AppCut StudioHow many creatives does app user acquisition need each month?
Concepts and variants are different things
A concept is a new idea: a different story, a different first scene, a different reason to install. A variant is the same idea with a new hook, size, language or end card. Networks learn fast from variants, but they only find new audiences when you feed them new concepts.
Mixing the two up is common. A team that ships 40 resizes of one idea has made 40 files and one concept.
How many to make
The table is a planning model, not a benchmark. Adjust the variant count to the number of networks and languages you run.
| Monthly budget level | New concepts | Variants per concept | Files per month |
|---|---|---|---|
| Small test | 4–6 | 4–6 | 20–35 |
| Steady UA | 10–15 | 4–6 | 40–90 |
| Aggressive scaling | 20+ | 6–10 | 120+ |
Where the hours go
- Resizing to every network format and checking safe zones.
- Replacing on-screen UI and prices for each language.
- Cutting new hooks onto proven bodies.
- Naming and exporting files so the media buyer can upload without asking questions.
Making 60 variants a month with a small team
Split the work. Automate what is mechanical: resizing, subtitle timing, translation drafts. Keep people on what decides performance: the first two seconds, the concept and the final check. A monthly plan with fixed limits for formats, languages and revisions keeps the volume predictable for both sides.
A brief that keeps variants cheap
Most rework comes from missing information. Send the master with its project files, a list of networks and sizes, the languages, any text that must stay identical in every market, and the store screenshots for each locale. With that in one message, a batch of 15 variants takes days, and nobody has to ask what the price on the end card should be in Ukraine.
Signs you need more concepts, not more variants
- Cost per install rises while click-through holds.
- New variants of the winning ad stop beating the original.
- The same two ads take most of the spend for weeks.
Start with a pilot pack
Send one master and your networks. You'll get up to 15 variants for $600, or one free adaptation to judge the quality.
Send a test briefFrequently asked questions
Should playables count toward the creative plan?
Yes, as their own track. They take longer to make and are tested separately from video.
Is AI-generated UGC worth testing for apps?
As an extra hook source, yes. Keep the app screens real and label AI content where the network asks for it.