Someone asks what your app's App Store ranking is, and the honest answer is a question back: which one? There are two, they are computed on different surfaces, and a number from one says nothing about the other. A position for one query is not a chart position, and a chart position tells you nothing about any query.
The short version, before the detail: a chart rank is a position on one of Apple's top lists, for one storefront, platform, category and chart type. A keyword rank is a position in the results for one query, in one storefront. Apple documents some of what drives the second and nothing about what drives the first. Everything else you will read about "the App Store ranking algorithm" is somebody's inference, and this post says so where it applies.
Two rankings, not one
| Chart rank | Keyword rank | |
|---|---|---|
| What it orders | Apps on Top Free, Top Paid or Top Grossing | Apps returned for one search query |
| What splits it | Storefront, platform, category, chart type | Storefront, the query |
| Apple documents its factors | No | Partly, on its App Store search page |
| Depth anyone can read | 100 positions per feed (measured) | 200 results per query (measured) |
| Where to read it here | /apps and /trending | /keyword-rank-checker |
The splits matter more than the depth. Fetched on 2026-09-09, Top Free in the United States led with ChatGPT; the same feed for the United Kingdom led with Lloyds Mobile Banking; the US Entertainment genre feed led with TikTok Pro. "#1 on the App Store" names one cell in a grid with more axes than most people quote.
What Apple says about search, in its own words
Apple's page on App Store search is the only first-party description of any ranking mechanism on the store, so it is worth quoting rather than paraphrasing. Fetched 2026-09-09; every quote below was checked against the raw markup.
On what the results are ordered by: "Search results are based on a number of factors, including text relevance (matches for your app's title, subtitle, keywords, and primary category), as well as user behavior (downloads, ratings and reviews, and more)."
On categories: "Your primary category and optional secondary category are indexed by our search algorithm."
On ratings: they "can influence how your app ranks in App Store search."
On the one field that does nothing: "promotional text doesn't affect your app's search ranking so it should not be used to display keywords."
That is the whole of it. Three text fields, two categories, and a behavioral clause whose last two words are "and more". Apple names downloads and ratings as inputs and declines to say how they are weighed, over what window, or against what else. The keyword field's own mechanical rules, which the same page sets out, are in the keyword field post.
What Apple says about the charts: nothing
There is no equivalent page for the charts. Apple's discovery page mentions them once, as a place customers look, alongside "handpicked collections and categories". Checked on 2026-08-20 for an earlier post, the obvious URLs, /app-store/charts/ and /app-store/top-charts/, return a "Page Not Found" body, and the App Store Connect help path answers HTTP 200 with the same body, which is why Apple pages here are verified by title and size rather than status.
The one place the charts appear as a mechanism in Apple's documentation is a prohibition. App Review Guideline 5.6.3, Discovery Fraud, forbids manipulating "charts, search, reviews, or referrals". Apple will tell you what happens if you game the chart. It does not tell you what the chart counts.
The ranking-factor lists are unsourced
Search for "App Store ranking factors" and the top results largely agree with each other. Between them they assert, as fact, that download velocity outweighs download volume, that retention and uninstall rate are ranking inputs, that update frequency is one, that an app with a keyword in its title ranks a specific percentage higher, and in one case that the algorithm is a product of relevance, velocity, conversion and retention.
None of that appears on Apple's page. Apple names downloads; it does not say velocity. Apple names ratings and reviews; it does not say retention, uninstalls or update cadence. The percentage figure has no stated method. That does not make every one of those claims false. It makes them undocumented, and this site does not assert what Apple does not, so none of them appear here as a factor.
The one Apple sentence that comes closest to strategy is the trade-off it states outright: popular terms "may drive a lot of traffic, but are highly competitive", while "less common terms drive lower traffic, but are less competitive". That describes competition, not a ranking rule.
What Apple exposes publicly, measured
The chart feeds stop at 100. Re-verified on 2026-09-09: topfreeapplications/limit=200 returned 100 entries; toppaidapplications/limit=200 returned 99, because Apple's paid list is short there rather than capped. The history behind /trending has asked for 100 since 2026-08-23, after a run of asking for 200 that changed nothing except implying it tracked 200 positions. Anything reporting a "top 200" from these feeds is reporting 100 apps and a hopeful parameter.
The search endpoint returns up to 200. A live search for "weather" in the US storefront with limit=200 returned 194 results on 2026-09-09, with AccuWeather at #3. The same query against the UK storefront returned 186, AccuWeather at #4. Both positions are real and neither is the other.
