Ask most founders how their Meta ads are performing and you will hear one number: ROAS. Return on ad spend. A 4x looks good, a 2x looks bad, and budgets go up or down on that basis.
The problem is that ROAS, as the platform reports it, answers the wrong question. It tells you how much revenue the platform is willing to claim. It does not tell you how many new customers you paid for, or whether you made any money doing it.
What ROAS hides
Platform ROAS counts every purchase the ad platform can connect to an ad. That includes your existing customers, people who were already on their way to buy, and people who clicked an ad after seeing your email. All of that revenue is real. Very little of it was created by the ad.
The result is an account that looks healthiest when it is targeting your warmest audience. The platform learns that returning buyers convert, spends more on reaching them, and reports a strong ROAS while the number of genuinely new customers flattens. Growth stalls and the dashboard says everything is fine.
The numbers to scale on instead
- Cost per purchase. What you actually pay for each order the ads produce. Simple, hard to flatter, and directly comparable to your margin.
- New customer acquisition cost. What you pay for each first time customer. This is the number that tells you whether the business is growing or recycling the same buyers.
- Contribution after ad spend. What is left from those orders once product, delivery and ad costs are paid. This is the number that tells you whether growth is profitable.
Judge every one of these against your margin. A cost per purchase that looks low can still lose money on a low margin product. A higher one can be very profitable if the customer goes on to buy again.
An illustrative example
Two months on the same £15,000 ad budget. These figures are illustrative, not from a client account.
- Month one reports a 4x ROAS, £60,000 of attributed revenue. Most of those orders are from returning customers. 300 new customers were acquired, a new customer acquisition cost of £50.
- Month two reports a 3x ROAS, £45,000 of attributed revenue. The account has been pushed towards cold audiences. 600 new customers were acquired, a new customer acquisition cost of £25.
On ROAS, month one wins and month two would be cut. On what actually grows the business, month two doubled the customer base at half the cost per new customer. Every one of those customers now enters your email flows and can buy again without you paying for them twice.
What this looked like at RB Club
When we started with RB Club, spend was spread across more than eight campaigns, each too small to give the algorithm a clear signal, and cost per acquisition was £32. The account was consolidated into a structure built to learn and scale, and measurement moved to first conversion incremental attribution, so decisions were made on new customers acquired rather than on platform ROAS that rewarded returning buyers.
Cost per acquisition fell from £32 to £13. Net new customer acquisition rose by 179%, and the account held a 5.2x return on new customer acquisition cost. You can read the full breakdown in the RB Club case study.
Why acquisition cannot be fixed on its own
What you can afford to pay for a new customer depends on what that customer is worth after the first order. That is decided by your store's conversion rate and your retention, not by the ad account. A brand with strong email flows and a high returning customer rate can afford to pay more to acquire, and outbid competitors who cannot. That is why the ACR Growth Method treats acquisition, conversion and retention as one system.
What to change in your own account
- Report new customers separately. Track new customer acquisition cost alongside cost per purchase every week.
- Stop scaling on platform ROAS alone. A falling ROAS with rising new customer volume can be the healthiest month you have had.
- Consolidate. If budget is spread across many small campaigns, the account cannot learn. Consolidation often lowers cost per acquisition before a single new ad is made.
- Set your ceiling from margin. Work out the most you can pay for a new customer and stay profitable, then scale up to it.
The free Meta Ads Testing System is the structure we run on every client account, and the growth calculator shows what better acquisition, conversion and retention could do together. If you would rather it was run for you, this is how we work as a Meta ads agency.