Hook Rate for Dropshipping Video Ads: How to Read It and Fix a Weak One (2026)

A video ad that stalls at a $30 CPA against a $14 target gets killed for the same reason nine times out of ten in most dropshipping ad accounts: nobody looked past CPA to see which part of the video actually broke. Hook rate, hold rate, and click-through rate are three separate numbers measuring three separate failures, and a creative that dies on all three needs a different fix than one that only dies on the first.

Hook rate tells you whether the first three seconds stopped the scroll. Hold rate tells you whether anyone who stopped kept watching. CTR tells you whether watching turned into a click. Collapse all three into "the CPA is bad" and you end up rewriting a hook that was already working, or scrapping a whole concept when the real problem was a CTA buried at second 22 of a 30-second video.

What follows: what each metric actually measures and where to find it in Meta and TikTok's ad managers, honest benchmark ranges for each one, a diagnostic grid that matches the pattern across all three to what's actually broken, and how to fix a weak hook without touching a video that's already converting past it.

What hook rate, hold rate, and CTR each actually measure

Hook rate is the share of people who saw the ad and stayed past the first few seconds, usually reported as 3-second video views (or ThruPlay, on Meta) divided by impressions. It measures exactly one thing: did the opening frame and first line earn a moment of attention. Nothing past that first beat affects it.

Hold rate — sometimes shown as average watch time or a retention curve, on Meta and TikTok both — measures what happens after the hook lands: does the middle of the video keep people watching, or does attention bleed out steadily until the CTA plays to an empty room. A video can have a great hook and a terrible hold rate, and the two numbers will show it plainly if you look at them separately.

CTR measures something different from either: did watching translate into a click on the link. A video can hook and hold attention well and still post a weak CTR, usually because the CTA is vague, arrives too late, or asks for a click the viewer doesn't yet feel enough urgency to make.

• Hook rate ≈ 3-second video views (or ThruPlay) ÷ impressions — measures the opening 2-3 seconds only
• Hold rate ≈ average watch time ÷ video length, or the retention curve shown in the ads manager — measures the body of the video
• CTR ≈ link clicks ÷ impressions (or ÷ 3-second views, for a cleaner read of post-hook conversion) — measures the CTA and offer
• CPA folds all three together with landing page and offer performance on top, which is exactly why it can't tell you which one is broken on its own

Where to actually find these numbers in Meta and TikTok's ad managers

Meta doesn't surface hook rate as a named column by default — you build it from two metrics that are there: add a "3-Second Video Plays" or ThruPlay column next to Impressions, then divide. Some accounts have a video retention graph under the ad's preview pane that shows the drop-off curve directly, which is the faster read once you know where to click.

TikTok Ads Manager is more direct about it: the Video Views breakdown surfaces 2-second and 6-second view-through rates as near-equivalents to hook and early-hold rate, and the built-in retention graph under a running ad shows the same drop-off curve Meta buries a level deeper.

Neither platform hands you these numbers unprompted in the default reporting view — you have to go add the columns or open the specific breakdown. That's the single biggest reason most accounts default to reading CPA alone: it's the number that's already on screen, not necessarily the one that explains what's happening.

See the ads that are winning right now

Benchmark ranges: what counts as a good hook, and what counts as weak

Take every number below as a working range gathered from what dropshipping media buyers commonly report, not a published platform standard — Meta and TikTok don't publish official hook-rate benchmarks, and what counts as strong shifts with format, placement, and how competitive the auction is that week. Use these to spot an obvious outlier, not to grade a creative to the decimal.

On Meta, a UGC-style video hitting 25-35% on 3-second view rate is a workable middle; past 40% is a genuinely strong hook worth studying and reusing the pattern from. Under 20% on a cold-audience placement is usually weak enough to be the first suspect when CPA is off target.

TikTok's native feed runs hotter across the board because attention there is already primed for short-form video — a 2-second view-through rate in the 45-55% range is closer to average, and anything comfortably above 60% is strong. That gap is the format, not your creative being worse on TikTok than on Meta.

