Just a few years ago, the growth formula for a mobile app was almost universal: build a product, buy installs, optimize the funnel. In 2026, this formula no longer works as it once did. The number of installs in the market is barely growing, acquisition costs continue to rise, and users increasingly make the decision to install not in the store but through recommendations from AI assistants or high-intent search queries.
At the same time, the market itself remains enormous. According to Fortune Business Insights, the global mobile app market was valued at approximately $298.4 billion** in 2025, and in 2026 it is projected at **$330 billion, with further growth to $1.017 trillion by 2034. Other estimates give comparable figures: a CAGR in the range of 14–17% depending on methodology. Within this large market, structural shifts are occurring that are more important than the overall volume.
The main shift of 2026: non-game apps have overtaken games in in-app purchase revenue for the first time. According to Sensor Tower, in 2025 IAP revenue from non-game apps grew 21% year over year to reach $85.6 billion**, while games brought in **$81.8 billion. Total IAP revenue in the market was $167 billion. This is not a cosmetic change: it means a redistribution of advertising budgets, investments, and developer attention.
Who is growing fastest: generative AI as the new axis of the market
If one were to look for a single category that explains most of the growth in the non-game segment, it is generative AI. According to Sensor Tower, from Q2 2025 to Q1 2026, IAP revenue in generative AI apps grew 232% year over year, reaching $6.1 billion** for the year. Quarterly revenue for this segment grew from less than **$60 million in Q1 2023 to $1.9 billion in Q1 2026 — more than 32 times in three years.
This changes the competitive picture on several levels at once. First, AI apps are taking user time and money from other categories — primarily from mobile games, whose revenues in certain quarters of 2026 fell 4.5% year over year, while non-game apps grew 14.6%. Second, AI apps are creating new monetization patterns: subscriptions with high willingness to pay, a fast hypothesis-testing cycle, and a low barrier to entry for small teams.
However, there is a flip side to this coin. The number of apps is accelerating: in the first half of 2026 alone, more than 3 billion apps were downloaded from Russia’s RuStore, and nearly 13,000 new apps are released worldwide every day — 106,000 in the App Store and 227,000 in Google Play in October 2026. Competition for visibility is growing faster than the market.
Acquisition economics: why CPI is no longer the main metric
The cost of an install in 2026 depends heavily on category and geography, and market averages are almost useless for planning. According to Adjust, in games the average CPI was $0.56** (up 30% year over year). In finance, CPI globally is **$1.13, in North America $4.13**, and in Europe **$4.75. In e-commerce, CPI in North America is $2.49**, and in Europe **$2.25. According to Appalize estimates, the average CPI on iOS exceeds $3.50**, and on Google Play it is **$1.20.
But the key problem is not the absolute cost of an install. The problem is that an install is no longer an indicator of business health. If an install costs $4 and one in ten users takes out a subscription, the cost per paying user is already **$40**. Budget decisions in 2026 are made based on CAC (cost per paying user), ROAS D7/D30 (return on ad spend over 7 or 30 days), and LTV (lifetime value), not CPI.
The basic rule — LTV must be higher than CAC — has not changed, but the gap between them is narrowing. Average CAC across all channels hides unprofitable channels, so it has to be calculated separately for each channel. This means the era of “pouring in traffic and seeing what happens” is over. Budget is allocated based on unit economics, not intuition.
ASO in 2026: from keyword optimization to intent management
App Store Optimization remains the most cost-effective growth channel because organic installs require no attribution and have zero marginal cost with each improvement in algorithmic signals. But ASO in 2026 is no longer “putting keywords in the title and description.”
Algorithms have become more complex. Apple and Google use engagement signals, retention data, and contextual relevance. Superficial keyword stuffing no longer works — genuine quality signals are needed. Apple evaluates text relevance (title — 30 characters with the greatest weight, subtitle — 30 characters, keyword field — 100 characters), download velocity, retention and engagement, volume and average rating of reviews, as well as conversion signals: CTR from search to the card and conversion from card to install. Google Play focuses on text relevance (title, short description — 80 characters, full description — 4,000 characters), install velocity, uninstall rate (a unique negative signal), Android Vitals (crash rate, ANR, battery consumption), and NLP sentiment analysis of reviews.
