Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
With high-end stores equipped with AI software, developers may think that the best bet for bringing value is to integrate artificial intelligence technology into their products. However, a new study that looked at app subscriptions for iOS, Android, and the web challenges the idea.
RevenueCatThe company, which provides subscription management tools used by more than 75,000 software developers, said in a statement 2026 State of Subscription Apps Report that AI integration does not guarantee long-term retention. In fact, AI-powered apps struggle to retain subscribers, with people canceling their annual subscriptions — a metric known as churn — 30% faster than non-AI apps, on average, according to the report.
The report is based on a survey of subscribers who use it RevenueCat tools to manage their more than 1 billion in-app transactions, which generate more than $11 billion in revenue for developers annually. As one of the most popular tools in this space, its data represents a great example of event research.
Among the many interesting findings, the report revealed that many applications using the company’s platform were not powered by AI. AI-powered software makes up 27.1% of software in all categories, compared to 72.9% of non-AI software. However, it’s a growing category, as nearly one in four apps now has AI.
(To be clear, the category of AI-powered software includes popular AI chatbots, such as ChatGPT and Gemini, as well as any software that markets itself as AI-powered.)
Graphics and video applications have the largest share (61.4%) of AI-powered applications, while games have the smallest share at 6.2%. Travel (12.3%) and Business (19.1%) also have low levels of AI.
The most surprising numbers are around the ability of AI software to retain paying customers. AI software doesn’t perform well on monthly and yearly savings, RevenueCat shows.
Annual retention, which focuses on the program’s ability to retain subscribers after 12 months, was 21.1% for AI programs, compared to a higher 30.7% for non-AI programs. On a monthly basis, AI software saw retention rates of 6.1% compared to 9.5% for non-AI software – a difference of 3.4 percentage points.
The only area where AI led to retention was on the front end of the week, where AI programs had retention rates of 2.5% compared to 1.7% for non-AI programs. It is worth noting that weekly subscriptions are not the most popular option for AI software.
These metrics can take into account the rapid evolution of AI technology, which can see users jumping between different AI programs quickly, as they try to find the one with the most advanced technology under the hood.
As customers test a number of AI applications, they may find that some do not meet their needs. The report shows that AI software has a 20% higher return on investment (4.2% vs. 3.5% median) than non-AI software.
The higher rate of return for AI software is also higher (15.6% vs. 12.5%), meaning there is “high volatility in operational costs and deep operational cost, innovation, and long-term quality,” the report said.
There are some advantages to being part of AI-driven software, which shows.
RevenueCat found that AI programs convert users from trials to paid customers 52% more than non-AI programs (8.5% vs. 5.6% median), and AI programs monetize their downloads around 20% better than non-AI programs (2.4% to 2% median).
AI software also generates 39% or more monthly lifetime realizable value (RLTV), a metric that measures the true value of a payer over time. AI software averages in this metric at $18.92 per month, compared to $13.59 for non-AI software. AI software also has a 41% or higher annual RLTV, at $30.16 vs. $21.37, also in the middle.
The takeaway from the report’s findings is that AI can drive revenue faster, faster, but these programs are struggling to improve their value with customers over time.
