Industry
The cheapest install is often the fastest cancellation
Two fitness ads can buy installs at the same price and build completely different businesses. The hook decides which cohort you bought. Here is why cost per install ranks the wrong creative first, which hooks survive platform review, how one take becomes a week of them, and what twenty honest hooks cost.
- Where a mismatched promise shows up
- day 30
- One finished thirty second ad
- 110 CR
- Longest source take accepted
- 3 min
An install is not a customer, and the hook decides which one you bought
Two ads can produce the same cost per install and completely different businesses.
One promises a body in twelve weeks. The other promises twenty minutes in a living room before the school run.
The first is cheaper on day one. It sells a fantasy, and fantasies convert.
It is also the reason the day thirty cohort chart falls off a cliff.
The second attracts fewer installs at a higher cost. A meaningful share of those people are still opening the app in March.
Optimising creative on install cost alone systematically buys the first kind.
That is not a media buying error. It is a measurement window that ends before the truth arrives.
So the useful creative question is not which hook converts.
It is which hook describes the product accurately enough that the person stays.
- Cost per install ranks the fantasy hook first and the accurate one second
- A promise the app cannot keep is a refund request with a two week delay
- Retention is set by expectation, and expectation is set in the first three seconds of an ad
Transformation creative meets policy before it meets results
Platforms restrict health and body imagery. Fitness is where that restriction bites hardest.
Before-and-after pairs sit near a line. So do body-part close-ups.
So does copy implying a negative view of the viewer's own body.
Ad accounts in this category get restricted for creative choices far more often than for targeting choices.
That is rarely how the internal post-mortem describes it.
The workable alternative is behavioural. What changed about somebody's Tuesday. What they stopped doing. What time they now get up.
None of that is a claim about a body. It is also more specific, which is the part that makes it perform.
I stopped skipping Mondays beats a number of pounds, and it survives review.
Write the restriction into the brief rather than discovering it at rejection. Rejections here cluster around exactly the creative the team was most excited about.
This is the whole editor
Highlight a phrase and a clip lands on those exact words. No timeline, no keyframes, no layers.
The ad is about the week, and this is the fastest way to cut a week into an ad
An interface tour is a poor cold ad anyway. Nobody installs a fitness app because a menu looked tidy.
So the cold creative has to be about the life the app fits into.
Upload one talking-head take, up to three minutes. Cutroom transcribes the read and drafts the cut.
It places b-roll against the words and returns a finished 9:16 MP4 with captions burned in.
You direct on the transcript. Highlight a phrase and a clip lands over exactly those words.
Delete a line that overpromised and the cut rebuilds. Change the pace and it re-cuts the whole piece.
App screens belong in the ad too. Record the walkthrough with any screen recorder, upload it to your workspace library, and swap it into the slot where the read names the feature.
That keeps interface footage where it works, in the second half and in retargeting.
Keep the cold opening on the living room floor at six in the morning, and on the fact that nothing has to be booked.
A real week into a cold ad
Pick one changed habit
Not a body claim
Film the read
One take, up to 3 minutes
Batch
100 CR, transcript and first cut
Export
9:16 MP4, 20 CR per output minute
Rotating twenty honest hooks beats repeating one dishonest one
A finished thirty second ad is about 110 credits. Basic is $39.99 for 2,500 credits, roughly 22 ads.
Premium is $79.99 for 5,000, roughly 45.
Published benchmarks put the share of creatives that carry spend at roughly 5 to 8 percent.
That comes from Motion's analysis of 550,000+ Meta ads. Twenty attempts is where a category like this becomes measurable.
Twenty honest hooks at 110 credits is about 2,200 credits. One month of Premium, with room to spare.
The behaviour that changes is who you film. Coaches, members, a founder who trains at 6am.
All of them can carry a specific week. None of them needs a crew or a booking.
The trial is 7 days and 300 credits with no card, which is two finished ads.
Run those two against your current best performer and the answer arrives inside a week.
Finished ads a month, by plan
- Trial, 300 credits2 ads
- Basic, $39.9922 ads
- Premium, $79.9945 ads
- Each $15 top-up9 more
What the phone films, what the library holds, and what lands in the ad account
The b-roll search covers people running, mats on floors and kitchens at dawn.
That carries the week the read is describing, which is what a cold ad needs.
Anything specific to your app comes from you. Record it, upload it to the library, and it is available to every ad afterwards.
Highlight the phrase and the clip lands there. No trip to another tool and no timeline.
Exports are 9:16 MP4 with captions burned in. That covers Reels, TikTok, Shorts and Stories.
Source takes cap at three minutes and one project makes one finished ad.
Brand presence is carried by the caption treatment across six packs, plus colour, weight, size and position.
Somebody has to speak on camera. A member describing a changed Tuesday is the strongest asset in this category.
It is also the one thing a generated presenter cannot supply, which is why the format keeps beating produced spots.
Questions people ask
- Can it show the app interface?
- Yes, through your library. Record the walkthrough with any screen recorder, upload the clip, then highlight the words where you describe the feature and it lands there. There is no screen capture built into the tool itself.
- What hooks work when body claims are restricted?
- Behavioural ones. What somebody stopped doing, what time they now train, what their Tuesday looks like. Specific behaviour is safer under policy and more convincing than a number.
- How many creatives does an install campaign need?
- More than most teams run. At roughly 5 to 8 percent winner rate, twenty attempts is where the numbers mean something, and twenty at about 110 credits is around 2,200 credits.
- Who should not buy this?
- Teams whose ads are interface demonstrations with nobody speaking over them. For the cold half of the funnel, twenty spoken hooks a month is what this makes and nothing else does at the price.
One member describing a changed Tuesday becomes three tested hooks by Friday, for about 330 credits. Compare them on day thirty and the accurate one is still there.