Which creative formats actually beat account average
Knowing an ad is stale is easy. Knowing what beats it is not. We scored every creative format against its own account's average, so a replacement can be chosen from evidence rather than from whatever the designer has time for.
Creative formats ranked
Relative performance against account average. 100 = that account's own average.
| Creative format | Vs the account's average | Results |
|---|---|---|
| Offer / price-led | Above average | +55% |
| Customer review | Above average | +20% |
| Unscripted real footage | Above average | +13% |
| Product demo | Above average | +7% |
| Styled lifestyle photo | Below average | −34% |
| Paid creator, scripted | Below average | −39% |
Methodology. 100 = that account's own average. First-party data, read from the platforms' reporting APIs, 90-day windows. No format is published from fewer than three accounts, and no individual advertiser's figures are ever disclosed.
Two things worth knowing before you build the replacement
01 — Unscripted beats scripted, every time we could compare them
In four accounts running both side by side, real unrehearsed footage beat paid creators reading a brief in all four, with a median advantage of 1.7x the return per dollar.
02 — The prettiest format is the worst performer
Styled lifestyle photography, the thing most brands spend their production budget on, indexed 66 and did not beat account average in a single one of the three accounts running it. Meanwhile the offer card, which costs almost nothing to make, beat average in every account that ran one.
These are format-level findings, not category quirks. They held across footwear, supplements, beauty, homewares and financial services — categories with nothing in common except a human scrolling past. That consistency is why they travel.
What sits behind the numbers
Questions, answered.
What does an index of 155 mean?
That format returned 55 percent better than the average creative in the same account over the same window. Indexing against the account's own average is what makes the numbers comparable between advertisers of very different sizes.
Why not publish absolute ROAS or CPA figures?
Because they would tell you more about the accounts than about the formats. A CPA is a function of category, price point, market and offer. An index against the account's own average strips that out.
Whose accounts are these?
Advertisers we run or have run paid media for. No brand is named and no individual advertiser's figures are disclosed, which is also why no format is published from fewer than three accounts.
How often does this update?
Quarterly. Each update is dated on this page, so you can see how recent the numbers are.
Build the replacement from the format that wins
Bring the ad that is carrying your account and we will build the alternatives the benchmark says should beat it.