Meta ads benchmarks for e-commerce, and how to use them
CPM, CTR, CPA and ROAS ranges for online stores — plus the reason a benchmark is a sanity check on your setup, never a performance target.
Everyone wants a benchmark and almost nobody should optimise towards one. Both halves of that sentence are worth taking seriously.
A benchmark is useful for exactly one thing: telling you whether a number is plausible. If your CTR is 0.12%, something is wrong with the creative or the audience, and knowing the typical range saves you a week of wondering. If your CTR is 1.3%, the range tells you to stop tuning the ad and go look at your product page instead.
What a benchmark cannot do is tell you whether you are making money. Only your margin does that.
The ranges
These are broad orientations for consumer e-commerce in Western European and North American markets, prospecting campaigns, conversion objective. Treat them as order-of-magnitude, not as targets.
| Metric | Usually low | Typical band | Usually high |
|---|---|---|---|
| CPM | under $6 | $8–$20 | over $30 |
| CTR (link) | under 0.6% | 0.9%–1.8% | over 2.5% |
| CPC (link) | under $0.40 | $0.60–$1.50 | over $2.50 |
| Add-to-cart rate | under 3% | 5%–10% | over 12% |
| Purchase rate (of clicks) | under 0.8% | 1.5%–3% | over 4% |
| Frequency (7 days) | under 1.3 | 1.5–2.5 | over 3.5 |
Deliberately absent from that table: ROAS and CPA. They belong in the next section, because quoting a range for them does more harm than good.
Why there is no useful ROAS benchmark
ROAS without margin is not a number, it is a rumour.
break-even ROAS = 1 / gross marginA store at 70% margin breaks even at 1.43 and is genuinely profitable at 2.0. A store at 30% margin breaks even at 3.33 and is losing money at 3.0 — the same 3.0 that would be an excellent result for the first store.
So when someone tells you “aim for 3x”, they are quoting a number that is either trivially easy or literally impossible depending on a fact about your business that they do not know. Work out your own break-even before you compare yourself to anyone, and read the six metrics that decide your results in that light.
The same applies to CPA. A $40 CPA is a disaster on a $50 product and a bargain on a $400 one.
What actually moves these numbers
Before you conclude that your CPM is “too high”, check the four things that move it more than your skill does:
Country. CPMs in the Nordics, Switzerland and Australia routinely run two to three times those in Southern or Eastern Europe. A $25 CPM in Norway and a $9 CPM in Portugal can be the same relative price.
Season. Q4 raises CPMs across every market — November in particular. A 40% increase between October and late November is normal and says nothing about your campaigns.
Audience width. Narrow audiences cost more per impression, always. Three stacked interests will produce a CPM well above broad targeting on the same product.
Placement mix. Feed impressions cost more than Stories and Reels impressions. An ad set that skews to Reels will show a lower CPM and often a lower CTR, and neither figure is comparable to a Feed-heavy ad set.
Category. Fashion and beauty see higher CTR and lower purchase rates than, say, tools or homeware. Consider the benchmark band as shifting by category, not fixed.
How to use a benchmark properly
Use it as a triage tool, in this order:
- Is CTR in band? If not, the creative or the audience is the problem, and nothing downstream matters yet.
- Is CPM in band for your country and season? If it is far above, your audience is too narrow or you are in an expensive window.
- Is add-to-cart rate in band? If CTR is fine and this is not, the ad is promising something your product page does not deliver.
- Is purchase rate in band? If add-to-cart is fine and this is not, your problem is checkout: shipping revealed late, forced accounts, missing payment methods.
- Only now, compare ROAS to your own break-even.
That order is the whole value of benchmarks. Each step isolates one part of the chain, so you find the broken link instead of guessing.
Build your own benchmark as soon as you can
Third-party ranges are a scaffold you should be trying to replace. After six to eight weeks of consistent spending you have something better: your own historical bands, per country and per category.
Keep a simple record — month, country, CPM, CTR, add-to-cart rate, purchase rate. Within two months it will tell you more than any published table, because it controls for your product, your market and your price point automatically.
That record is also what makes forecasting a new campaign possible at all.
AdPilot’s simulator uses ranges like these, adjusted for your market and category, to project impressions, clicks, conversions and expected ROAS for a budget before you commit it.