If you ask ten ecommerce founders what a 'good' ROAS is, you'll get ten different answers — usually anchored to whatever number they happened to hit last quarter. The truth is that a healthy return on ad spend depends on your margins, your channel mix, and how mature your owned audience is. Benchmarks only help if you read them in context.
The benchmarks we see across $10M+ in managed media
Across the NZ and Pacific stores we manage, blended paid-media ROAS typically lands in these ranges for established brands with decent creative refresh cycles:
- Google Ads (Search + Shopping): 4x–8x for mature accounts, 2x–3x during aggressive scaling.
- Meta Ads: 2.5x–5x blended; prospecting often sits at 1.5x–2.5x with retargeting carrying the number up.
- Klaviyo / email automations: effectively 20x–40x because the media cost is near zero — this is why owned channels distort blended ROAS so heavily.
Platform-attributed ROAS undercounts real revenue by 20–40% post-iOS 14.5. We coach clients to measure blended ROAS (total revenue ÷ total ad spend) and use platform numbers only as directional signals.
When a 'low' ROAS is actually winning
A 1.8x ROAS looks bad in isolation. But if your gross margin is 65% and that acquisition feed your email automations — which then return 25x over 90 days — the 1.8x is the top of a very profitable funnel. We've run accounts where a sub-2x prospecting ROAS produced a 6x blended return once retention kicked in.
"Stop optimising the platform number. Optimise the customer and the platform number fixes itself."
— James Sharma, Founder — Pulse Advertising
How to use these benchmarks
Use them as a gut-check, not a target. If your Google Shopping sits at 1.2x while the market sits at 5x, the gap is almost always tracking, feed quality, or bidding strategy — not budget. Before you add spend, fix attribution and creative first. Then scale into a number you can actually trust.
Want this applied to your brand?
Book a free 15-minute growth audit and we'll show you where the revenue is hiding in your funnel.