Visual assets have become more important than text. The icon is the most important creative element, appearing everywhere: in search results, in charts, on the home screen. Screenshots should tell a story, not just show screens. A preview video can increase conversion by 20–35% if it is shorter than 30 seconds and the first 3 seconds work.
Customization for campaigns. Apple allows up to 35 custom product pages with unique screenshots and videos to align ad creatives with landing pages, seasonal promotions, and pages for specific features. Custom pages have begun appearing in organic search — this blurs the line between paid and organic ASO.
Apple Ads have changed. Since March 3, 2026, a second ad placement has appeared in App Store search results below the results. You cannot choose the position or place a separate bid on it — the placement is determined by auction. Apple also renamed Apple Search Ads to Apple Ads, signaling expansion beyond search. For ASO teams, this means increased competition for visibility and the need to account for paid placements in organic growth strategy.
Google Play is moving toward a visual storefront. The Play Shorts format has launched — short videos in the apps section for discovery through video rather than through the product card. For games, a PC section, cross-platform badges, and trial versions have been added.
Acquisition channels: why the “big four” no longer guarantee results
Global mobile marketing spend grew to $109 billion in 2025**, of which **$78 billion went to User Acquisition and $31 billion to retargeting. At the same time, the effectiveness of the main channels is declining: install costs are rising, and install growth is stagnant.
The “big four” — Google, Meta, Apple, and ad networks — still deliver volume, but rising CPI and shrinking incrementality mean most teams need at least one channel outside this group. A healthy UA structure in 2026 is not choosing one winner but a combination of paid search, social media, ASO, and new inventory sources that do not compete for the same auction.
One such source is mobile OEM inventory: preinstalls, on-device display, and alternative app stores sold directly by manufacturers (Samsung, Xiaomi, vivo, Huawei, HONOR, Transsion, OPPO). This has ceased to be a niche experiment and has become a real acquisition channel. For developing markets (India, Indonesia, Brazil), diversification is especially important because the distribution of Android devices there favors OEM channels and alternative stores rather than a “Google-first” strategy.
Another shift is search through AI assistants instead of the store. Some users formulate a query in a chatbot or voice assistant rather than typing keywords into the App Store. This changes the nature of ASO: semantic relevance and the structure of app data become more important than exact keyword matching.
Retention: the only metric that multiplies all the others
In 2026, retention has finally become the main KPI of a mobile app. The logic is simple: if the user does not return, no investment in acquisition pays off. Retention multiplies LTV, justifies higher acquisition costs, and creates organic growth through word of mouth and positive reviews.
Classic metrics — D1, D7, and D30 retention — show what percentage of users returned to the app on the 1st, 7th, and 30th day after install. According to RevenueCat, which analyzed more than 115,000 apps with combined revenue of over $16 billion, average first-year retention for subscription apps is below 30%, and most paying subscribers churn before the first renewal. At the same time, the subscription app market has become “winner takes all”: the top 25% of apps increased MRR by 80%+ year over year, while the bottom 25% lost more than 33% of MRR.
For teams, this means the battle is not for the install but for the second week after install. If the user does not return on day 7, the probability that they will stay and pay drops sharply. Therefore, UA strategy and product strategy can no longer exist separately: traffic quality is determined not by CPI but by how well the app retains the users who come.
How to choose a niche: six steps that save months of development
One of the most expensive mistakes in mobile development is choosing a niche “for yourself” or based on one country. If you lack a convenient expense tracker, that does not mean hundreds of thousands of other people are looking for one and that existing solutions do not suit them. A category that is overheated in the US may be almost empty in Germany or Brazil. An exam-prep app launched in July will see a drop in downloads and draw incorrect conclusions about demand.
