The First Two Seconds: Writing a Reel Hook That Holds

A reel does not fail at the end. It fails at second two, before the viewer has consciously decided anything. So the whole job is the opener, and how long the reel runs barely matters if that opener misses.

What makes a reel hook hold in the first two seconds

• Hold: open on tension, a bold claim, a question, or a visual pattern break.
• Lose: a logo, a title card, or "hi guys, welcome back". Three seconds gone.
• Say the most interesting thing first, not last.
• If the first line would not stop your own thumb, rewrite it.

The hook is visual before it is verbal

Most advice about hooks is about the words. The words are half of it. A viewer scrolling a feed sees your first frame before they process a single syllable, and if that frame looks like nothing is happening, the thumb keeps moving while your excellent opening line plays to no one.

A strong opening frame has motion already underway, a face mid-expression, or something visually unusual enough to interrupt the pattern of the feed. A weak one is a person sitting still about to start talking, which reads as a video that has not begun yet.

• Start mid-action rather than before it. The hand is already reaching, the pour is already happening.
• Put your on-screen text up on frame one, not after a beat. Many viewers read before they listen.
• Avoid opening on a wide establishing shot. There is no room in two seconds to establish anything.
• Check the first frame as a still. If it would not stop you as a thumbnail, it will not stop anyone as a hook.

How fast the decision actually happens

Reported figures put the average continue-or-scroll decision at around 1.7 seconds. Not two seconds as a rule of thumb: less than two, before most viewers could tell you they had decided anything.

What follows from that is the number worth managing. Three-second retention, sometimes called hold rate or the inverse of skip rate, is the share of viewers still watching at the three second mark. Typical skip rates sit around 20 to 35 percent, and anything above 40 percent means the opening is not doing its job.

Why that particular number carries so much weight is visible in how the platform ranks: Instagram's ranking explainer describes reels distribution as driven largely by predicted completed watches and reshares. A viewer lost at second two contributes to neither, so hold rate is not one metric among several. It is the gate the others sit behind.

Five hook shapes that work

Published three-second retention benchmarks by hook type point in a consistent direction, and it is worth noticing which end of the table the familiar marketing openings sit at.

Social proof sitting last is the useful surprise. Opening on a testimonial or a rating feels persuasive and performs worst of the six, most likely because it reads as advertising immediately. The visual pattern interrupt performs best, which is also the only one on the list that survives being watched on mute with no captions.

The contradiction

State something that runs against what the viewer expects to hear. It works because a contradiction creates an open loop, and open loops are uncomfortable enough to hold attention until they close.

The specific claim

A precise, concrete statement about a result or a method. Specificity signals that there is real information coming, whereas a general claim signals an advertisement. Numbers help here, provided they are true.

The named audience

Address exactly who the reel is for in the first line. You lose everyone else instantly, which sounds bad and is usually good: a smaller, correctly-targeted audience watching to the end beats a large one leaving at second three.

The mistake

Open on the error the viewer is probably making. This works because it is about them rather than about you, and because the implied promise, that you will explain the fix, is one people want kept.

The visual pattern break

Something unexpected happens on screen with no verbal setup at all. Useful when the product itself does something visually interesting, and the only shape on this list that survives being watched on mute with no captions.

The promise problem

There is a way to write hooks that holds attention at second two and destroys the reel at second six. It is the hook that promises more than the video delivers. Retention graphs show it clearly: a strong start followed by a cliff.

This matters beyond a single reel. An audience that learns your openings oversell stops trusting them, and the hooks get progressively less effective on the same people. Whatever the hook implies has to arrive, and ideally arrive quickly.

How to actually write them

• Write the hook last, once you know what the reel genuinely delivers. Writing it first tends to produce a promise the video then has to chase.
• Write ten and throw away eight. The first three are almost always the obvious ones.
• Read them aloud. A line that reads well and sounds stiff is a line nobody will believe on camera.
• Cut every word before the interesting one. Most hooks improve by deleting their first four words.
• Test hooks against each other rather than debating them. Same reel, two openings, let the retention decide.

That last one is the only reliable method, and it is why volume matters so much in short-form. Opinions about hooks are cheap. Retention data is not, and you only get it by publishing. More on that in why consistency beats one viral reel.

Why the hook decides distribution, not just attention

The first two seconds are usually discussed as an attention problem. They are really a distribution problem, which is why the stakes are higher than they look.

Instagram's own ranking explainer describes reels distribution as driven largely by how likely a viewer is to watch through and to reshare. Both of those are decided at the opening: a reel abandoned in the first two seconds cannot be completed and will not be sent to anyone. The hook does not merely win the view in front of it, it determines whether the platform offers the reel to the next thousand people.

That is the mechanism behind the gap in performance between a strong and a weak opening. It is not that a good hook is somewhat better; it is that a weak one removes the reel from circulation before the rest of the video is ever assessed. How long a reel should be covers the completion side of the same signal, and why your reels stopped getting views covers what it looks like when this goes wrong at account level.

Testing hooks without remaking the reel

The practical advantage of treating the hook as a separate component is that it can be tested on its own. Keep the body identical, change only the first two seconds, and publish three versions across three weeks. Any difference in performance is attributable to the opening, because nothing else moved. That is a real experiment, and it costs one edit rather than three productions.

Doing that monthly builds a list of openings that work for your specific audience, which is worth more than any general advice about hooks including this article. If the constraint is producing the variants rather than writing them, that is what our social media reels service handles.