A viral hook is the attention-grabbing opening of a video, typically lasting 1-3 seconds. It's designed to stop viewers from scrolling and compel them to watch the rest. Effective hooks use curiosity, controversy, or immediate value. Examples include "Stop scrolling if you..." or "I tried this for 30 days and..."
Why hooks decide whether an ad ships or dies
On auction-driven feeds (TikTok, Reels, Shorts, Facebook), ad ranking is heavily influenced by early-session retention. TikTok's creative guidelines explicitly flag the first three seconds as the "hook window," and Meta's own performance data shows that a 1-second improvement in view-through rate correlates with roughly 10% lower CPA on most DTC verticals. In practice this means a weak hook does not produce a slightly-worse ad — it produces an ad the platform refuses to distribute, which zeroes out the rest of the creative no matter how polished it is.
This is why every serious performance marketer tests hooks as a separate variable from the rest of the creative. Ship 4–6 hook variants on the same body script, kill the ones with <25% hook rate, and only then optimize the payoff.
The seven hook archetypes that consistently perform
1. Pattern interrupt. A visual or verbal anomaly the viewer cannot instantly categorize — "Wait, why is she holding that?" — that buys another second of attention.
2. Contrarian claim. "Stop using [popular advice]. Here's why." Works because the viewer must watch to evaluate whether to update their belief.
3. Stakes reveal. "I spent $12,000 testing this so you don't have to." Credential + loss signals combined — the viewer gets to steal the lesson for free.
4. Identity hook. "If you're a [specific audience], this is for you." Most effective when audience-specific enough that the viewer self-selects — "If you're a Shopify store owner under $50k/month" out-performs "If you're an entrepreneur."
5. Tangible outcome. "I went from 2x ROAS to 6x using one change." Specific number, specific direction, specific timeframe.
6. Negative opening. "This $50 product is a scam. Here's what works instead." Negative framings have 1.3–1.8× higher hook rate on TikTok per several public creator case studies.
7. Question hook. "Why is nobody talking about [X]?" Works because curiosity is asymmetric — the cost of watching another 5 seconds is low, the payoff of the answer feels high.
Example: hook variants on a single script
A DTC supplement brand ran one script ("I tried this for 30 days and my sleep score went up 22 points") through five different hooks on the same AI creator. The identity hook ("If your Oura ring sleep score is under 80...") produced a 38% hook rate. The contrarian hook ("Stop buying melatonin. Here's what actually works.") produced 41%. The body was identical in every variant. The difference in ROAS across the five variants was 3.2× from best to worst.
This pattern — same creative, different hook, materially different ROAS — is the reason hook testing is the highest-leverage A/B test in paid social.
How AI UGC changes hook testing economics
Traditional UGC makes hook testing expensive because each variant requires a reshoot or re-delivery from the creator. AI UGC makes hooks effectively free to test — you can ship 10 hook variants on one script in under an hour. The practical consequence is that hook rate stops being a once-a-quarter research project and becomes a weekly optimization loop.
UGC Copilot generates multiple hook variations automatically for every script, pulled from the seven archetypes above plus trend-specific variations based on what's currently working in your category. This turns "pick the best hook" into "test the top five and let the data decide."
Related concepts
A hook is measured by hook rate (3-second view-through). The rest of the video is evaluated on watch time and conversion rate. Hooks are the single biggest driver of creative fatigue — when a hook stops working, refresh it long before the payoff.