Hook Rate Calculator

The Hook Rate Formula

40,000 impressions and 10,400 three-second plays gives a hook rate of 26.0%. Roughly a quarter of the people you paid to reach stopped long enough to see what the ad was about. The other three quarters never got past the first frame.

Impressions go in the denominator because that's what Meta bills you for. Hook rate is the share of paid impressions that turned into actual attention, which is why it's also called thumbstop rate. Same formula, different name.

Don't compare a TikTok rate with a Meta one. TikTok counts a view at 2 seconds, Meta at 3, so the same video will usually post a higher rate on TikTok. Benchmark each platform against itself.

Adding Hook Rate to Meta Ads Manager

Ads Manager doesn't show hook rate by default. You build it once as a custom metric and it's there in every report after that:

1. In Ads Manager, open Columns and choose Customize columns.
2. Click Create custom metric and name it Hook Rate.
3. Set the formula to 3-second video plays ÷ Impressions and the format to percentage.
4. Add a second metric for hold rate: ThruPlays ÷ 3-second video plays.
5. Save the column preset, and read both at the ad level, not the campaign level.

“Didn't Meta get rid of 3-second views?” Not for ads. What Meta retired in June 2026 were the unique 3-second viewer counts in Page and post Insights, the organic side, as part of Graph API v25. We checked Meta's own ads reporting fields in November 2026 and 3-second video plays is still there, defined as plays of at least 3 seconds per impression, excluding replays. If a reporting tool has stopped showing it, check it's pulling from the ads API and not Page Insights.

What Is a Good Hook Rate in 2026?

The largest published dataset with a stated method is Billo's. Across 88,329 sales-objective Meta video ads with more than 1,000 impressions each, run between 1 January and 30 June 2026, the average hook rate was 25.44%, up from 24.42% in the second half of 2025.

Source: Billo, Video Ad Hook Rate Benchmarks H1 2026. Meta only, sales objective only. Billo doesn't publish a margin of error or a split by placement, so read these as reference points, not pass marks.

So, in practice:

• Under 20%: below every industry average. The opening isn't stopping anyone.
• Around 25%: average. Fine to run, worth testing against.
• 30% or more: ahead of every industry average in the panel, since none averaged above 29%.

Your category matters more than the headline figure. 24% is above average for a pet brand and below it for health and beauty. Pick your industry in mode 1 and the verdict adjusts. For the rest of the funnel, our Meta ads benchmarks and TikTok ads benchmarks cover CPM, CTR and ROAS.

Hook Rate vs Hold Rate

Hook rate tells you whether the opening worked. Hold rate tells you whether the rest of the video kept the people you hooked.

A ThruPlay is a play to the end, or to 15 seconds on longer videos. With 10,400 three-second plays and 2,288 ThruPlays, hold rate is 22.0%.

Here's the catch. Some teams calculate hold rate as 15-second views ÷ 3-second plays, and some divide by impressions instead. Those give very different numbers for the same ad, so a “20% hold rate” from someone else's deck means nothing until you know the denominator. That's also why we don't show a hold rate benchmark: we haven't found a large published panel that states its method. Compare hold rate against your own other ads.

The two read together like this:

• Low hook, any hold: fix the first 1 to 2 seconds. Nothing later matters if nobody sees it.
• High hook, low hold: the opening over-promises or the middle drags. People came for one thing and got another.
• High hook, high hold, few clicks: the content works and the ask doesn't. Look at the offer and the call to action.

Why a High Hook Rate Isn't a Profitable Ad

Hook rate is an attention metric, and attention isn't revenue. Billo's own H1 2026 data makes the point: Toys & Games had the best hook rate of all fourteen categories and the third-worst ROAS. Some of the highest-hooking health and beauty ads barely broke even.

The cleanest way to see where an ad leaks is to split click-through rate into two parts. This split is exact:

Say an ad gets 60,000 impressions, 13,200 three-second plays and 780 link clicks. CTR is 1.30%, which on its own just says “below average”. Split it and it's 22.0% hook × 5.9% of hooked viewers clicking. The hook is a bit under the 25.44% average. The click-through among people who watched is well under the roughly 7.6% you get by dividing Billo's average CTR (1.93%) by its average hook rate. That's a rough guide, since an average of ratios isn't a ratio of averages, but it's enough to point you at the right half of the ad. Here, it's the body and the call to action.

Mode 2 does that split for you and names the weaker stage.

How Long to Test a Hook

Hook rate reaches a reliable answer far sooner than anything you actually get paid for, because 3-second plays are hundreds of times more common than purchases.

Two ads, 10,000 impressions each. One hooks at 26%, the other at 28%. That two-point gap is already 99.9% likely to be real. Now say the same ads drove 30 and 40 purchases. That looks like a big win for the second ad, but the confidence is only 76.9%, and you'd need around 55,000 impressions per ad to confirm a gap that size.

So use hook rate to kill weak openings early and cheaply. Use purchases, given enough volume, to pick the ad you scale. Mode 3 runs both tests side by side and tells you when they disagree.

One honest caveat: these tests treat every impression as independent, and they aren't, because the same person can see an ad several times. Treat the confidence figure as a best case. If frequency is high, check it with our ad frequency and fatigue estimator first.

How to Raise a Low Hook Rate

In our experience running DTC creative, the fixes that move hook rate are nearly all in the first second or two:

• Start mid-action. Cut the logo intro and the slow pan. Open on the product being used, the problem happening, or a face talking straight to camera.
• Put the hook in text on screen. A lot of feed video plays muted. A one-line overlay in the first frame does the job the voiceover can't.
• Make it look native. Creator-style footage usually out-hooks polished brand video in the feed, which is why we run so much of it. Our UGC ad examples show openings that are working right now.
• Test openings, not whole ads. Keep the body and swap only the first 3 seconds. That isolates the hook and makes mode 3's comparison fair.
• Call out the buyer. Naming the audience or the pain point (“If your dog won't stop scratching”) trades a little reach for a lot of relevance, and it tends to hold its click-through too.

Then check mode 4. A better hook is worth real money at the same spend, but only if the people it pulls in still click and buy.

What This Calculator Can't Tell You

• Why an ad hooks. It measures the result. The creative judgement is still yours.
• Hook rate by placement. Reels, Stories and Feed hook differently. Break the ad down by placement in Ads Manager and run each row separately if the mix changed.
• Whether the benchmark fits you. The averages come from sales-objective ads. Awareness and engagement campaigns get delivered to different people and hook differently.
• Incremental sales. Purchases here are what the platform reports. They include people who would have bought anyway.
• Hold rate benchmarks. There's no published panel we trust for it yet, so the tool compares hold rate only against your own ads.