New vs Returning Customer ROAS
Why blended ROAS hides the truth.
Our best-looking campaign for a year was called "Warm audiences 2". ROAS 6.8. We fed it like a prize pig. When we finally split its revenue by first-time versus repeat buyers, 84% of it came from people who had bought in the previous 60 days. We had been paying Meta to send our own customers a reminder we could have emailed for free.
The one-sentence version
Blended ROAS mixes new customers with people who were coming back anyway; new-customer ROAS is the part you actually bought. Two campaigns with the same blended ROAS can be a great acquisition engine and an expensive coupon.
The setup
Split the number
Try it: one campaign, two numbers
Blended ROAS vs new-customer ROAS
Blended ROAS (what Ads Manager shows)
4.0x
Share of revenue from returning customers
60%
Blended ROAS
4.0x
New-customer ROAS
1.6x
Revenue you paid for twice
$24,000
on $10,000 spend
This campaign is actually acquiring.
Most of its revenue is first-time buyers. Scale this one; it is building the list the other campaigns will later harvest.
The split does not make the campaign worse. It just tells you which ones cost money.
Three numbers to keep apart
And what each is for
Blended ROAS
All revenue over spend. Good for a quick health check of the whole account. Useless for deciding between campaigns.
New-customer ROAS
First-order revenue from first-time buyers over spend. The acquisition number. Sort campaigns by this.
MER (marketing efficiency ratio)
Total revenue over total ad spend, across everything. The number finance should track weekly. It cannot be gamed by attribution.
Why it matters
Returning customers are the easiest to "win"
Retargeting, brand search, and email all specialise in people who already know you. They post enormous ROAS for the same reason a coupon does. Left unsplit, they absorb budget from the prospecting that filled the list in the first place. Six months later the list stops growing and everyone wonders why the "best" campaigns are fading.
"Warm audiences 2" is still running. Its budget is a tenth of what it was. Its job title changed from "acquisition" to "reminder", which is what it always was.
The fix
Split by order history, not by audience name
Shopify knows whether an order's customer has ordered before. Order-verified attribution carries that flag onto every attributed order, so each campaign gets a new-customer share and a new-customer ROAS without any audience setup. In TrueROAS it is a column, and the column is sortable. Sort by it once and you will not go back.
Honest footnote
When blended is fine
One-time-purchase products with almost no repeat rate: new and blended are the same number. Everyone else, split it.
Questions people actually ask
Straight answers
What counts as a returning customer?
Someone with a previous order on the same store, matched by email or customer ID at checkout. Guest checkouts with a new email look new even when they are not, so treat the returning share as a floor.
Is a low NC-ROAS always bad?
No. A retention campaign is allowed to sell to returning customers. The problem is a retention campaign that is being budgeted as acquisition because its blended ROAS looks amazing.
What is a good new-customer ROAS?
Whatever covers first-order margin plus the share of lifetime value you are willing to invest. Many stores are profitable at 1.5x NC-ROAS if repeat rate is strong. There is no universal number.
Does Meta show this?
Only if you give it a customer list to exclude, and then only roughly. Order-verified attribution knows from Shopify whether the buyer had ordered before.
See your own numbers, verified against your orders.
Free under $10k/month in revenue. 14-day trial above that. Nothing changes in your ad accounts. You will know within a week which platform has been telling stories.
"I built this because I was making $50k-a-month decisions on numbers I knew were wrong. If TrueROAS does not show you something Meta or Google got wrong in the first two weeks, tell me and I will help you cancel."
Rasmus Arvidsson
Founder, TrueROAS. Answers support himself.
Keep going