In measured numbers
Every figure on this page is read from a code path, not written by hand — the same discipline the product applies to your own data.
What the model is given
The prompt is assembled from your account, not from a category template. Nothing in it is invented:
- Your current rank for every tracked keyword, per storefront, with its measurement state.
- The apps ranked above you and their rating counts — the wall you would have to climb.
- The demand signal for each term from store autocomplete.
- Your current title, subtitle and keyword field, with their remaining character budget.
- Discovery candidates, each still carrying the source it came from.
How it works
- You ask for a plan for one app and one storefront.
- The measurements are packaged and sent with the request. Terms whose state is not measured are sent as unknown, not as zero.
- The suggestion comes back as a list you can act on — proposed keywords, a proposed field within 100 characters, and a reason per line.
- You decide. Nothing is written to your store listing. The planner produces text; shipping it is a separate, manual act in App Store Connect or Play Console.
What you do with it
Fit a field
A hundred characters is a packing problem. A shortlist ordered by measured opportunity beats a blank box.
Learn what fits in 100 charactersCut the unreachable
Terms whose ratings wall is out of reach get dropped before you spend a release on them.
Measure the wall above youPlan per storefront
The same app in another storefront is another measurement, so it gets another plan.
Generate candidates firstWhat this does not do
The model ranks evidence. It does not supply it.
Every input in the prompt is a measurement from your own account. What comes back is a shortlist you act on, not a forecast:
- It does not write to your store listing. There is no path from this screen to your published metadata, by design.
- It does not predict rank. No model can tell you what position a word will reach; the honest output is a shortlist.
- It does not replace the measurement. After you ship, the answer comes from the next run — not from the model that proposed it.
- It does not invent demand. A term the stores never suggest stays a term the stores never suggest.
What each plan includes
| Free | Pro | Studio | |
|---|---|---|---|
| AI suggestions | — | — | Yes |
| Suggestion runs per day | — | — | 3 |
| Keywords the plan can draw on | 25 | 250 | 500 |
The planner is a Studio feature (60 $/month, 54 $/month billed yearly). The measurements it reads are collected on every plan, including the free one.
Frequently asked
Does the AI invent keywords out of nothing?
No. It is given your measured ranks, the rating counts of the apps above you, the autocomplete demand signal and your current metadata, and it chooses between those. Terms that could not be measured are passed as unknown rather than as zero.
Can it publish metadata to the App Store for me?
No, and that is deliberate. The planner produces text; shipping it stays a manual step in App Store Connect or Play Console.
Will it tell me what rank I'll reach?
No. No model can measure that. You find out from the next measurement run after you ship the change.
Which plan includes it?
Studio, at three suggestion runs per day. Pro and Free collect the same measurements but do not include the planner.
See this for your own app
Paste your store link, add your keywords. Rank, the apps above you and demand are measured every day.
Start free →