Rating counts are per storefront. That same AccuWeather listing carried 1,496,325 ratings in the US response and 97,985 in the UK one, the same shape the per-storefront ratings post recorded for ChatGPT: roughly 8.9 million in the US against 1.4 million in Germany. A rank and a rating count are both facts about one storefront.
What is not exposed, anywhere public: a download count for any app but your own, impressions, conversion rate, the chart formula, the search weighting, and any position below #100 on a chart or #200 in a search. Apple's search page points you to App Analytics for "impressions, conversion rate, downloads, and more". Those are yours, about you, and nobody else's are published.
Why the charts move every day
The charts are recomputed on a cadence Apple does not publish, but the visible effect is measurable. This site snapshots the three US charts daily, and the history holds 35 consecutive days, 2026-08-06 through 2026-09-09.
Across those 34 day-to-day pairs on Top Free, a mean of 90% of the apps that stayed on the chart changed position overnight, and a mean of 9.3 apps entered the top 100 each day. A one-day rank is a photograph of a moving object, which is why the movers on /trending use a seven-day window rather than yesterday.
Seven days of that history, 2026-09-02 against 2026-09-09, diffed at full depth:
| Chart | Positions replaced | Held their place | Median move among holders | Top 10 still in the top 10 |
|---|---|---|---|---|
| Top Free | 14 | 86 | 5 places | 7 |
| Top Grossing | 11 | 89 | 3 places | 9 |
| Top Paid | 18 | 81 | 7 places | 9 |
Top Grossing is the most settled of the three, this week as in the August measurement: the fewest replacements and the smallest typical move. Nine of the top ten on Top Grossing and Top Paid were the same apps a week earlier, seven on Top Free. The churn is further down.
What a 7-day mover means, and what it refuses to mean
Four rules govern how a mover is reported, and each exists because the alternative published a falsehood.
- The baseline is the snapshot closest to seven days before the latest, not exactly seven. A missed day shifts the window rather than voiding it.
- Delta is signed so that positive means improved. Ranks are inverted; a delta computed the other way colors every decline green.
- Delta is null, never zero, for an entrant or a drop-off. An app that left the top 100 landed somewhere past it, and the distance is unknown. A zero-place move would be a fabrication.
- Drop-offs are suppressed when the newer snapshot is below 90% of the baseline's size. A half-fetched morning would otherwise publish as fifty apps leaving the store overnight. A chart with no baseline at all produces no movers, since with nothing to compare against every app on it would file as a new entrant.
The per-genre movers on each category page come from the same diff, run once at build time, so the two pages cannot disagree about who climbed.
How to check any app's rank
For a chart rank, open the app on /apps. Each bundled app carries the position it held on the US Top Free, Top Paid and Top Grossing charts in the latest daily snapshot, and the "hide the giants" filter shows the apps ranking at your scale rather than the household names on top. For movement, /trending lists the week's climbers, fallers, entrants and drop-offs on those three charts, and each category page carries the same four lists for its own genre.
For a keyword rank, put the app and the query into /keyword-rank-checker and pick a storefront. It reads Apple's public search endpoint 200 deep and reports the position, or that the app is not in the first 200. Signed in, a checked combination can be tracked daily, which builds the history the rank-tracking post is about.
One thing to know when a rank disagrees with a product page. Apple's search index and its lookup endpoint are not always the same snapshot: recorded here on one app, seconds apart, search returned version 2.14.21 dated six days later than the 2.14.20 the lookup returned, and a rating count of 144,886 against 144,851. A keyword rank comes from the search index. If it seems a day ahead of the listing, it may be.
What this cannot tell you
- Why any app ranks where it does. Apple documents inputs for search and none for the charts. Nobody outside Apple has the weighting.
- What a position is worth in downloads. A chart position is an ordinal and the estimates that get fitted to it arrive coarsely rounded; the downloads-to-chart post covers why that quotient is a guess on a guess.
- Anything past #100 on a chart or #200 in a search. Not ranked in the visible depth is the whole of what a null means.
- Whether a change you make will move a rank. Apple names the fields it indexes. It promises nothing about what editing them does, and neither does this post.
- Any storefront you did not check. Both rankings are per storefront, and so are the rating counts that Apple says feed one of them.
Read a chart position as one cell in a grid that turns over daily, and a keyword position as one query in one country. Then check the ones you care about on /apps, /trending and /keyword-rank-checker, and read a null as unmeasured, never as zero.