Open almost any phone on a morning train and the home screen tells a familiar story: a few apps worn smooth with use, and a long tail of icons nobody has touched since installation day. Sensor Tower’s numbers for 2025 back it up. Downloads worldwide grew by a mere 0.8 percent. Hours in apps grew by 3.8 percent, which averages out at roughly 3.6 a day per user.
Everyone with an icon to defend feels that squeeze: retailers, broadcasters, bizbet and the rest of the sports operators. A fresh install proves little. Staying on the first screen is the real contest, and it gets played out on handsets with half-full storage and a signal that comes and goes between stations.
The quiet part of the report
Flat downloads sound like stagnation. Read further and the same report says something else: people spent more on non-game apps than on games for the first time, and India ranked first for total time spent. Sensor Tower even calls it a monetisation-first era. Revenue per user matters more now than the next million downloads.
Asia Pacific is where that gets interesting. According to the GSMA, close to 1.8 billion people in the region are unique mobile internet users, a shade over 60 percent of the population. Coverage alone does not explain who is online. The GSMA’s 2025 connectivity report counted 3.1 billion people worldwide who sit inside a coverage area and still do not use mobile internet, and it points to handset prices as one of the barriers.
Voucher codes, one tab at a time
Discount codes are a neat test of how awkward a phone can make a simple chore. A VoucherCodes.co.uk survey from September 2025 found that almost 60 percent of people in Britain check for a code before paying. On a laptop that costs a second window. On a phone it means abandoning the retailer’s app for the browser, copying a string of capital letters, and getting back before the basket resets.
Conditions make it fiddlier. Some codes lapse within days, some need a minimum spend, some quietly exclude half the catalogue. If a code is rejected, the shopper heads back to the search tab, and each trip is another chance for a patchy connection to end the session. Retail had a tougher year overall. Sensor Tower logged fewer downloads and less time in retail apps as the expansion of Temu and Shein slowed.
Live scores on the same connection
Sports apps live and die on freshness. Sensor Tower found global downloads of sports betting apps up 24 percent in 2025, helped by launches in new markets such as Brazil, while fantasy sports apps drove growth in India. Late scores get noticed within seconds, which means a weak signal hurts a sports app sooner than a shopping one. The betting and fantasy figures also describe different products and audiences, so adding them together would mislead.
Scores, account access and payments all share one connection, though not equally. Text updates are cheap on data. Video is expensive. On a capped plan the difference shows up within days.
What earns a place on the phone
Storage fills with photos and chat attachments long before anyone thinks about it. Each new installation, a bizbet download among them, has to fit in whatever space remains. The store listing helps: it states the file size and the date of the last update, and a long silence since that date says something about whether anyone is still looking after the app.
Worth checking before installing anything:
- Free storage, with room left for updates.
- How big the first download is. A large one is better saved for Wi-Fi.
- Permissions: what the app asks for, set against what it actually does.
- How it behaves offline or on one bar of signal.
- Whether the file comes from the publisher or the official store page.
An app that asks for access unrelated to its job deserves a second look. And large updates are better kept for Wi-Fi, since a few of them can drain a monthly allowance.
Where the evidence ends
Sensor Tower’s figures are global, the GSMA’s regional, the voucher survey British. None of them isolates voucher use or sports app use among Asian phone owners, so claims about that group are inferences. The full reports may break the numbers down further. The public summaries do not. Anyone concluding a specific market would need local survey data, which none of the three sources provides.