• Meta 3-second view rate: under 20% weak, 25-35% workable, 40%+ strong
• TikTok 2-second view-through: under 40% weak, 45-55% average, 60%+ strong
• Hold rate (average watch time ÷ length): under 25% weak on either platform, 35-50% workable, 55%+ strong for anything past 15 seconds long
• CTR (link clicks ÷ impressions): under 1% weak for a cold dropshipping campaign, 1-2% workable, 2.5%+ strong — read this one against your own account average as much as any fixed number

The diagnostic grid: matching the pattern to what's actually broken

Read the three metrics together, not one at a time, and the fix becomes obvious instead of a guess. This is the part most sellers skip — they see a bad CPA, rewrite the hook because that's the easiest thing to touch, and never check whether the hook was the part that failed.

• Weak hook rate, whatever hold rate and CTR do: the first 2-3 seconds aren't earning attention. Fix the opening frame and the first line before touching anything else in the video — nothing downstream matters if nobody stays to see it
• Strong hook, weak hold rate: the opening worked and the middle lost people. Usually a pacing problem — too much setup before the payoff, a slow demo, or a claim that took too long to prove. Cut the body, don't rewrite the hook
• Strong hook, strong hold, weak CTR: attention was never the problem — the offer or CTA is. Move the CTA earlier, make it more specific, and the video is already doing its job
• Strong hook, strong hold, strong CTR, weak CPA: the ad is doing exactly what it should and the problem lives past the click — landing page, price, or offer, not the creative. Stop iterating the video and go look at the page it sends people to

Fixing a weak hook without touching what's already working

A weak hook rate almost always traces to one of three things: the first frame doesn't signal what the video is about fast enough, the opening line states a fact instead of a problem, or the pattern has simply burned out from being seen too many times by the same audience — frequency climbing past 3.5-4 in a week degrades hook rate on a video that scored well the week before, independent of the video itself.

Test hook variants as a separate line from body and CTA changes. Cutting three different opening 3 seconds onto the same body and CTA isolates exactly what the hook contributes, instead of confounding it with a body edit that happened in the same round of changes.

• First frame: does it show the product doing the thing, mid-action, in the first still frame someone would see if the video never played — a static product shot rarely earns the stop a mid-action frame does
• First line: lead with the problem or the surprising claim, not a description of the product itself — "I didn't believe this worked until" beats "this is a neck massager" every time hook rate is compared between the two
• Frequency check before a rewrite: if the same hook scored well two weeks ago and is weak now with frequency past 3.5-4, the concept isn't dead — it needs a new execution of the same angle, not a different angle entirely

Where this fits into a normal testing budget

None of this replaces the testing-floor math that decides when a creative has earned a real read in the first place — roughly 100-150 landing page clicks and a handful of purchases before any of these numbers are worth acting on. What it adds is a faster read once that floor is spent: instead of waiting for CPA to confirm a creative failed, the hook, hold, and CTR pattern shows which part failed while there's still budget left to fix that one part instead of killing the whole test.

Trackira's ad library shows which hook styles and pacing patterns are keeping ads running for months across Meta, TikTok, Pinterest, and Google Shopping on a given product, which is the fastest way to see what a genuinely strong hook rate looks like in your exact niche before you shoot your own version. Once a concept is validated, the same workspace carries a winning product through to a Shopify draft.

What is a good hook rate for a dropshipping video ad?

There's no official platform benchmark, but practitioner-reported ranges put a workable Meta 3-second view rate at 25-35% and a strong one above 40%; on TikTok's 2-second view-through, 45-55% is average and 60%+ is strong. Read your own number against your account's own history as much as any fixed range.

What's the difference between hook rate and CTR?

Hook rate measures whether the opening seconds stopped the scroll — it's driven entirely by the first frame and line. CTR measures whether watching the video turned into a click on the link, which depends on the offer and the call to action, not the hook. A video can score well on one and poorly on the other, and the fix for each is different.

How do I find hook rate in Meta Ads Manager?

Meta doesn't label it directly. Add a 3-Second Video Plays (or ThruPlay) column next to Impressions and divide one by the other, or open the retention graph under the ad's preview pane for a faster visual read of the same drop-off.

My hook rate is strong but CPA is still bad — what's wrong?

If hook rate, hold rate, and CTR are all healthy, the creative isn't the problem — the drop-off is happening after the click, on the landing page, the price, or the offer itself. Stop rewriting the video and audit what happens once someone lands on the page.