The hook is the first thing someone sees or hears in your ad: the opening line, the first image, the first second of video. Its only job is to make a scrolling person stop and pay attention. If the hook fails, nothing after it gets seen.
A hook is the opening element of an ad or a page that captures attention: the first line of a headline or caption, the first frame and the first spoken sentence of a video, the subject line of an email, the title of a push notification. It sits in front of the angle: the angle is the argument, the hook is the moment that earns the right to make it. Common hook types are a direct question to the audience, a surprising statement or number, a visible problem, a pattern interrupt (an unexpected image or movement) and a specific promise.
The first seconds carry a disproportionate share of a video ad’s effect. Meta has reported in its video creative guidance that a large share of a video ad’s value is delivered in the first three seconds, which is why platforms recommend putting the brand, the product and the core message at the start, and designing for sound-off viewing with captions. Buyers measure the hook with hook rate, usually three-second video views divided by impressions, and with click-through rate on static formats; a hook that lifts these numbers without lifting conversions is pulling in the wrong people.
Hooks are the cheapest element to test. Keeping the angle, visual style and offer fixed and changing only the opening line or the first two seconds isolates the variable, and new hooks are also the fastest response to creative fatigue: an audience that ignores a familiar ad will often stop for the same ad with a new opening. That makes hook libraries, collections of openings that worked by angle and audience, one of the most useful assets a media-buying team keeps.
The ethical limit is simple: the hook must be true to what follows. Fake notifications, fake news captions, “you have won” openings, or footage that implies a celebrity endorsement are hooks that stop the scroll and fail compliance, and they raise scrub and refund rates because the people they attract were promised something else.
A budgeting app pays a 9 USD CPA per verified registration. You keep the same 20-second video and change only the first two seconds, spending 600 USD per variant on social.
| Hook (first 2 seconds) | Hook rate | CTR | Registrations | Cost per registration |
|---|---|---|---|---|
| Logo and app name | 18% | 0.7% | 48 | 12.50 USD |
| “Where did 340 dollars go last month?” with a phone showing transactions | 34% | 1.3% | 92 | 6.52 USD |
| Person shouting “STOP SCROLLING!” | 41% | 1.1% | 55 | 10.91 USD |
The loudest hook wins the attention metric and loses on cost per registration, because it stops people without telling them what the ad is about. The question hook is specific to the audience’s problem and the only variant under the 9 USD payout. It becomes the template for the next hook round. Numbers are illustrative.
The algorithm handles targeting now; creative is what you control.
Advanced · 12 min readThe angle is the highest-leverage decision in a campaign — the specific reason THIS audience wants THIS offer.
Core · 11 min readHow paid social traffic works for affiliates: Meta and TikTok mechanics, strict policy and bans, algorithm-driven targeting, creative-led performan...
Beginner · 6 min readWhat media buyers actually do day to day: testing creatives, reading data, managing spend and scaling the campaigns that turn a profit.
The opening of an ad (the first line, frame or seconds) designed to stop a scrolling person and make them pay attention to what follows.
A metric for video ads, usually three-second views divided by impressions. It shows how many people the opening held; it should be read together with click-through and conversion rates.
Because that is when people decide whether to keep watching or scroll on. Meta’s creative guidance stresses front-loading the message for this reason.
Use the audience’s own words for their problem, lead with something specific (a number, a situation, a question), show the product or problem immediately, and test several openings on the same body.
References are listed as plain text on purpose; look them up by title and publisher. Updated: 2026-10-06.
Run them inside a network with real tracking, roughly 48-hour payouts and a dedicated manager.
Join the network