Hook Rate vs Hold Rate: What Each Metric Predicts (2026)
Hook rate measures the percentage of viewers who watch past the first 2 to 3 seconds of a video. Hold rate measures the percentage who stay through 50 or 100 percent of the duration. Hook rate predicts algorithmic reach: will the platform show this video to more people? Hold rate predicts engagement depth and conversion: will viewers who see it take action? Optimize hook rate first, because hold rate is irrelevant for viewers who scroll past.
Definitions: hook rate and hold rate
Measures initial attention capture. On TikTok, X is 2 seconds. On Meta, X is 3 seconds. Also called "2-second retention" or "3-second video view rate" in ad platforms.
Measures sustained engagement. Y is typically 50 percent (mid-point retention) or 100 percent (full-watch rate). Also called "retention rate," "average percentage viewed," or "completion rate" depending on the platform.
Note that the denominators differ. Hook rate uses impressions (total times shown). Hold rate uses total viewers (people who started watching). A video with 10,000 impressions and 3,500 viewers who watch past 2 seconds has a 35 percent hook rate. If 700 of those 3,500 viewers watch to 50 percent of the duration, the hold rate at 50 percent is 20 percent.
What each metric predicts
| Aspect | Hook rate predicts | Hold rate predicts |
|---|---|---|
| Reach | Initial distribution volume | Extended distribution push |
| Engagement | Surface-level interest | Depth of interest, intent |
| Conversion | Awareness (top of funnel) | Action (bottom of funnel) |
| Creative quality | Hook strength | Body and CTA strength |
| What it tells you | "Did they notice?" | "Did they care?" |
How algorithms use both metrics
Short-form video algorithms use a two-stage distribution model. The first stage is the hook gate. When a video is published, the platform shows it to a small initial audience (typically 200 to 500 viewers on TikTok). If the 2-second retention rate exceeds the platform's threshold, the video advances to broader distribution.
The second stage is the engagement gate. Once a video passes the hook gate, the algorithm evaluates average watch time, completion rate, shares, and saves. These deeper engagement signals determine how far the video is pushed: to thousands, tens of thousands, or millions of viewers.
This means hook rate is the gatekeeper and hold rate is the accelerator. A video with a strong hook rate but poor hold rate will enter broad distribution but stall quickly as the algorithm detects poor engagement. A video with a poor hook rate never reaches the second stage at all, regardless of how good the body content is.
2026 benchmarks for both metrics
| Metric | Below avg | Baseline | Good | Elite |
|---|---|---|---|---|
| Hook rate (TikTok, 2s) | < 25% | 30 to 35% | 35 to 40% | 40%+ |
| Hook rate (Meta, 3s) | < 20% | 25 to 30% | 30 to 40% | 40%+ |
| Hold rate at 50% (organic) | < 15% | 20 to 30% | 30 to 40% | 40%+ |
| Hold rate at 50% (paid) | < 10% | 15 to 20% | 20 to 30% | 30%+ |
The four quadrant framework
Every video falls into one of four quadrants based on its hook rate and hold rate combination. Each quadrant tells a different story about the creative.
The ideal. Strong hook stops the scroll and strong body delivers on the promise. This creative will receive maximum algorithmic distribution and convert well. Scale it.
Clickbait pattern. The hook promises something the body does not deliver. Good initial reach but poor engagement and conversions. Rework the body to match the hook's promise, or tone down the hook to set accurate expectations.
Hidden gem. The content is excellent but nobody sees it because the hook fails to stop the scroll. This is the highest-ROI optimization target: rewrite the hook and you unlock all the value already in the body.
Weak creative. Neither the hook nor the body is working. Do not iterate on this creative. Go back to concept development and start fresh.
What to optimize first
Always optimize hook rate first. The logic is sequential: hold rate improvements only benefit viewers who made it past the hook. If your hook rate is 15 percent, doubling your hold rate still means only 15 percent of your audience ever sees the improved body content. But doubling your hook rate from 15 to 30 percent means twice as many people see everything that follows.
The exception is the "high hook, low hold" quadrant, where the hook is already performing well but the body is causing drop-off. In this case, fix the body first because the hook is not the problem.
For AI-powered pre-production QA, Hooklayer's score_hook evaluates hook strength, and predict_virality evaluates the full script (hook plus body). Together, they cover both metrics before any filming occurs.
How to improve each metric
Improving hook rate
- *Lead with the payoff, not the setup. "This $12 serum outperformed my $80 one" instead of "So I have been trying a lot of serums lately."
- *Use proven psychological triggers: curiosity gap, negative bias, social proof, pattern interrupt.
- *Stack visual and verbal hooks. Text overlay plus spoken hook plus visual reveal.
- *Cut dead air. The first syllable should land within 200 milliseconds of the video start.
- *Pre-score hooks with Hooklayer score_hook before filming. Kill anything below 70.
Improving hold rate
- *Deliver on the hook promise within the first 10 seconds. Do not tease indefinitely.
- *Use pacing changes: speed shifts, energy transitions, and visual cuts to maintain novelty.
- *Open nested loops. Start answering one question, then briefly introduce a second question, then resolve both.
- *Match the script structure to the hook trigger. A curiosity gap hook should resolve with a specific answer, not a vague conclusion.
- *Keep videos tight. For most niches, 15 to 45 seconds hits the sweet spot for hold rate optimization.
Frequently asked questions
What is the difference between hook rate and hold rate?
Hook rate is the percentage of viewers who watch past the first 2 to 3 seconds of a video. Hold rate is the percentage who watch through a significant portion (typically 50 or 100 percent). Hook rate measures initial attention capture. Hold rate measures content delivery and viewer satisfaction.
Which matters more, hook rate or hold rate?
Both matter, but they operate in sequence. Without a strong hook rate, nobody sees the content. Without a strong hold rate, the content does not convert. For algorithmic reach, hook rate is the gatekeeper. For conversions and brand building, hold rate is the predictor.
Can you have a high hook rate and low hold rate?
Yes. This happens when the hook promises something the video does not deliver (clickbait pattern). The hook stops the scroll, but the content disappoints, causing rapid drop-off. Over time, this pattern damages your account because the algorithm learns that your content does not retain.
What is a good hold rate on TikTok?
For TikTok, a hold rate (measured at 50 percent of video duration) of 15 to 25 percent is typical for paid content. Organic content from followed creators often sees 25 to 40 percent. Full-watch rates (100 percent completion) for 15 to 30 second videos range from 10 to 20 percent on average.
How do algorithms use hook rate and hold rate?
TikTok primarily uses 2-second retention (hook rate) as the initial distribution signal. Videos that pass this threshold enter broader distribution, where average watch time and completion rate (hold rate) determine how far the algorithm pushes them. Instagram Reels and YouTube Shorts use similar two-stage logic.
Should I optimize hook rate or hold rate first?
Always optimize hook rate first. Hold rate improvements only matter if viewers get past the hook. A video with a 15 percent hook rate and 40 percent hold rate reaches fewer people than a video with a 35 percent hook rate and 20 percent hold rate. Fix the hook. Then fix the body.