What we have

One game, four surfaces.

Weekly columns = Jul 13–19, the last full week with all four surfaces open (web was gated Jul 22 – Aug 4 and is recovering; Android remains unlisted). iOS: 2.1× monthly growth and 81% of all games played. June's mobile-web reach was a paid campaign — high traffic, low intent.

Retention · Apple App Store Connect

Retention plateaus instead of decaying.

Apple's own daily-retention curve, days 1 through 23. After the first week it flattens in the mid-30s and holds there for more than three weeks — essentially no decay.

Apple App Store Connect · Daily Retention · opt-in devices · cohort average · full retention cohorts (Apple + ours) ->

Cohorts & depth · our data

Weekly retention and games per player.

Weekly cohort retention

How to read each row is a group of users by the week they first opened the app. wk 1 = the share of that group still active 7+ days later; wk 2 = 14+ days; and so on. Higher = stickier. Brighter cell = higher retention. "·" = week hasn't elapsed yet.
low high

Games per player · lifetime, share of players

Where they play

Who the players are.

Current daily

Where it sits now.

All organic — the only paid campaign (Apple Search Ads) ran Jun 30 – Jul 17 and has been off since. The base peaked at an all-time high on Aug 26 (70 installs, 160 daily active, 34 sign-ups); after that late-August surge it has settled to a steady organic run-rate of ~40–65 installs and ~110–145 daily active per day.

Day by day · last 14 complete days (UTC)

Installs = new iOS devices, deduplicated per physical device (floor — sub-5s bounces uncounted); daily active = iOS devices that opened the app; sign-ups = non-guest registrations. Internal, QA and junk traffic excluded. Restated Sep 14 on a stricter device-level identity — earlier snapshots counted guest and signed-in states of one device separately.