Healthy ad returns were hiding weak new-customer margins.
The homeware brand had strong repeat purchasing, and Google and Meta both showed attractive ROAS. First-order profitability was weakening nonetheless, because acquisition reporting included returning customers and retargeting credit. The business needed to understand what media spend added, not simply which platform reported the most revenue.
Twelve weeks of new-customer performance
Over twelve weeks, new-customer CAC fell from €48 to €37, approximately 23%. New-customer orders increased 19%, and first-order margin after variable costs improved by seven percentage points.
The twelve-week results combine lower new-customer CAC, more first-time orders and stronger first-order margin after variable costs. They describe a healthier acquisition mix; platform attribution remains distinct from an incrementality claim about every channel.
Give acquisition and retention separate jobs
Order records distinguished new and returning customers. Channel decisions used first-order costs and margin, with future purchase assumptions kept separate from value already earned.
Product profitability and availability shaped promotion. Creative tests addressed buying concerns, while the review followed new-customer orders, acquisition cost and money retained after relevant costs.
Separate new-customer acquisition from returning demand.
Order records distinguished who was new, what they bought and the margin left after variable costs. Google and Meta received clear jobs rather than competing claims on the same customer.
| Economic distinction | Question for allocation | Misleading shortcut |
|---|---|---|
| New versus returning customer | What demand is the activity expected to add? | Treating all credited orders as acquisition |
| First-order contribution | What remains after relevant acquisition and fulfillment costs? | Using revenue alone to justify the spend |
| Product margin and stock | Can the offer create and fulfill worthwhile demand? | Funding a product only because its ROAS looks strong |
| Future value | What repeat behavior is supported rather than forecast? | Using optimistic retention to excuse every first-order loss |
Retargeting could still be worthwhile. Its results could not stand in for the economics of winning someone new.
Let profitability and buyer concerns guide the work
Retention activity still served a purpose, but it needed its own objective and budget. Expected repeat value could not be an unrestricted justification for poor first-order acquisition economics.
Margin and inventory also changed which demand was worth buying. A campaign could improve revenue while directing spend toward products that contributed less or could not reliably be fulfilled.
Google and Meta could serve different jobs without competing for the same sale. New-customer economics, contribution and inventory made allocation more useful than comparing each platform’s claimed return.