How do you build a first-time customer P&L?
Strip repeat revenue out entirely, then take new-customer gross revenue down through discounts, returns, COGS, and operational cost. What is left is the real budget you have to acquire a customer. Compare it against ad spend and you know whether growth is paying for itself.
- $2,420 · ProfitIt's what month one earns. At $60K of spend in month three, it's negative.
- $50K · Ad spendMonth one spends $50K against $52.4K of contribution margin.
- About $21K · OperationsPick and pack, payment processing and customer service all scale with orders.
- $42.4K · COGSCOGS runs about 36.6% of net revenue, which leaves $73.6K of gross margin.
- $84K · Discounts and returns30% off sitewide and a 12% return rate give away 42 cents of every gross dollar.
Blended numbers mix two businesses.
Most DTC brands spending $50K+ a month on Meta judge scaling decisions on blended MER or blended ROAS. Those numbers mix two completely different businesses, acquiring strangers and selling again to people who already trust you.
Peel back a healthy-looking blended number and you can find that 60% of the attributed revenue came from customers who've already bought three or more times. Most of that revenue would have arrived anyway. It's propping up new-customer acquisition that may not be paying for itself, and the blended number can't tell you which is happening.
Before you build anything, check what share of total revenue comes from first-time customers. If it's only 25 to 30%, the business depends heavily on repeat buyers, and every blended metric you look at is mostly measuring your existing customers rather than your growth.
Build it top down, on new customers only.
Each line answers a question the line above it hides.
- 01Gross revenue from first-time customersLeave total revenue out of it. If you can't segment this, that's the first thing to fix.
- 02Net revenueSubtract discounts and returns. Most of the damage happens here, and almost nobody looks.
- 03Gross marginWhat's left after COGS and shipping on those orders.
- 04Contribution marginNext come the operational costs that scale with orders, like pick and pack, payment processing and customer service.
- 05Acquisition contribution marginTake off ad spend and you have the number. It tells you what one month of acquiring customers earned or cost you.
Month one looks fine, and month three goes negative.
Take a brand where first-time customers produce about 40% of total revenue, which is a reasonably healthy split.
Start with $200K of gross revenue from new customers. Run 30% off sitewide to bring them in, carry a 12% return rate, and you've given away 42 cents of every gross dollar before COGS has been touched. That $200K is now $116K of net revenue, a 58% net revenue margin.
COGS runs about 36.6% of net revenue, which leaves a 63.4% gross margin, or $73.6K. Take off roughly $21K in operational costs and contribution margin is $52.4K, a 45.2% margin. That $52.4K is the entire budget available to acquire these customers.
Month one spends $50K on ads against that $52.4K. Profit is $2,420, an acquisition contribution margin of 2.1%. It's positive, and it feels fine.
Month three scales ad spend to $60K while margins hold steady, and the brand goes to negative $2,338. By month six it's negative $9,475. Nothing broke, and efficiency didn't collapse. Spend grew past the contribution margin that was funding it, and blended MER never showed it because repeat revenue kept climbing at the same time.
Scaling becomes a question of contribution margin.
Once this exists, scaling stops being a question about ROAS. It becomes a question about how much contribution margin one month of acquisition produces, and whether the ad spend fits inside it.
It also makes the dependency visible. In the same example, with 8% monthly churn eating into the returning base, turning Meta off produces a 44 to 46% revenue drop. Know that number before somebody else finds it for you.
When it doesn't hold.
There are five cases where this framework misleads.
- 01It's a single-period viewA brand with strong retention can rationally run a negative first-order P&L, which is what a cohort payback model is for. Use both.
- 02It needs a clean splitYou have to separate first-time from repeat orders cleanly. Guest checkout, multiple emails and marketplace orders all blur that line, and a bad split produces a confident wrong answer.
- 03The percentages are one brand's shapeDiscount depth, return rate and COGS vary enormously by category. Run your own numbers through the structure instead of adopting these.
- 04Subscriptions and consumables break the framingThe first order is deliberately sold at a loss against a contracted stream, rather than a hoped-for repeat.
- 05It says nothing about incrementalitySome of those first-time customers would have found you anyway, and this model counts them as acquired.
Run your own numbers through the five lines.
Free. It opens as your own copy in Google Sheets.
Get the weekly email.
One tactic a week, to 3,475 operators.