Market research before starting development removes these risks in a few hours. The basic framework includes six steps:
Assess the market as a whole in tools like Store Insights: how many apps there are, how many new ones are released, how games and non-games divide, paid and free, which categories lead in downloads and revenue. Release dynamics matter: in October 2026, there were 13.6% fewer new apps in the App Store than in September, while in Google Play there were 52.8% more. For a newcomer, this is a signal that competition for visibility in Google Play is growing faster.
Break down the category by three indicators: average monthly revenue, average monthly downloads, and total number of apps. Example: the Food & Drink category in the US App Store has 41,430 apps, gathers an average of 23.6 million downloads and $6.2 million** in revenue per month. By downloads it ranks 13th, by revenue only 39th. Each download accounts for about **$0.26. The gap is explained by the fact that top brands earn from food orders, not from in-app purchases.
Compare the top by downloads and the top by revenue. Categories that appear only in the first list gather an audience but monetize poorly. This does not mean you cannot earn there — it means you need to understand who exactly pays and for what.
Assess the monetization structure. In the US App Store, 76.3% of apps are free, meaning roughly one in four is paid. In Google Play, 90.7% are free — only one in ten is paid. The share of paid apps in the App Store is 2.5 times higher, so for one-time payment it makes more sense to start with iOS.
Check seasonality and geography. The same category can be overheated in one country and open in another. Starting markets should be chosen not by size but by the ratio of demand, competition, and willingness to pay.
Assess the realism of acquisition. If large brands with big advertising budgets dominate the category and search demand is weak, even a good product may fail to gain users. Niche research is not a guarantee of success but a way to avoid spending months on a product the market is not waiting for.
A strategy that works in the US may fail in Latin America, and ASO rules in China differ radically from Western ones. Below is a study of four key markets: the US, the Spanish-speaking segment, English-speaking countries (outside the US), and China.
🇺🇸 The US: the largest market that has stopped growing
The US remains the world leader in spending, but install volumes and in-app purchase (IAP) revenue are stagnating or falling. In Q2 2026, IAP revenue in the US declined 3% year over year, although the country retained first place with $14.8 billion. Downloads totaled 3.08 billion, 0.2% less than a year earlier.
Nevertheless, the US is the market with the highest monetization value per user. Average CPI here is $3.50–$6.50 for non-game apps, explained by high competition and strong LTV that justifies premium rates. In the finance category, iOS CPI in North America reaches $4.13**, and in Europe **$4.75.
Key trend: the US was the first to feel the structural shift toward non-game apps. The Dating & Social Discovery category in the US contracted 34% year over year, while in Latin America it is growing at double-digit rates. This is a signal: the American market is oversaturated, and growth there is possible not through volume but through depth of monetization — subscriptions, retention, and LTV.
🇪🇸 The Spanish-speaking segment: Latin America as a growth zone, Spain as an “expensive entry”
The Spanish-speaking world is heterogeneous. Spain is a mature market with 47 million users and 92% smartphone penetration, but CPI there has grown 20–30% since 2023 and has approached Western European levels. iOS penetration in Spain is about 45%, higher than in most Latin American countries.
Latin America is the opposite. Installs and sessions in the region grew 3% and 1% respectively in Q1 2026. In Mexico, app installs grew 20%, and in non-game categories 28%. CPI in Latin America is significantly lower: $0.70–$1.80 in Mexico and $0.80–$2.00 in Brazil.
However, subscription conversion in Latin America is 30–50% lower than in the US for the same app. This means low CPI does not guarantee high payback. Those who use hybrid monetization models — advertising, IAP, and subscriptions simultaneously — win. Already 10% of non-game apps in the region monetize through IAP and subscriptions.
Strategic takeaway: Latin America is a market of volume and experimentation, not quick payback. For startups with limited budgets, it is an opportunity to acquire users and test product hypotheses before entering expensive markets.
🇬🇧 English-speaking markets outside the US: the “second tier” with high competition
The UK, Canada, and Australia are first-tier markets by competition but with smaller volume than the US. CPI in the UK is $2.80–$5.00, in Australia $3.00–$5.50, and in Canada $2.00–$4.00.
