Hook Rate vs Hold Rate: What Each One Tells You About a Creative

· Creative

Hook Rate vs Hold Rate: What Each One Tells You About a Creative

The video with the highest hook rate in one of our nicotine alternative accounts stopped 62% of the people it was served to over the last 90 days. That is more than double the 26% median across the account's top-spending videos. It came in at a $143 CPA, against a $98 median. The lowest CPA in the same set belonged to a video with a 12% hook rate, at $73 per purchase.

Teams read hook rate as a creative quality score, and that habit sends a lot of creative reviews sideways. Hook rate grades the opening of an ad while hold rate grades the body, so neither one predicts sales on its own. Read together, the pair shows which part of a creative is broken and what to change. That makes it far more useful than a single number on a leaderboard.

What hook rate and hold rate each measure

Hook rate is 3-second video plays divided by impressions. It measures the share of people who stopped scrolling long enough to register a view. In practice it grades the opening frame, the first line of audio and the on-screen text over them, with very little else in play.

Hold rate is ThruPlays divided by 3-second video plays. A ThruPlay is a view of 15 seconds, or the full video if it runs shorter. Some teams divide by impressions instead. We divide by 3-second plays, which counts only people the hook already caught and isolates the body from the opening.

That separation is what makes the pair useful. The hook has to earn attention while the body has to justify it, so an ad can fail at one job and still do the other well.

The top half by hook rate converted worse

We took every video in that account that spent more than $45k over the last 90 days. That gave us 24 ads with $1.89M in combined spend, which we split at the median hook rate.

Metric

Bottom half by hook rate

Top half by hook rate

Median hook rate

16.7%

37.4%

Median CTR

1.24%

1.18%

Median CPA

$92

$109

The top half more than doubled the share of people who stopped. None of that extra attention carried through to the click, with median CTR flat at 1.18% against 1.24%. Median CPA came in 19% worse.

The 62% ad at the top of the list sharpens the picture. It drew a 2.44% CTR, about double the top-half median, and still converted at $143. The attention survived the click and stalled before the purchase.

This is a single account, so treat it as a pattern to check against your own data. The underlying point holds either way, because a hook rate tells you how many people you caught and nothing about whether they were the right ones.

Hold rate tells you whether the body earns the hook

Hold rate is the closest read you have on whether the message lands with the people the hook brought in. A low hold rate behind a strong hook usually means the opening promised something the next 10 seconds did not deliver. The classic case is a pattern-interrupt visual or a shock line with no real link to the product. People stop, work out what they are looking at, then scroll on.

A strong hold rate behind a weak hook is the more valuable finding. It means the ad works for anyone who gets past the first 3 seconds, with the opening holding it back. Keep the concept and test 3 or 4 new openings on the same body, which is the lowest-lift creative test you can run.

Hold rate has its own blind spot, and it shows up most in yapper ads (talking-head videos where one person speaks straight to camera). A good yapper can keep people watching to the end and still fail to sell if the product shows up late or the CTA is soft. A high hold rate earns an ad a closer look at CTR and CPA before it earns more budget.

Reading the two together gives you 4 diagnoses

High and low here mean relative to your own account median. Industry hook rate benchmarks vary too much by vertical and placement to be worth much. Once you have both numbers for a video, it falls into one of 4 buckets.

  • High hook, high hold, weak CPA. The creative is doing its job, so the problem sits after the click or with who the ad reaches. Check landing page conversion rate and the offer before you touch the ad.

  • High hook, low hold. The opening is borrowing attention the body cannot keep. Tie the hook closer to the product or rebuild the first 10 seconds. Do not scale an ad like this on hook rate alone.

  • Low hook, high hold. This is the best iteration candidate in the account. Keep the body exactly as it is and test 3 or 4 new openings on it.

  • Low hook, low hold. Turn it off unless the CPA is strong. A strong CPA here usually means the ad has found a small pocket of high-intent buyers, which makes CPMr the next number to check.

The last bucket is where the wrong call is easiest to make in either direction. Teams kill quiet ads that were doing real work, or keep one alive on a good ROAS while it serves the same warm audience on repeat.

CPMr tells you who is actually seeing the ad

Both metrics measure behavior among the people who were served the ad. Neither says whether those people are new. That gap matters, because hook rate can hold steady or climb while an ad's reach shrinks. A warm audience that already knows the brand can stop for it more readily than a cold one.

This is why we pull CPMr into every creative review. CPMr is the cost to reach 1,000 unique people, calculated as CPM × frequency, and Phil Kiel, our Director of Paid Social, popularized it as a leading indicator of saturation. If a high-hook ad shows CPMr climbing week over week, check whether frequency or CPM is driving it.

Rising frequency with flat CPM means the ad is cycling through the same people. In that case the strong hook rate is partly a product of familiarity, which calls for tighter exclusions and fewer overlapping ad sets. Rising CPM with flat frequency points to auction pressure instead, a separate problem with a separate fix.

Meta's Andromeda system uses the creative itself as a targeting signal, so the opening of an ad also shapes who Meta shows it to. A hook built around the product's real problem pulls in people who have that problem. Openings built purely for the stop can widen reach in a direction that never converts.

Pull hook rate and hold rate for every video that spent in the last 30 days. Put them next to CPA and CPMr, then sort each one into a bucket. Brief new openings this week for the low-hook, high-hold ads. Stop judging the high-hook, low-hold ads by their hook rate.

Data from a single nicotine alternative account: 24 video ads that each spent more than $45k, $1.89M combined, over the 90 days to October 6, 2026. Some ads are relaunches of the same creative in scaling campaigns. All data anonymized.


Back to all articles