Search "Meta hook rate benchmark 2026" and you will get a confident answer within about four seconds. Median hook rate 28%. Top decile 45%. Aim for 30% thumbstop. The numbers appear across dozens of pages, stated flatly, often in a tidy table.
We went looking for where they came from. They do not come from anywhere.
The most-cited version traces to a single post that attributes its figures to three other sites. Those three sites publish no methodology, no sample size, no date range, and no named analyst. Follow them further and the citations become circular — each one sourcing the others.
We could not find a single 2026 hook rate, thumbstop rate or hold rate benchmark published by a source with a stated sample, a stated window and a named methodology. Not one.
Hook rate is the metric most brands use to kill or keep a creative. If the threshold you are judging against has no methodology behind it, you are making real production decisions — cutting concepts, reshooting, reallocating budget — against a number nobody stands behind.
Notably, Motion — which runs the most rigorous public creative dataset in the industry — does not publish hook rate or hold rate at all. They have $1.29 billion in Meta spend across 578,750 creatives and 6,015 advertiser accounts, and they declined to publish those two metrics. That is a meaningful silence. The people with the best data chose not to make the claim.
Motion's benchmarks do publish something more useful: hit rate. An ad counts as a hit if it spends at least 10x the account median and at least $500. It is a blunt definition, and it has a real limitation — it measures how Meta allocates budget, not ROAS or revenue. But it is transparent, and it is enormous.
Roughly one ad in twenty becomes a real winner, and those winners absorb over half of all spend. About half of all ads never receive meaningful budget at all.
That single ratio explains more about Meta performance than any hook rate threshold. It reframes the question from "is this ad good enough" to "how many shots have I taken."
Here is what accounts actually launch per week, by spend tier — median, then top quartile:
Hit rate rises with tier too — roughly 3.8% at the smallest accounts to 8.2% at enterprise. Bigger accounts are not just launching more, they are hitting more often, because volume is how you find the angle that works.
Run the arithmetic at a 5% hit rate: 20 ads gets you roughly one winner. 50 ads gets you two or three. If you launch four creatives a quarter, you are not testing — you are guessing, and then attributing the outcome to hook rate.
Andrew Foxwell's read on the same dataset is the line worth remembering: a high hit rate may actually signal that an account is not testing enough.
That inverts how most brands think about it. If nearly everything you launch performs, you are not being efficient. You are being conservative, and you have stopped looking for the outlier that would have doubled the account.
Selected asset types, over Motion's September-to-January window:
Lifestyle product imagery also placed above UGC on Motion's leaderboard, which puts UGC fourth overall rather than third.
By hook: offer only 9.29%, confession 8.74%, curiosity 7.77%, bold claim 7.19%. By visual style: letter format 10.83%, unconventional text placement 9.63%, ASMR 8.58%, founder 8.57%.
Two honest caveats, because the numbers get repeated without them:
And the part that complicates our own position: UGC ranked fourth, not first. We would rather say that than pretend otherwise — the fair claim for UGC is that it beats expensive polish at a fraction of the cost and can be produced at the volume the auction requires, not that it wins every head-to-head.
For actual account performance, Triple Whale's dataset — 40,000+ brands, twelve months ending July 2026 — is the most credible DTC-specific source we found:
The pairing at the top of that list is the interesting one. CTR up 16% while conversion rate fell 4.7%. More people are clicking and fewer are buying. Click-side metrics are getting easier to hit and meaning less — which is a good argument for judging creative on downstream outcomes rather than on the top-of-funnel numbers that are inflating industry-wide.
Stop benchmarking against strangers. Benchmark against yourself.
Our best-performing asset to date hit a 28.98% thumbstop rate on 292K impressions, with a 5.62% hold rate and 1.79% CTR. We publish it as one result from one campaign — not as a benchmark, because a single account is not a dataset. That distinction is the whole point of this post.
The confident benchmark numbers circulating for Meta creative in 2026 have, as far as we can trace them, no methodology behind them at all. The organization with the best creative dataset in the industry deliberately declined to publish the same metrics.
What is real: about 1 in 20 ads wins, winners take over half the spend, and the accounts that find more winners are simply launching more. That is a production problem, not an analytics problem.
If your constraint is getting enough distinct concepts into the auction, the Content Bank is built around exactly that math — 20 assets from 8 concepts and 12 hook variations in one production cycle.
Sources: Motion Creative Benchmarks 2026 ($1.29B Meta spend, 578,750 creatives, 6,015 accounts, Sept 1 2025–Jan 1 2026); Motion, "How many creatives do you actually need to launch" (Mar 2026); Foxwell Digital, "Motion Creative Benchmarks 2026: 8 Key Takeaways" (Mar 2026); Triple Whale Facebook Ads benchmarks, 40,000+ brands, Aug 2025–Jul 2026 (published Aug 18, 2026). Get Reel Ads campaign figures are first-party, from the Dog Is Human campaign.
The Content Bank delivers 20 assets from 8 concepts and 12 hook variations in a single production cycle.