These markets are often combined with the US into one campaign due to language similarity, but this is risky: Canada and the UK have their own cultural and consumer characteristics. The Canadian market often “sticks” to US campaigns and receives residual traffic, leading to wasted budget.
Key feature: English-speaking markets outside the US benefit from English-language content — localization is often not required, which reduces entry costs. But competition for user attention remains high, and without localized visual assets (screenshots, icons, videos), conversion from card view to install falls.
🇨🇳 China: the largest ecosystem with the highest barriers
China is the largest market by revenue after the US, but the rules of the game there are fundamentally different. It is estimated that in 2026 the Chinese app market (including iOS and legal Android channels) will be about $118 billion**, and the total market (including illegal channels) is estimated at **$120–140 billion, which may exceed the American one. The number of smartphone users exceeds 1.1 billion, and mobile internet MAU is 12.76 billion (including multiple devices and accounts).
Monetization and commissions: Apple reduced the standard commission for the Chinese App Store from 30% to 25% as of March 2026, but external payment links and third-party stores remain prohibited. This means developers cannot bypass Apple’s commission through web payment, as is possible in some Western markets.
Regulation and localization: China requires strict compliance with local norms, including content registration, data compliance, and licensing. For foreign developers, this creates a high barrier to entry. At the same time, the Chinese market is not monolithic: there are alternative app stores (Huawei AppGallery, Xiaomi GetApps, Tencent MyApp), each with its own audience and rules.
ASO in China requires a separate strategy: local store algorithms take into account not only keywords but also social signals — reviews, ratings, activity on social networks. Without content localization and adaptation to local cultural codes, entry into the market is practically impossible.
📊 Summary table: CPI and market potential
| Market | CPI range (non-game) | Key feature |
|---|---|---|
| US | $3.50–$6.50 | High LTV but stagnating volume |
| Spain | ~$2.50–$4.50 (estimate) | Mature market, high iOS share |
| Latin America | $0.70–$2.00 | Cheap traffic, low subscription conversion |
| UK | $2.80–$5.00 | English-speaking, high competition |
| Canada | $2.00–$4.00 | Often “sticks” to the US |
| Australia | $3.00–$5.50 | Small volume, high CPI |
| China | High entry barrier | Largest market, strict regulation |
Sources: SEM Nexus, RocketShip HQ, Adjust
🌍 Global context: two trends more important than geography
The first trend — non-game apps have overtaken games in revenue. In Q2 2026, non-game apps brought in $24.4 billion** in revenue versus **$19.2 billion for games. Generative AI became the main driver: spending in this category grew 108% year over year, and ChatGPT holds about 60% of the category’s revenue with 82% growth.
The second trend — retention has become the main metric. Median D1 retention across all categories is 25.4%, but apps with session depth of 4.2+ per day show 37% higher D30 retention. Paid traffic accounts for 38% of all installs globally, and CPI inflation averages 12% on iOS and 7% on Android. This means buying installs is becoming more expensive, and the return on them less predictable.
💎 Strategic conclusions
For the US: Do not chase volume. The market is saturated, and depth of monetization wins — subscriptions, retention, LTV. If your product does not have a clear value proposition for the American user, entering there will be expensive and risky.
For Spanish-speaking markets: Latin America is cheap traffic and a testing ground, but not a guarantee of payback. Spain is the “gateway to Europe” with high CPI and a mature audience. Separate your strategies for Spain and Latin America: they are different markets with different economics.
For English-speaking countries outside the US: Use the language advantage but do not merge campaigns into one. Canada, the UK, and Australia require separate budgets and localized creatives, even if the language is the same.
For China: Without localization, a partner, and readiness for regulation, it is not worth entering. This is a market for those willing to invest in long-term presence, not for quick experiments.
Main takeaway of the study: In 2026, geography is not just “where traffic is cheaper” but “where your product can retain and monetize the user.” A market that looks cheap on entry may turn out to be expensive on exit. And vice versa: high CPI in the US is justified if the user’s LTV exceeds the acquisition cost several times over. Choosing a market is choosing an economic model, not just setting up geo-targeting.