The four numbers that run a DTC brand

A typical DTC dashboard stack shows forty numbers. Shopify shows revenue and conversion, Meta shows ROAS, Google shows a different ROAS, Klaviyo shows attributed revenue that overlaps both, and a spreadsheet somewhere tries to reconcile them. Most operators respond by checking everything daily and trusting nothing fully.

You need four numbers: blended CAC, MER, contribution margin, and payback. Everything else is either an input to one of these or a distraction from them. This essay defines each one precisely, shows the computation on a worked example, and covers the trap each number sets when it's read alone.

One deliberate omission up front: platform ROAS is not on the list. Meta's reported ROAS is a claim by an interested party, useful for comparing ads against each other inside Meta and dangerous for judging whether Meta is worth the money. Why the platform number diverges from your bank account is its own essay. Here we only need the conclusion: budget decisions run on the four numbers below, not on any platform's self-grade.

Every formula below uses the same brand

Every formula in this essay uses the same illustrative brand so the numbers connect:

Blended CAC: what a customer actually costs

Total marketing spend divided by new customers. All of it: ad spend, agency fees, tools, creator payments, creative production. And only new customers in the denominator, never all customers.

For the worked example: ($96,000 + $26,000) ÷ 2,100 = $58.

The question it answers: what are we really paying to acquire a customer, with no channel taking credit and no channel hiding? Platform CAC numbers argue with each other because each platform claims the same conversions. Blended CAC cannot argue. Dollars out, new customers in.

The trap: the denominator. Count all customers instead of new ones and a strong repeat business flatters acquisition, sometimes by 40% or more. Your repeat buyers were coming back anyway, mostly for free. Blend them in and you'll scale spend believing acquisition is cheaper than it is, which is precisely how brands grow revenue while losing money on every new cohort.

Compute it monthly. Weekly blended CAC is noise for most brands under $25M; monthly is signal.

MER: the whole engine in one ratio

Total revenue divided by total marketing spend. The worked example: $420,000 ÷ $122,000 = 3.4.

The question it answers: is the marketing engine as a whole earning its budget? MER is the honest weekly heartbeat. It ignores attribution entirely, which is its whole value: no platform can inflate it, no cookie loss can break it.

The trap: MER has no opinion about growth. Cut spend and MER rises, because the sales your brand generates without ads (repeat, organic, word of mouth) stay in the numerator. A rising MER can mean improving efficiency or a shrinking ambition, and the ratio won't tell you which. Read it next to spend and new-customer count, never alone.

The second trap is treating a target MER as a preference. Your minimum MER is arithmetic, not taste, and it falls out of the next number.

Contribution margin: the number that sets all the others

Revenue minus all variable costs, per order. Product cost, inbound freight, shipping, fulfillment, packaging, payment fees, and the expected cost of returns. Not ad spend (we're computing what's left over to pay for ads) and not fixed costs like salaries or rent.

The question it answers: how many dollars does each order generate to pay for marketing, fixed costs, and profit? At $40 per order, first-order break-even CAC is $40. Our worked brand pays $58 blended, so it loses $18 of contribution on a customer's first order and needs repeat purchases to climb out.

That isn't automatically bad. It's a decision, and most operators have never seen it stated as one.

Contribution margin also sets your minimum MER. Break-even MER = 1 ÷ contribution margin rate. The chart below is the whole relationship:

Our example brand runs a 3.4 MER against a 2.1 break-even, so the month as a whole is profitable even while first orders are underwater. Both statements are true at once, and you need both numbers to see it.

The trap: computing margin once a year. COGS moves, carriers reprice, returns creep. A brand that computed 47% in January and runs 41% by August is making every downstream decision with stale arithmetic.

Payback: the speed limit

How long a new customer takes to return their CAC in contribution margin. First order contributes $40 against a $58 CAC: $18 still owed. If the average new customer's repeat purchases contribute another $18 within about seventy days, payback is roughly ten weeks.

The question it answers: how fast does a dollar spent on ads come back as a dollar you can spend again? That's a cash question, and cash, not ROAS, is what actually caps growth for a bootstrapped or lightly funded brand. Two brands with identical LTV:CAC ratios can have completely different ceilings: the one that recovers CAC in six weeks can compound its budget monthly; the one that recovers in nine months is quietly financing its growth with inventory debt.

The computation without a data team: take a cohort of customers acquired in a single month, sum their contribution margin month by month, and mark where the cumulative line crosses that cohort's CAC. Shopify order exports and a spreadsheet are enough.

The trap: LTV faith. "Our 24-month LTV is 3x CAC" is a fine sentence for an investor deck and a dangerous one for an operating plan, because the 24-month customer value arrives over 24 months while the ad invoice arrives in 30 days. Payback is LTV with the wishful thinking removed: it only counts margin, and it only counts it when it lands.

The four constrain each other in a loop

The loop runs like this. Contribution margin sets what you can afford: break-even CAC per order and break-even MER for the blended month. Blended CAC tells you what you're actually paying. The gap between them is your first-order profit or your deliberate first-order loss. Payback tells you how fast that loss turns into cash, which sets how aggressively you can scale spend. MER watches the whole engine weekly and flags when something below it needs a closer look.

When the four disagree, the disagreement is the diagnosis. MER healthy but blended CAC rising: your repeat base is subsidizing worsening acquisition, and the problem is hiding in one of the five gates. Blended CAC flat but payback stretching: your new cohorts are repeating less, which is a product or offer question, not a marketing one. CAC fine, margin shrinking: the finance conversation you've been postponing.

Your business model sets the patience, not the numbers

The four numbers are universal. Their healthy ranges are not, and most benchmark charts skip this.

• One-time purchase brands (furniture, gifts, occasion products) have to clear break-even CAC on the first order, or very close to it. There's no repeat curve coming to the rescue, so payback is nearly instant or the model doesn't work. These brands live and die on contribution margin and message-matched conversion.
• Consumable and repeat brands (skincare, supplements, coffee) can run a first-order loss on purpose, but only as deep as their real repeat curve, measured from their own cohorts, not from a category average. The trap is borrowing a subscription brand's payback tolerance with a one-time brand's repeat rate.
• Subscription brands can pay the most for a customer and are also the easiest to fool: churn assumptions hide inside every payback number. A subscription cohort's payback claim is only as good as its month-six retention, so the four numbers pick up a fifth companion here (churn), and payback gets re-computed as cohorts age, not projected once at acquisition.

Same arithmetic everywhere. What changes is how much patience the model has earned, and the evidence standard for that patience is your own cohort table.

The weekly sheet

The operating version of this essay fits in one small spreadsheet, eight cells a week:

• Revenue, total marketing spend, MER (computed)
• New customers, blended CAC (computed)
• Contribution margin rate (updated monthly, carried weekly)
• Spend as % of revenue
• Current cohort payback estimate (updated monthly)

Fill it every Monday for a quarter and most dashboard anxiety disappears, because the four numbers move slowly and mean something. The forty numbers can go back to being inputs.

If you want help wiring this up against your own P&L, book a CAC audit: a free 30-minute review, no deck, no pitch. You leave with the top three things to fix first, whether or not we work together.