Every badge on an app page answers one question: how many apps have ever done this? Here is exactly how we compute it, and what the number does and does not mean.
An App Achievement is a download threshold that an app has crossed. It is defined by three parameters: the metric (cumulative downloads), the threshold (from 500 to one billion) and, for speed achievements, a time window measured from the app’s release. Nothing else enters the definition. The badge you see on an app page is only the visual for it.
The app eventually reached this many downloads, with no time limit.
The app reached this many downloads within 7, 30, 90 or 365 days of release. A far smaller number reached in a week is often rarer than a large one reached over a decade.
A tier is not a ranking of apps. It describes how rare the achievement itself is: what share of the apps in the reference population ever crossed that threshold. Every app holding the same badge in the same population therefore carries the same tier.
share = apps that crossed the threshold ÷ apps in the reference population
| Badge | Tier | Share of apps reaching it | In plain terms |
|---|---|---|---|
 | Legend | ≤ 0.01 % | at most 1 in 10,000 |
 | Elite | 0.01 – 0.1 % | about 1 in 1,000 |
 | Master | 0.1 – 1 % | about 1 in 100 |
 | Challenger | 1 – 5 % | about 1 in 20 |
 | Rising | 5 – 15 % | between 1 in 20 and 1 in 7 |
 | Discover | > 15 % | more than 1 in 7 |
Concrete example: 6 of the 1.4M+ iOS apps we track have passed one billion downloads — 0.00043%. That is a Legend tier.
The same threshold is not equally hard depending on when an app launched. So each badge is measured against five populations: all apps we track, and the apps released in the same year, half-year, quarter and month. A badge therefore carries five readings, not one.
Why it matters: an app released in 2024 that reached a million downloads did so among a very different set of peers than one released in 2012. Comparing it to the whole catalogue would flatter it; comparing it to its own release cohort answers a different, sharper question.
A reading is published only if its reference population contains at least 200 apps. Below that, a share stops being informative: one app out of four is 25%, which describes nothing. This excludes the narrowest early cohorts, mostly monthly ones from 2008 to 2011. No lifetime reading is ever affected.
We deliberately set no floor on the number of apps holding a badge. Both counts come from the same measured population, so a very small numerator is not noise — it is the finding. A minimum-holders rule would have deleted precisely the rarest achievements, which are the most informative ones we publish.
All figures are worldwide download estimates produced by MWM Intelligence, our live market data platform tracking 15M+ apps. They are estimates, not publisher-reported numbers — which is precisely why we never present an exact download count for an app, only estimated thresholds it has crossed.
What we estimate is the moment an app crosses a meaningful milestone. Because a threshold is a far more robust thing to estimate than a precise total: being confident that an app has passed a million downloads does not require knowing whether it sits at 1.02 or 1.14 million.
Achievements are recomputed daily; the figures shown on app pages are refreshed weekly and always carry the date they were computed. That date is the one to cite.
The date an app first crossed a threshold is frozen the first time we observe it, and never recalculated afterwards. Without this, a revision upstream would silently change a fact we had already published.
A tier is not a position in a ranking of apps. “Legend” means the threshold was reached by at most 0.01% of the population, and that this app is one of them — it does not mean the app ranks in the global top 0.01% on any other measure.
Tiers from different cohorts are not comparable. Apps released in 2012 have had fourteen years to accumulate downloads, so a larger share of that cohort crossed any given threshold. Narrowing the cohort makes an older app look less exceptional, not more.
Readings for recent cohorts will drift. A 2024 badge shown today at 0.33% will keep rising for years as more of that cohort catches up. That is why every figure is dated. For comparisons across cohorts, we use figures normalised to equal elapsed time instead.
Every figure on this site is an estimate produced by MWM Intelligence. Cite it as an estimate, and cite the date it carries.