Growth that costs too much
Spend goes up without the unit economics under it. CAC drifts away from what a first order returns. Top line climbs while contribution margin quietly compresses.
Four pillars. Twenty checks.
01 · Offer
Does a new customer pay back?
- What the first order returns in gross profit sets the ceiling on what a customer is allowed to cost. Every other number is judged against it.
- Five checks: unit economics, what lifts order value, which products carry the spend, what repeat is worth, whether the page justifies the price.
Most CPA problems are offer problems wearing a creative costume. If the first order cannot pay for the customer, better ads only buy the loss faster.
02 · Conversion
Does the site turn the click into the order?
- The ad does its job. We check whether the page echoes the promise it made, on a phone, above the fold.
- Then the funnel: where between session and checkout people leave, and whether testing is a habit with an owner or an occasional idea.
Post-click is where conversion rate goes quietly, and it is the one discipline nobody owns by default.
03 · Distribution
Is the budget going where it compounds?
- Every campaign, what it is for, and whether your own customers are being bought back as if they were new.
- Then the curve: the budget where profit after ads peaks, what the next dollar returns, what spending past the peak costs a month.
This is where "I pushed spend and ROAS dropped" comes from. The money had nowhere clean to go, so the algorithm spent it on people you already had.
04 · Creative
Is the next winner already in testing?
- New ads launched against what you spend, how many of them hit, and how long a winner lasts before it fades.
- How much of the account rides on the top two ads, and whether anything is queued behind them.
Creative is stock, not a one-off. It wears out, and the replacement has to be in testing before the current one goes. Model it, or find out the week you can least afford to.
