New Customer ROAS ncROAS
New customer ROAS ignores the revenue from people who had bought before. It counts the rest.
ncROAS = revenue from new customers ÷ ad spend
A campaign spends $5,000 and brings in $25,000. Of that, $14,000 came from first-time buyers and $11,000 from existing ones.
- Standard ROAS: 25,000 ÷ 5,000 = 5.0x
- ncROAS: 14,000 ÷ 5,000 = 2.8x
Both numbers are true. Only one of them describes growth.
The retargeting trap
Ad platforms optimise for conversions. The cheapest ones going are people who already meant to buy. Somebody who knows the brand, has the product in a basket, and sees a retargeting ad on the way back is an easy conversion — and often would have converted anyway.
The platform books it regardless. So the account looks excellent while the customer base stops growing, and the budget quietly turns into a discount on purchases that were already coming. Standard ROAS cannot show this. Both dollars look identical from where it sits.
Getting at the number
Measurement depends on what your stack can tell you apart:
- Shopify segments new against returning customers in its own reports, and the analytics tools built on top of it expose the same split
- GA4 has a new users segment that can be applied to conversion reports
- Meta offers a new customer acquisition objective in some regions and campaign types
Shops without any of that use a rough substitute: exclude the existing customer list from targeting and read the remaining ROAS as an approximation. It overstates a little, since some excluded customers buy anyway, but it moves in the right direction and takes an afternoon to set up.
Reading the two together
Watch them side by side rather than picking one.
| What it shows | |
|---|---|
| Standard ROAS | how the advertising performed overall |
| ncROAS | how much of that was actual growth |
| MER | whether total marketing spend is efficient at all |
The signal to look for is divergence. Standard ROAS holding steady while ncROAS falls means the business has stopped acquiring and started recycling. That can run for a long time before it shows up in revenue, and by then the customer base has aged.
A low ncROAS is not automatically bad
It depends on what a new customer is worth over their lifetime. If first orders are followed by repeat purchases, paying more than one order’s worth to acquire someone is the correct decision, and an ncROAS below break-even on the first purchase is the price of that. Where most customers buy once and leave, the first order has to pay for itself, and the same ncROAS is a problem.
So the figure needs customer lifetime value beside it before anyone acts.