Mobile Applications people actually keep

Mobile stopped being the “second screen” a long time ago. For most consumer businesses it’s the first one — and increasingly the only one. The question is no longer whether to be on mobile, but whether you’re there as a website someone tolerates on a small screen, or as an app they actually keep.

Mobile is where commerce happens now

The behavioural shift is settled. In the US, mobile accounted for 47.7% of online sales through the first seven months of 2024 — about $280 billion, growing 10.2% year over year (EMARKETER / Adobe Analytics). And the time people spend on phones is overwhelmingly in apps: eMarketer found apps capture roughly 88% of US adults’ mobile internet time, not the mobile browser. Globally, app stores saw 257 billion downloads and $171 billion in consumer spend in a single year (data.ai, State of Mobile 2024).

That last point matters because of what it implies about conversion. When someone is in your app, they convert dramatically better than on the mobile web:

Vertical	In-app conversion	Mobile web	Gap
Retail	18%	4%	~4×
Travel	20%	6%	~3×

(Criteo Global Commerce Review — older data, but the gap has proven durable.)

Retention is the metric that humbles everyone

Getting installs is the easy part. Keeping users is where most apps quietly fail. The benchmark data is sobering: across categories, average Day-1 retention sits around 25%, and by Day 30 only about 5–6% of installs are still active (AppsFlyer, via Business of Apps). The curve looks like this:

Users still active

The cliff is in the first 72 hours. That’s where the engineering actually pays off: a fast, native-feeling first run; push notifications that are useful rather than annoying; offline support so the app works on a subway; and an onboarding flow that gets the user to value before they bounce. We design for the shape of that curve, not just the launch-day install count.

Cross-platform, without doubling the team

You don’t usually need two separate native codebases to get a great app. React Native and Flutter are now the two leading cross-platform frameworks — used by 8.4% and 9.4% of all developers respectively (Stack Overflow Developer Survey, 2024) — precisely because one team can ship a high-quality app to both stores from a shared codebase, dropping to native modules only where it counts (camera, biometrics, push, payments). Shared logic and a thin native bridge feed both the iOS and Android builds; platform-specific code is the exception, so fixes land everywhere at once.

In practice most of the app is genuinely shared — the same screens and the same business logic on both platforms — with native APIs like payments, camera and biometrics handled underneath. The shared layer is the rule; bespoke native code is the exception.

When a project genuinely calls for it — heavy real-time graphics, deep platform integration, an AR feature — we go fully native (Swift, Kotlin). The framework is a decision we make from your product’s requirements, not a religion. What stays constant is the goal: an app that’s fast on day one and still on the home screen on day thirty.

References

• EMARKETER — Mobile commerce / retail sales, 2024
• eMarketer — The majority of mobile time is spent in apps
• data.ai — State of Mobile 2024 (report PDF)
• Criteo — In-app vs. mobile web conversion
• Business of Apps — Mobile app retention benchmarks
• Stack Overflow — Developer Survey 2024: Technology