Mobile app for business: when it actually pays off
An app earns its keep when users come back often enough to keep the icon on screen — not because it looks modern. Without that frequency, the budget belongs on a website instead.
The core test: how often people come back
People delete apps they use less than once a month almost immediately if they see no reason to return. Taxis, food delivery, banking, a coffee chain's loyalty card — frequency is natural there, and the app earns its screen space.
A dental clinic, a law consultation, a furniture showroom — customers show up once a year or less. An app does not work here: it either never gets installed or gets deleted within a week. A mobile-friendly website converts better than an app nobody opens.
The economics
A simple app — catalogue, ordering, payment — typically costs 15,000,000 to 30,000,000 UZS per platform in Tashkent, at 2-3 months. One with accounts, geolocation, push notifications and an admin panel runs 40,000,000-60,000,000 UZS over 3-4 months.
Add maintenance on top: updates for new iOS and Android releases, developer certificate renewals, fixes from user feedback — typically 10-15% of the build cost per year. An app launched and forgotten breaks on new OS versions within about eighteen months.
When a website wins
If the main acquisition channel is ads and shared links, a mobile website wins outright — a link has no install barrier, and content shows in seconds. An install funnel loses part of the audience at every step: store redirect, download wait, permissions.
A PWA (progressive web app) covers the need for a home-screen icon and offline access in most cases for a fraction of native app budget — covered separately in the PWA alternative article.
A practical test before committing
Ask three questions: will the user return at least once a month, does the product need camera, geolocation or push access, does transaction volume justify the build cost. Two 'yes' answers out of three is reason enough to budget seriously.
If the answers are uncertain, test demand on a website or a Telegram bot first, and hold the app budget until frequency and audience are confirmed by numbers rather than a hunch.