He made $16,000 in three weeks and kept $500

A store owner came to us three weeks after turning on Google Ads. His dashboard looked like a win: about a hundred orders and almost $16,000 in sales on roughly $2,700 in ad spend. Most new advertisers would have screenshotted it and called the campaign a success. He asked a question instead: "At what point does advertising start consuming the business?"

We did the math with him on the call: a 20 percent margin on $16,000 in sales is about $3,200 in gross profit, and ads had taken $2,700 of it. He kept about $500. We gave him the only answer the arithmetic allowed: he works for Google now. The W-2 is in the mail.

Per order, the picture sharpens. His average order was about $160 and carried roughly $32 in gross profit, and winning it cost about $27 in ads. He cleared five dollars an order. Three weeks in, that can be fine, and it can even be the plan, but only when two things are true.

Ad spend that buys a repeat customer is an acquisition cost. Ad spend that buys one order at break-even is a donation. A store whose buyers reorder every quarter can afford to break even on the first sale, because the second and third orders arrive without an ad bill attached.

Three weeks of history cannot prove a repeat rate, but you do not have to wait a year to see one forming. Group customers by the month of their first order, then watch what each group spends in the months that follow. If the January buyers are still buying in April, the first order was a down payment. If every cohort goes quiet after one purchase, the store is renting its revenue from Google one click at a time.

At 20 percent there is no cushion, and the next click-price increase comes straight out of his $500. The two levers are pricing and order size, and small moves on either one change the whole equation.

Suppose a bundle or a free-shipping threshold lifts the average order from $160 to $190. At the same $27 cost per order, his keep goes from five dollars to eleven. A thirty-dollar bigger cart, and the ad bill never moves. Pricing works the same way: a modest increase flows almost entirely to the profit line if conversion holds, and conversion is a thing you test with a hundred visitors rather than debate in a meeting.

Both conditions roll up into two numbers we told him we needed before touching the budget: what a customer is worth over twelve months, and what he would pay to get one.

The first number is measured. Customer lifetime value sounds like consultant vocabulary, but it is a division problem waiting on data the store already collects: take a cohort of buyers, add up what they spend across a year, multiply by the margin. The second number is decided, not discovered. Once you know a customer is worth, say, $120 over a year, you choose how much of that you will pay to acquire the next one, and that choice becomes the ceiling on every bid in the account.

We advised him to hold the budget where it was until both numbers existed. Until then, scaling the spend is a guess, and so is shutting it off.

This is the work he brought us in to do, and it is less glamorous than the ads themselves. Pull the order data and measure the cohorts. Set the twelve-month customer value. Choose the target cost per customer and cap the bidding with it. Then spend moves toward the campaigns and products whose buyers come back, the single break-even orders get cut, and the margin work runs alongside: price tests, bundles, an order-size target the ads can bid against. When the account is rebuilt around those two numbers, the owner stops asking whether advertising is consuming the business, because the dashboard answers it every morning.

If yours looks like his, run his math before you touch the budget. Multiply revenue by margin, subtract ad spend, and read what is left. That number tells you whose payroll you are on.

The store and its figures are lightly disguised to keep the client anonymous. The math works the same.