quarter, cheapest to dearest, with the multiple of the narrowest and one line on what each is actually measuring. The spread sits on its own row.
Cohort Payback sheet
One row per acquisition quarter with its own CAC, its own monthly net revenue retention, cumulative contribution at twelve, twenty-four and thirty-six months, and the payback month. The blended figure sits beside them as its own row, next to the recurring revenue it ties to in the ARR pack.
Sensitivity sheet
The newest cohort stressed on acquisition cost, opening price, gross margin and retention separately, then price and retention together. Retention cases use a level an earlier cohort actually reached, with months saved on every row.
Unit Economics Memo
Written for a board pack or a financial data room. Leads with the trend, shows every cohort rather than the best one, splits the movement into cost and retention, and hands off to the investor update financials once it is safe to show an outsider. It drops straight into the Unit Metrics sheet of the board financial package, which is where a metric has to arrive with its definition and the query behind it.
The rule set
Eleven rules the space applies without being asked, from nothing computed before the note is signed to publishing the amounts under every ratio, which is what an outside reader asks for first.
Five agent prompts
Fix the Definitions, Build the Cohorts, Run the Sensitivity, Test a Price Change and Write the Memo, each wired to the sheets so a new quarter of billing data rebuilds the cohorts without rebuilding the model.
A price change model
Separates new business from the installed base, phases the rise across the actual renewal calendar, and gives the break-even win-rate penalty, because a price rise that costs conversion raises cost per logo at the same time.
How to use it
- 1
Open in River, or download it
Open the pack in River and let the agent classify your own spend lines and build the cohorts from your billing export, or download the blank Word documents and CSV sheets and fill them in. Both routes are free.
- 2
Sign the definition first
Compute the spread before anyone quotes a number. Seeing the same quarter produce a thirteen times range is what makes the definition conversation take ten minutes instead of recurring at every board meeting.
- 3
Build cohorts from billing records
Group by the quarter of first paid subscription and let each cohort recover its own cost out of its own contribution. Keep the customers who churned in the row, because dropping them describes only the survivors.
- 4
Stress the newest cohort, not the average
Run each lever separately against the cohort you are acquiring right now, using retention levels an earlier cohort actually reached, then say who owns the lever that wins.
Frequently asked questions
Is this template free?
Yes, all of it. Word documents and CSV sheets, no signup and no card, the same as everything in the template library. Edit with AI is a separate optional route that opens the pack in River and builds it from your billing and spend exports.
Why does the definition matter so much?
Because there is no standard behind it. In the worked example one quarter of spend divides seven defensible ways into acquisition costs from 4,672 to 61,820, a thirteen times spread on identical books. Every disagreement about whether the economics work is really a disagreement about which of those is on the slide.
Is this a real question outside the boardroom?
It is a question the SEC staff has put to filers in writing. Warby Parker was asked in 2021 whether its definition was commonly understood to measure costs against active customers rather than newly acquired ones, and whether the term could confuse investors. The comment letter is public.
What is wrong with a blended payback?
It divides this year's acquisition cost by this year's new customers, then recovers it from the gross profit of a base mostly acquired in earlier years at lower cost and higher retention. The worked example blends to 21.7 months while its newest cohort takes 37, and the gap grows every quarter the trend continues.
Do I have to publish the amounts, not just the ratio?
An outside reader will ask, and a regulator already has. The SEC staff told Chewy in 2019 to provide the CAC and LTV amounts per customer per period rather than the ratios alone, and questioned an LTV built on cumulative contribution profit. That letter sets the bar this pack writes to.
Can I report one good cohort?
You can, and the reader will assume the ones you left off are worse. The SEC staff asked Clear Secure in 2021 to extend a single-year cohort disclosure to earlier years where a material trend existed, per its comment letter. The pack reports every cohort by default.
What data do I actually need?
A billing export with the first paid date on every customer, sales and marketing spend by month at line-item detail rather than a single summary figure, and cost of revenue by month. The line detail is what lets the definition separate acquisition spend from spend that serves existing customers.
Settle the number before the meeting
Download the blank pack as Word and CSV files, or open this exact pack in River and let the agent show you how far apart your own definitions are.
Edit with AI