Why Returning-Customer Revenue in ROAS Misleads Acquisition Decisions
Meta and Google ROAS includes revenue from buyers who would have purchased anyway. Splitting new vs returning ROAS exposes the real cost of acquiring a customer — and usually changes the budget call.
Quick answer
Platform-reported ROAS counts revenue from customers who already knew your brand and would have repurchased without the ad. That blend can make Meta and Google look 2-3x more efficient than they are at acquiring new customers. Split ROAS into new-customer ROAS and returning-customer ROAS — the new-customer figure is the only number that should drive acquisition budget decisions.
Returning-customer revenue in ROAS
Revenue from repeat buyers attributed to ad clicks that's blended into platform ROAS, inflating apparent acquisition efficiency.
Meta Ads Manager, Google Ads, and most ad platforms report a single ROAS figure that sums all attributed conversion revenue and divides by spend. That figure does not distinguish between a first-time buyer and a returning customer who saw a retargeting ad on the way to a purchase they were already going to make. For brands with a healthy repeat rate, 40-70% of platform-attributed revenue is often repeat purchase revenue — money that would have arrived anyway. The result: blended ROAS overstates the marginal return on acquisition spend, sometimes by a factor of two or three.
This isn't a tracking bug. The platform is doing exactly what you asked it to: count every purchase that touched an ad. The problem is interpretive — you treat the number as if it measures acquisition efficiency, when it actually measures attributed revenue from a mixed audience.
Why platform ROAS inflates acquisition efficiency
Three mechanisms compound. First, retargeting campaigns deliberately show ads to people who already visited your site — many were coming back regardless. Second, broad prospecting campaigns on Meta increasingly target lookalikes of existing customers, so the algorithm surfaces past buyers. Third, branded search on Google captures demand you generated through email, podcast, or organic — the ad just intercepted an existing intent signal.
An apparel store running €40k/month on Meta might see a reported 4.2x ROAS. Split it: new-customer ROAS lands at 1.6x, returning-customer ROAS at 9.8x. The acquisition campaigns are losing money on the first order — but the blended figure hides it inside the retargeting halo.
The repeat-rate trap
The healthier your repeat purchase behaviour, the more your platform ROAS lies. Subscription brands and beauty refill SKUs are the worst-affected — their reported ROAS can sit 3x above true new-customer ROAS, masking acquisition unprofitability for quarters at a time.
How to detect the gap in your account
The first signal is a divergence between platform ROAS and Shopify's customer acquisition reports. If Meta says 4x but Shopify shows that paid social brought in 200 new customers at €85 AOV against €40k spend, your true new-customer ROAS is 0.43x. That delta is the returning-customer revenue you've been counting as acquisition.
A second signal: scale spend by 30% and watch what happens. If reported ROAS holds but new orders barely move, you were buying retargeting impressions on existing customers. The blended number is a flattering mirror; the new-customer count is the truth.
How much blended ROAS overstates new-customer ROAS, by repeat-purchase rate
| Repeat-purchase rate | Reported blended ROAS | True new-customer ROAS | Overstatement |
|---|---|---|---|
| 15% (low-repeat, e.g. mattress) | 3.2x | 2.7x | 1.2x |
| 30% (typical apparel) | 4.0x | 2.1x | 1.9x |
| 45% (beauty, skincare) | 4.5x | 1.8x | 2.5x |
| 60% (refill / subscription) | 5.1x | 1.5x | 3.4x |
How to fix it: split ROAS at the source
The clean fix is to pass a new_vs_returning flag into your conversion event. In Shopify, you can read customer.orders_count at checkout and send it as a custom parameter on the Purchase event to Meta and GA4. Filter Ads Manager by that parameter and you get two ROAS columns instead of one — the only view that should inform acquisition budget.
For a faster directional read without dev work, build a Shopify segment of first-order customers in the last 30 days, divide that revenue by paid spend in the same window, and compare against blended ROAS. It's coarse but it surfaces the gap in an afternoon. The detailed walkthrough lives in our guide to splitting ROAS in Meta Ads Manager and the companion piece on new-customer ROAS in Shopify + GA4.
Validate with a holdout
Even split ROAS is correlational. Run a geo-holdout or ghost-bid incrementality test on prospecting campaigns to confirm the new-customer ROAS you're measuring is actually caused by the ads — not demand that would have converted through organic or email. See our note on incrementality tests for new-customer ROAS.
What changes once you report it correctly
Most teams discover that 20-40% of their paid social spend is buying repeat purchases at a premium. Reallocating that budget to email, SMS, and retention typically improves contribution margin without dropping new-customer count. The retargeting campaigns don't disappear — they shrink to the budget the incremental new orders actually justify.
The harder organisational shift is reporting. Finance, the CEO, and the agency all anchor on blended ROAS because it's the number the platform shows. Expect a quarter of recalibration before new-customer ROAS becomes the default in budget reviews — and read our piece on why operators keep reporting blended ROAS anyway for the politics of that change.
Common questions about new vs blended ROAS
Meta's conversion pixel fires on every purchase regardless of customer history. The platform has no native concept of 'new customer' unless you pass it as a custom parameter, so the default report aggregates all attributed revenue. It's a measurement choice, not a bug — but the default does flatter the channel.
For most online stores in the €1M-€15M band, a new-customer ROAS of 1.0-1.5x at first order is acceptable if your second-order rate within 90 days is above 25%. Below 1.0x you're betting heavily on LTV; above 2.0x you're probably under-spending on acquisition.
Not necessarily — but you should size it to incremental lift, not attributed revenue. A ghost-bid test on retargeting often shows 30-60% of the conversions would have happened organically. Cap retargeting spend at the incremental portion and redirect the rest to prospecting or retention.
Blended CAC mixes paid spend across all new customers (including organic-acquired ones); paid CAC isolates the cost per new customer from paid channels. New-customer ROAS is the revenue-side mirror of paid CAC — both strip out the noise that makes paid look more efficient than it is.
Yes, and branded search is the worst offender. Branded search ROAS often reports 15-30x because it intercepts existing demand. Splitting by new vs returning customer reveals most branded clicks are repeat buyers — useful to know before you scale that budget.
Subscription products generate frequent repurchase events that all fire the conversion pixel. With 60%+ repeat rates, the majority of platform-attributed revenue is recurring billing or refills the customer would have made anyway. The reported-vs-real ROAS gap can exceed 3x.
That's a good directional check but it doesn't attribute first orders to specific campaigns. You still need the channel split — Shopify tells you how many new customers arrived, Meta (with the new-customer parameter set) tells you which campaigns brought them.
Yes. MMM and incrementality answer 'how much did this channel contribute?' at a macro level. Split ROAS answers 'which campaign within the channel is acquiring profitably?' — you need both, at different decision cadences.
On Shopify with the Meta and GA4 native integrations, passing a customer_status parameter takes a developer half a day, or you can use a tag-manager rule and have it live in two hours. Without dev work, a weekly export-and-segment workflow gets you 80% of the value.
Usually yes. Most teams find at least one campaign whose reported 4x+ ROAS hides a sub-1x new-customer ROAS. The first response is rarely to kill it, but to cap spend at the incremental new-customer contribution and shift the rest to prospecting or retention.
Track CAC, channels, and funnel conversion in one place
Metricuno connects ad spend, funnel events, and revenue so you can see CAC by channel, cohort, and campaign — without stitching together five tools.