State of Mobile 2026: people now spend more on apps than games, and AI is why

Sensor Tower published State of Mobile 2026 in January 2026, summarising full-year 2025 data from iOS and Google Play. Global app downloads reached nearly 150 billion, up 0.8% and above the previous high set in 2023.

In-app purchase revenue totalled US$167 billion, up 10.6%. Time spent on mobile reached 5.3 trillion hours, up 3.8%, or about 3.6 hours per day for each mobile user.

The report also names the United States as the largest market by revenue, with nearly US$60 billion in consumer spending. Overall, downloads barely grew while spending and time grew faster. In our view, that points to a maturing market: fewer new installs, and more value from the apps people already keep.

The most discussed finding is that 2025 was the first year non-game apps out-earned games. TechCrunch reported app spending of about US$85 billion, up 21% and nearly 2.8 times the level of five years earlier. Games came in at nearly US$82 billion, up only 1.3%.

Both Sensor Tower and TechCrunch point to generative AI services as a key driver of this growth. In our view, the figures show that consumers will pay for apps they use regularly and that deliver clear value, not only for entertainment such as games.

Generative AI apps were a major driver. Downloads doubled to 3.8 billion, in-app revenue nearly tripled to over US$5 billion, and time spent reached 48 billion hours, about 3.6 times the 2024 total. Sessions passed one trillion.

TechCrunch noted that ChatGPT earned US$3.4 billion in global in-app revenue, and that OpenAI and DeepSeek together accounted for nearly 50% of AI app downloads, up from 21% in 2024.

A side effect for every other app is higher user expectations. Once people are used to typing a question and getting an answer straight away, apps that make them tap through several layers to find information feel slow and dated.

The report shows credit and lending app downloads up 18%, restaurant and food delivery apps up 14%, and sports betting downloads up 24% globally. Social media users averaged 90 minutes a day.

By contrast, global downloads and time spent in retail apps declined. That is worth noting for stores considering their own app, because consumers appear to be concentrating shopping in a few apps they already use.

This is global data and the public report does not break out Thailand, but the overall picture helps answer a question we hear often: should we build an app? Our answer depends on frequency of use. An app pays off when customers return often, as with financial services, food ordering, membership and loyalty programmes, or tools people use every day.

For typical retailers whose customers buy occasionally, a fast mobile website combined with a LINE Official Account and marketplace stores is usually better value than a separate app, because customers do not need to download anything.

If you do build an app, plan from the start for recurring revenue or ongoing value to customers, and consider whether AI features such as plain-language search or a help assistant would genuinely remove steps for users.

Before investing in an app, try to answer these questions.

If the answers point away from a native app, that is not a step backwards. A fast, well-designed website can deliver most of the same value at lower cost, and you can still build an app later when usage justifies it.

• How often will customers open the app each month? If less than weekly, consider a web app first.
• What can an app do that the web cannot, such as notifications, scanning or offline use?
• How will you earn from or measure the app: subscriptions, repeat purchases, or lower service costs?
• Test AI search or help features at the points where customers get stuck most often.
• Budget for yearly maintenance from day one, not only the build.