Mobile App Support and Growth

The release isn’t the finish: an unsupported app degrades within a year

A mobile app lives in a hostile environment: new iOS and Android every year, tightening store policies, new devices, aging SDKs. A product abandoned after release degrades on schedule: crashes pile up, the rating creeps down, installs get pricier. In our delivery case, a year without support cost the rating a slide from 4.7 to 3.9 — and a month of stabilization to return. Support isn’t an option after development but the product’s living condition.

Health 24/7: problems visible before the reviews

Monitoring reads the vitals continuously: crashes and ANR by device and version, cold start speed, post-release degradations. Alerts reach the team instead of accumulating in angry users’ reviews. The health bar — 99.5% crash-free sessions — holds as an engineering norm. When degradation does happen, the SLA works: a critical incident gets a 2-hour response, an out-of-turn fix and a post-mortem with anti-recurrence measures.

Betas in advance: September stops being a fire

Every autumn Apple and Google ship new OS versions, and unprepared apps fall for millions of users at once. We test the betas from the first builds — months before the public release: catching API breaks, updating SDKs, shipping compatibility ahead. In the digital product director’s review, September passed routinely for the first time — after years of fire-fixes-and-apologies-in-the-reviews. Store policies run the same way: changes get tracked and closed before deadlines, not after a block.

Growth sprints: a team without hiring

The business needs features monthly, and keeping a mobile team for them is expensive: salaries, taxes, vacations, churn. The sprint format gives a team without hiring: we run the backlog together, features get estimated transparently, releases follow the calendar, and every feature carries an expected effect and the fact after. In the booking service CEO’s review, keeping their own developers would’ve cost triple at their volume. Support economics computes easily — and usually in its favor.

Foreign code, reporting and the place in the stack

Half our cases are apps after other teams: the vendor vanished, the code undocumented, the build lost. The procedure is practiced: an audit, build recovery, stabilization, monitoring — then a normal life. Hopeless code we’ll call hopeless honestly, but more often we save it. The report is monthly: health, releases, money. Nearby sits the whole mobile stack: iOS and Android development, the backend, integrations — the cases are in the trio below. The product lives with one vendor.