AI Creative Hook Rate Benchmarks 2026

AI Vidia publishes these ai creative hook rate benchmarks to answer the one question a growth-stage media buyer asks before signing off on a video budget: what thumbstop rate should AI-produced ads actually hit on Meta, TikTok, and YouTube in 2026. The short answer is that a strong hook rate sits above 30 percent on Meta Reels and above 40 percent on TikTok in-feed, measured as three-second video views divided by impressions, and the AI Vidia studio bench holds 38 percent on Meta Reels and 46 percent on TikTok across 1,834 AI videos shipped. Hook rate is the first gate on every paid social account, because an ad that loses the viewer in the first three seconds never earns the CTR or ROAS the rest of the funnel depends on.

Why hook rate decides the paid social quarter

Hook rate, also called thumbstop rate, is the share of impressions that watch past the first three seconds of a video ad. It is the single metric that gates every downstream number on a paid social account. A Meta Reels ad below a 20 percent hook rate rarely clears the 8 to 12 percent hold rate needed to drive a profitable CPA, and it burns budget at full CPM while it fails. The auction reads weak early retention as low relevance, so a poor hook rate raises CPM at the same time it starves CTR, which is why it sits at the top of every quarterly review in 2026.

The stakes are concrete on a mid-market account. A DTC brand spending EUR 50,000 a month on Meta video with a 25 percent average hook rate wastes roughly one in three impressions on viewers who scroll past inside three seconds. Moving that account to a 35 percent strong bench shifts about EUR 5,000 of monthly spend from wasted impressions into watched impressions, before any change to the offer or the audience. Meta for Business reports that campaigns with 5 or more creative variations see 30 to 50 percent lower CPA, and hook rate is the metric that decides which of those variations the algorithm keeps alive. Wyzowl 2025 reports 91 percent of businesses use video marketing and 30 percent cite production cost as the largest barrier, which is why most brands cannot test enough openers to find the strong band on their own.

Hook rate benchmarks by platform and format

The table below is the 2026 bench AI Vidia uses on commercial scoping and quarterly reviews. Each row maps a placement and format onto four bands: a weak hook rate that signals a broken opener, the category average, a strong top-quartile band, and the AI Vidia studio bench at steady state. Numbers come from AI Vidia studio runs over the last 12 months and live placement data across 12 brands in flight on Meta, TikTok, Instagram, and YouTube. Hook rate is three-second video views divided by impressions on every video row, and scroll-stop rate on the single static row.

Two rows decide most video accounts. The TikTok in-feed row sets the highest bar, with a 30 percent average and a 41 percent strong band, because sound-on autoplay and a fast native feed reward a hook that lands in under two seconds. The Meta Reels row sits lower, with a 26 percent average and a 34 percent strong band, because the placement mixes sound-off viewing with a slower scroll. A brand that ports a single 9:16 master across both placements without recutting the first two seconds leaves 8 to 12 points of hook rate on the table, which is the most common and most expensive mistake in the table.

Format fit moves the numbers as much as the opener itself. A 1:1 crop dropped into a 9:16 Reels or TikTok slot suppresses hook rate by 25 to 40 percent, because the letterboxed frame reads as an ad before the viewer processes the content. The static scroll-stop row runs on a different scale entirely, where a 17 percent strong rate on Meta Feed is a genuine result, so image and video benchmarks should never be compared on the same axis. For the CPM, CTR, and ROAS numbers that sit downstream of hook rate, see the companion 2026 benchmark report on CPM, CTR, and ROAS.

Framework 1: The AI Vidia Hook Rate Diagnostic

The Hook Rate Diagnostic is the strategic model AI Vidia runs on every account inside the first scoping call. Five checks, one verdict, and the output is the specific reason an account sits below the strong band. Run it on the last 90 days of ad data and the fix ranks itself before any new creative is briefed.

1. Step 1. The per-placement split. Pull hook rate for every placement separately instead of reading a blended account average, since a 28 percent blended number often hides a 42 percent TikTok line and an 18 percent Reels line. The blended average is the number that hides the problem, and the per-placement split is the number that names it. Fix the lowest placement first, because that is where the wasted spend concentrates.
2. Step 2. The first-frame audit. Freeze the first frame of every active video and score whether the subject, the motion, and the claim are all present before the half-second mark. Ads that open on a logo, a slow establishing shot, or an empty set lose 15 to 30 percent of available hook rate before the viewer sees anything worth staying for. The strong band starts with a subject and a reason in frame one.
3. Step 3. The hook-type read. Score the first three seconds of each concept against a six-part taxonomy: problem reveal, pattern interrupt, named claim, creator handoff, product close-up, and price tag. Accounts that ship fewer than three distinct hook types per concept cap their own hook rate, because a single opener style fatigues fast on a warm audience. Distinct hooks, not one clever frame, are what hold the bench over a quarter.
4. Step 4. The ratio and caption check. Confirm every creative runs in the native ratio for its placement and carries a sound-off caption in the first second, since most feeds play muted by default. A 9:16 native cut with a burned-in opening caption lifts hook rate by 25 to 40 percent against a reused 1:1 master with no caption. This is a production fix with no marginal generation cost on an AI pipeline.
5. Step 5. The hook-to-hold ratio. Divide hold rate by hook rate to catch openers that win attention and then lose it, which is the failure a raw hook rate number hides. A 40 percent hook that collapses to a 4 percent hold is worse than a 28 percent hook that holds at 12 percent, because the second ad actually reaches the offer. This ratio is the number AI Vidia optimises against, not hook rate alone.

Run the diagnostic once and the next 90 day plan writes itself. Accounts failing Step 1 or Step 4 do not need new creative concepts; they need a per-placement recut and a caption pass on the openers they already have. Accounts passing the format checks but failing Step 3 need a wider hook library before they generate another asset. Accounts that clear the first four checks but fail Step 5 are chasing a vanity hook rate and need the hold rate wired back into the brief, which is exactly what the second framework below covers.