Hook fatigue kills ads from the top. The first three to five seconds are the only creative element every impression is guaranteed to see, so the hook accumulates frequency faster than the body. That means you need a hook rotation schedule that is separate from the ad set. A body can still be winning while the hook has already burned out, and swapping the whole ad set then is waste.
Why hooks fatigue before the rest of the ad
Every impression charges the opening. In Meta and TikTok feeds, the auction counts an impression when the first frame enters view, not after a completed view. On YouTube in-stream, the first five seconds are the guaranteed window before the skip. The body is only seen by the people who stay.
That asymmetry is the mechanic behind hook fatigue. At Meta Reels' median hook rate of 28 percent, 100,000 impressions means 100,000 people saw the first frame, but only about 28,000 reached the body. The opening is carrying four times the exposure load. Meta internal research says CTR drops 45 percent after a fourth exposure to the same creative, and the repeated part is almost always the hook. Weekly frequency of 2.5 on Meta prospecting is where decline begins, but a retargeting pool or a high-frequency account can hit that in days.
It is a concentrated version of creative fatigue. The first seconds die earlier because the rest of the ad has not had the same number of completed views.
Typical hook lifetime by placement
Placement changes the clock. Our 2026 fatigue benchmark puts platform medians from a composite of published studies covering more than 12,000 ads. The refresh cadences below are operational ranges we use on accounts, not study findings.
| Placement | Median hook rate | Frequency ceiling | Hook refresh cadence we use | What usually fatigues first |
|---|---|---|---|---|
| Meta Reels | 28% | 2.5 | Fortnightly | First 2 seconds and text overlay |
| Meta feed | 28% | 2.5 | Fortnightly | Opening frame or visual hook |
| TikTok | 33% | about 3.0 | Weekly | First beat, movement and caption |
| YouTube in-stream | 22% | 3 to 7 per week | Every two to three weeks | First 5 seconds before skip |
Vertical changes the window too. Food and beverage can hit a 40 percent CTR decline in 9 days. Beauty and DTC takes roughly 18 days. B2B SaaS can stretch to around 28 days. The pattern stays the same: the hook approaches its ceiling first, then the body follows.
How to keep a hook bench warm
A hook bench means you never promote a new, unproven opening directly into a scaled campaign. Keep two hook variations per body that have both passed the same performance bar. The primary runs at volume. The secondary sits in the same ad set or a small test campaign with enough spend to stay measurable. When the primary hook rate drops below the secondary, you swap. Then you promote a third hook into the bench.
On TikTok, most accounts should refresh hooks weekly. On Meta feed, a fortnight is usually enough. We treat a hook rate drop of five percentage points as the warning sign. If a TikTok hook falls from 33 percent to 28 percent while the body holds, the fix is the opening, not the offer.
Use the ad fatigue calculator if you want a starting point for refresh timing on a specific account.
Why hook rotation must be separate from ad set rotation
Most teams treat creative fatigue as one problem: the ad set is tired, so they pause it. That hides a different problem. The body may still be fine. The opening is dead.
A separate hook schedule means tracking hook rate and CTR in the first seconds, then swapping only that part while leaving the body and targeting live. One testimonial or demo body can run for weeks with three or four hook variations in front of it. Most active campaigns need 8 to 20 live variations in rotation, but that count includes hook variants, not just body concepts.
Our benchmark report found brands shipping 15 to 50 creative variants a month see 3 to 5 times longer campaign lifespan than quarterly refreshers. Many of those variants are new openings, not new bodies.
Where Genyad fits
Genyad is built for this because it treats a hook as a script and shot selection problem, not an edit problem. You upload the source footage once. It transcribes and tags every clip. Each variation is a fresh script, shot selection, voiceover, caption set and export built from that library rather than a re-cut of the same timeline. A new hook can pull different lines or footage from the existing library.
One standard variation costs 1 credit. The free plan gives 5 video ad variations, so you can test a hook bench without paying. Starter is €29 for 15 credits, and Growth is €99 for 65 credits. Credits never expire.
Genyad does not generate AI avatars or synthetic presenters. If you need an AI actor reading a hook, use Arcads or HeyGen; Arcads was listed at €100/month and HeyGen at $29/month in August 2026. Genyad is for teams that already have footage and need new openings cut from it.
Frequently asked questions
What is hook fatigue?
Hook fatigue is the decline in hook rate or CTR caused by the first few seconds of an ad being seen too often, before the body has fatigued. It happens because the opening is the only creative element present in every impression. The wider 2026 fatigue benchmark shows creative CTR declines 15 to 20 percent in the first two weeks, and the hook is usually where that decline shows up first.
How often should you change hooks?
On TikTok and app install campaigns, most accounts should refresh hooks weekly. On Meta feed, a fortnight is usually enough. YouTube in-stream can stretch to every two or three weeks. These are starting points, so adjust when hook rate drops below the account's own baseline.
Does hook rotation mean pausing the ad set?
No. You often leave the body, audience and budget live and swap only the opening. Pausing the ad set because the hook is tired wastes a body that may still have weeks of life. Keep two hook winners warm so you can promote the second without restarting the campaign.
Can you reuse the same body with different hooks?
Yes. If the body is still holding conversion rate or view-through, the problem is the opening. Genyad makes this easier because it rebuilds variations from the original footage rather than re-cutting the same timeline. Each new hook variation costs 1 credit, so running a hook bench is cheap relative to losing a winning ad set.