We posted six promo shorts for our roulette app (3 app screen recordings, 3 Chichi animations) on YouTube, TikTok and Instagram and measured for seven days. Views were 681 for the screen recordings and 588 for the animations, and installs confirmed through the store link or UTM were 0 for both. No format won. The only thing the numbers settled is that short form video was not an install channel for this app.
What We Tested
We released our spinning wheel app, Chichi's Decision Roulette, on Play on October 17. You enter your options and spin, and Chichi, the cat boss, announces the result. It is a simple tool. Short form video was the only way we had to promote it, and since we did not know which style would work, we split it into two.
A is a recording of the real app with a question caption like "What should we eat for lunch?" over the first 3 seconds. B is an animation of Chichi spinning the wheel, and only the last 2 seconds show the app screen. Each pair has the same topic and the same result, and both videos went up on the same day at the same time.
For three days starting October 22, we posted one pair at 7 PM each evening on YouTube Shorts, TikTok and Instagram Reels, with three topics: what to eat for lunch, who pays today, and picking a dare. The other two pairs we had made did not fit the account concepts, so we did not post them.
The Numbers After Seven Days
We made the call on November 2. Views came from the YouTube Data API, Instagram Insights and the public play counts on TikTok. Installs were counted from the UTM campaign table in Play Console and the first launch records in the app's GA.
| A app screen recording | B Chichi animation | |
|---|---|---|
| Videos | 3 | 3 |
| Views, 7 day total | 681 | 588 |
| YouTube / TikTok / Instagram | 264 / 11 / 406 | 231 / 0 / 357 |
| Views per video | 227 | 196 |
| Installs confirmed by UTM and GA | 0 | 0 |
| Installs per 1,000 views | 0 | 0 |
Each Pair Had a Winner, but There Was No Overall Winner
In pair 1, A won 222 to 123. In pair 2, B won 270 to 219. In pair 3, A won 240 to 195. That makes A 2 wins and 1 loss. Our rule at the start was "4 or more of 5 pairs go the same way." We got only three pairs, and even those three did not lean to one side.
So the verdict is "no winner." However, B costs more to make because we build separate Chichi scenes and still include the A recording. When views are in the same range, the cheaper one should be the default, so from now on we post A.
We Looked into Why Installs Were 0
Over the seven days, the only Play listing visitors with a recorded source were 2 from "other," and nobody arrived under the UTM names we set up. In the same period, the app had 33 first launches, and most of them were 30 with no source (direct visits) and 3 from Play organic search.
There are limits too. The only places where a link could be clicked were the YouTube description and the pinned comment. Links in TikTok and Instagram captions are not clickable, so we left them out. Of the 1,269 views, the only ones who ever stood in front of a link were the 495 YouTube views. If Instagram and TikTok viewers went to the profile, searched the store and installed, they would have shown up in this table as "direct visits."
The conclusion is the same either way. There is no way to tell whether the 30 direct visits came from the videos, and every slot where we could tell (UTM) was 0. A short form view is a number that goes up each time playback starts (YouTube has counted views with no minimum watch time since March 31, 2025), so 1,269 does not mean 1,269 interested people.
What We Did Next
We closed the small app short form experiment on November 3. In the 17 days since release, the app had 41 installs, and 0 of them were confirmed to come from short form video. The sources were 36 direct visits and 5 Play organic. The lottery draw app and the truth or dare app we had planned to build alongside it were not made.
We are leaving the roulette app up. It costs nothing to keep, and the numbers confirmed that installs arrive through Play organic search. What we will work on next is not videos but the store name and keywords. When we checked open search terms, there were no giant apps on terms like Indonesia's "dadu," Brazil's "sorteio de times" and Japan's "コイントス," and adding these as modes inside the roulette (dice, team splitting, coin toss) is cheaper than building a new app.
What We Will Keep to Next Time We Post Short Form Video
The numbers did not tell us much, but one thing is clear. Views and installs are separate ledgers, so if views go up and the install column stays empty, that video only made a number.
- To see installs, first build a place where a link can be clicked. Do not pin hopes on captions that cannot be clicked
- Write down installs per 1,000 views for every video. The benchmark we had collected earlier was 0.5 to 5, and this time we were below it
- If there is no winner, do not compare more formats. Go with the cheaper one
- A small app's install path is store search. Fix the name and keywords before the videos
FAQ
Do more short form views lead to more app installs?
Not in this experiment. Six roulette app shorts got 1,269 views over seven days, but installs from short form video were 0 in the UTM links, Play Console and GA. Views and installs have to be measured separately.
How do you measure how many app installs short form video brings?
Add UTM parameters (utm_source, utm_campaign) to the link in the description or pinned comment, then look for the same names in the UTM campaign table in Play Console and in the first user source in GA. TikTok and Instagram, where caption links cannot be clicked, cannot be measured this way.
Which is better, app screen recording or character animation?
Views were slightly higher for the screen recording at 227 per video against 196 for the animation, but the animation won one of the three pairs. Installs were 0 for both, so we could not separate them, and we set the cheaper screen recording as the default.
Does this mean short form video is useless for app promotion?
That would be going too far. This is the result of one app, six videos and seven days, and YouTube was the only place where a link could be clicked. All we can say is that for this app it did not lead to installs.
More from the blog
ComparisonOnsen and bathhouse games: 8 compared by what the store pages sayOnsen and bathhouse games from Steam, Google Play and the App Store in one table: price, rating, idle or not, Korean and Japanese support, plus the weak spots of our own game.
ComparisonIdle tycoon games that work offline: 8 picks sorted by store labelsWhich idle tycoon games work offline? Google Play tag, store wording, list price and ads in one table, plus our own game's weak spots.
Games likeFinished Kairosoft? 14 games to play next, sorted by what you lovedIf you have played through Kairosoft, here are 14 games sorted into four flavors: short grow-it sessions, running a shop, watching residents live, and layout puzzles. Steam list prices, review scores and UI languages checked by hand.
ComparisonInternet Cafe Simulator Games in 2026: 17 Picks, iPhone IncludedLooking for an internet cafe simulator on iPhone, Android or Steam? We compared 17 games using prices, ratings and offline notes read straight off the store pages, plus our own game's weak spots.