The Update Treadmill

The active core is bigger than the pessimists claim. 12.1% of apps shipped an update in just the last 30 days, and 59.8% were updated within the past year. For all the talk of abandonware, the majority of the store is being actively maintained: patched, adapted to new iPhones, kept current. An app you download today is, more likely than not, one someone is still working on.

But the tail is real. 28.3% of apps have not been touched in over two years, and on a platform that changes annually, that's roughly the point where "still available" stops meaning "still working." A lot of these apps still launch fine, but they fail in quieter ways: logins break, compatibility with the newest iOS slips, security holes just sit there unpatched. They clutter search results and inflate category counts even though they're basically frozen in time. Honestly, a two-year gap without an update is the single clearest red flag on a listing.

Look at the timestamps for a second. The median app last got an update 222 days ago, about 7.3 months. Split the store in half and the two sides barely resemble each other: the freshest tenth of apps had been touched within 26 days, while the stalest tenth hadn't seen a release in over 2,179 days, roughly 6 years. Those three numbers alone paint the picture: a sizable core that's actively maintained, and a tail that's basically been left for dead, with not much sitting in between.

That gives you a rough yardstick to use. If the app you're looking at was updated more recently than 7.3 months ago, it's living in the fresher half of the store. If the last update stretches back years, it's fallen into that stale tail, and that's worth questioning before you trust it.

Keeping an app maintained does more than just keep it running. It shows up directly in how much people actually like it. Among apps with real rating volume, the ones updated in the last year average 4.44 stars, while the ones left untouched for over two years average 4.26. That gap looks small at first glance, but on a compressed rating scale it actually counts, and no matter how you slice the data, it points the same direction.

The clearest version of this shows up in the share of apps people genuinely like: 66.1% of recently-updated apps score 4.5 or higher, against just 47.1% of stale ones. Good developers keep shipping, and updates fix the things that annoy people. Both happen together. A recent update date and a high rating tend to go hand in hand.

Abandonment is really a long tail problem. It barely touches the apps people actually use. Look at the 500 most rated apps in the store, the household names, and 99% of them were updated in the past year. Widen that out to the top 2,000 and the number only slips to 97.5%. Popularity and upkeep move together at the top of the chart. You can't hold onto a huge, active audience without maintaining the app, and that same audience is what funds the maintenance.

That's why the 28.3% stale figure can feel off compared to your own experience. The dead apps are real, they just cluster among obscure titles almost nobody opens. The apps actually competing for attention on the charts and in search are, with very few exceptions, still getting worked on.

Not every category moves at the same speed. Finance apps (78.4% updated within a year) and Developer Tools keep pushing out updates because they run on live data, servers, and competitors breathing down their necks. Stickers apps (25.2%) sit at the other extreme, where simple tools and reference titles often get left alone once their makers decide they're done. That gap says a lot about how much upkeep a category's apps actually get.

Two honest limits. First, an update date records that a release happened, not what it contained: a genuine bug-fix and a token version bump to dodge Apple's "app not updated recently" warning look identical to us, so freshness is a signal of attention, not a guarantee of improvement. Second, "fresh apps rate 4.44 versus 4.26" is a correlation, not proof that updating raises a score — well-run apps tend to do both, and we cannot separate the two here. Every figure — the 222-day median, the 28.3% stale share, the 99% of top apps still maintained — is counted directly across 1,043,643 apps, never modelled.