Hook rate describes how often a video ad holds attention through its opening. One defined version divides three-second video plays by impressions and multiplies by 100. The viewing threshold and denominator can vary, so label the formula and compare only results measured the same way. A strong opening is a useful signal, not proof of sales.
- Write the viewing threshold and denominator beside every percentage.
- Compare the same platform, placement and reporting window.
- Read early attention alongside clicks, conversions and delivery conditions.
What does hook rate actually measure?
A hook is the opening idea that gives someone a reason to keep watching. It might be an immediately recognizable product, a specific problem or an unexpected visual. Hook rate turns a chosen opening-view count into a proportion. It is a derived attention measure, rather than one universal standard shared by every advertising platform.
The phrase alone is incomplete. “Our hook rate is 24%” leaves out how long a view must last, what it is divided by and where the ad ran. “Three-second plays per impression, Instagram Stories, November 1–7” is much easier to interpret. Keep the platform's original column names in the export so another person can reproduce the calculation.
A video can also attract attention for the wrong reason. An intriguing opening that has little connection to the offer may retain viewers who never become customers. Use the metric to investigate the opening, then check whether the remainder of the ad fulfills its promise.
How do you calculate hook rate?
Opening-view rate = qualifying opening views ÷ stated denominator × 100. The following numbers are hypothetical teaching examples. They are not AKOOL campaign results or performance targets.
Suppose an ad receives 10,000 impressions, 8,000 video starts and 2,400 three-second plays in the same report. Using impressions, the calculation is 2,400 ÷ 10,000 × 100 = 24% three-second plays per impression. Using starts, it becomes 2,400 ÷ 8,000 × 100 = 30% three-second plays per video start. Both calculations are arithmetically correct, but they answer different questions.
The impression-based version connects opening views to recorded delivery. The start-based version asks how often a recorded start reaches the threshold. Do not switch denominators between reporting periods and interpret the change as improved creative. If the denominator is zero, report the rate as unavailable rather than dividing by zero or writing 0%.
For a group of ads, add the qualifying counts and denominators first, then calculate the combined rate. An unweighted average of percentages gives a tiny ad the same influence as a heavily delivered ad. Keep separate rows when the underlying definitions or placements differ.
Why do Meta and TikTok opening metrics differ?
Meta and TikTok expose their own video metrics. A custom calculation using Meta's three-second-play column is not equivalent to one using TikTok's two-second-view column. The time threshold differs before you even consider placement, autoplay behavior or audience.
Use Meta's current video-metric help and the column tooltip in your account to confirm the applicable definition, including treatment of short videos and replays. Meta's help may require login. Do not replace an unavailable definition with an agency's shorthand.
TikTok's video-play reference describes its two-second view as a video playing for at least two seconds within an impression session. Its other viewing measures have their own conditions. In particular, do not assume every six-second or focused-view metric is a pure six-second retention count; read the selected column's full definition.
For cross-platform reporting, show the original metrics in separate rows with definitions. A larger number on one platform does not establish a better opening when the thresholds differ. You can still compare each platform's creative against its own consistent baseline.
How is hook rate different from hold rate and CTR?
- Opening attention: qualifying opening views ÷ impressions, or another explicitly chosen denominator. It diagnoses the first part of the video; it does not show whether the message persuaded anyone.
- Hold: choose and name a later viewing threshold and its denominator. For example, completed views ÷ video starts describes completion among starts. A completion rate for a ten-second ad is not directly comparable with one for a sixty-second ad.
- Link CTR: link clicks ÷ impressions × 100, when those are the selected platform columns. This describes clicking, not watching or purchasing. Keep link clicks separate from all clicks.
- Conversion outcome: use the platform or analytics system's defined conversion event and attribution window. A purchase and a landing-page visit are different outcomes.
These measures can move in different directions. An opening may increase viewing while attracting less qualified interest. Another may make the product obvious, lose casual viewers and still produce more useful visits. Decide the campaign objective before choosing which tradeoff matters.
What is a good hook rate?
There is no dependable universal threshold for every account and video. Start with comparable ads in the same account, using the same metric definition, placement, objective and reporting period. Note audience composition, video length and meaningful changes in delivery.
Build a baseline from enough delivery to be useful for your decision. A rate based on a handful of impressions is unstable. For example, one additional qualifying view changes a ten-impression sample by ten percentage points. Report the underlying counts beside the percentage so that uncertainty is visible.
Avoid presenting an observed difference as statistical significance unless the experiment design and analysis support that claim. Platform testing tools can help structure comparisons, but a screenshot of two percentages does not establish causation.
What should you check when an opening is weak?
- Slow reveal: can someone identify the subject in the first frame, or must they wait through an introduction? Try removing a nonessential setup shot.
- Unclear product: is the object recognizable at phone size? A beautiful wide shot may hide the detail that explains the offer.
- Unreadable text: does the opening ask someone to read several lines while the scene moves? Reduce the message and check its contrast and placement.
- Broken promise: does the body answer the question or demonstrate the claim introduced at the start? Remove an unrelated curiosity hook.
- Delivery differences: did the audience, placement mix, frequency or objective change? Inspect those conditions before assigning the entire difference to creative.
A useful revision follows the observed problem. If the product is hidden, show it earlier. If the text is hard to read, simplify it. Adding faster cuts to every weak opening may create more distraction without clarifying the message.
How can you plan opening variations with AKOOL?
Use AKOOL to create visual alternatives, then measure them in the advertising platform. AKOOL does not calculate your campaign's hook rate or automatically identify a winning audience. The following is an illustrative test plan, not a completed campaign experiment.
Keep an approved product reference, body, offer and CTA consistent. Plan three openings: an immediate product close-up, a problem-led scene and a restrained motion reveal. Change the opening idea deliberately rather than changing the entire advertisement.
For a photo-based asset, follow the Image to Video help workflow: upload the reference, choose an available model and describe the intended movement. Model settings and credit requirements vary. Record the model version, source image, prompt and settings used for each attempt.
An illustrative, untested prompt is: “Use the supplied bottle photo. Begin with the product clearly visible. Make one gentle camera move closer against a neutral background. Keep the product shape and label close to the reference. No generated text.” Inspect the actual result for changed packaging or invented features. Add exact claims and lettering as controlled overlays in an editor.
The product-photo video-ad guide covers the asset workflow in more detail. Keep this measurement exercise focused on the opening and its stated hypothesis.
What belongs in a useful test log?
- Variant and hypothesis: product close-up, problem scene or motion reveal; explain what each opening is intended to clarify.
- Definition: platform, column names, viewing threshold, denominator and calculation.
- Scope: date range, placement, objective, audience setup and any delivery differences.
- Counts: impressions, starts, qualifying views, link clicks and the defined conversion event.
- Production record: model, prompt, source asset and edited opening file.
- Decision: retain, revise or gather more data, with the reason and unresolved limitations.
Use AKOOL's video generator to explore a specific opening idea. Review the generated asset before building the ad, then let consistent campaign measurements determine the next revision. A useful hook connects the first frame to the actual offer.
