Health Apps Make Most of Their Money in the First Two Weeks. The Insight Has to Arrive Before the Trial Ends

Health apps are built on data that gets more valuable over time. A sleep baseline needs a week or two of nights, a recovery trend needs a few weeks, and a change worth telling someone about needs a baseline to change from. Apple’s new readiness score needs seven nights of data before it appears at all [4].

The money does not wait for any of that. RevenueCat’s 2026 report, built on more than 115,000 apps, shows Health & Fitness earning more per install in the first two weeks than any other category, and collecting about two-thirds of a paying user’s first-year value in the first month [1]. Most health apps are paid, or not paid, before their data has had time to say anything personal.

That is the design problem this post is about. First, the benchmarks.

The 2026 benchmarks for health and fitness apps

Source: RevenueCat, State of Subscription Apps 2026 [1]. “Value of a paying user” is RevenueCat’s realized lifetime value, the net value of an average paying user over the period, per payer rather than per install.

For context, the median gaming app earned $0.08 per install by day 14 [1]. Health apps convert well because people arrive with a goal and a reason to pay. The rest of the numbers say how quickly that reason has to be met.

Per 1,000 installs, the business is decided in about a fortnight

Run the medians through 1,000 installs:

• 69 people start a trial (6.9%), and about 57 of them start it on the day they install (82.1%).
• 26 of those trials convert (37.7%). With a few people paying without a trial, about 29 pay within 35 days (2.9%).
• Revenue reaches about $480 by day 14 and $660 by day 60.

Two ratios matter more than any single number:

• About 73% of the first 60 days’ revenue arrives in the first 14 ($0.48 of $0.66).
• About 68% of a paying user’s first-year value arrives in the first month ($24.23 of $35.64).

Both compare medians, so treat them as approximate. The direction is not in doubt: whatever a health app is going to earn from a cohort, it mostly earns while that cohort is still in its first month.

Health value compounds slowly. Revenue doesn’t.

This is the structural problem for any product built on health data. The trial, five to nine days for most apps, starts in the first session [1]. That is roughly the time it takes to collect a first week of sleep, the minimum many scores need before they say anything. A user who connects a watch on day 0 and sees “check back in a week for your insights” has spent the whole trial waiting.

The renewal year has a second cliff. Across all categories, first-year retention on annual plans fell from 31% to 28% between the 2023 and 2024 cohorts. The first month accounts for 35% of annual cancellations, cancellations fall to 3 to 7% a month through the middle of the year, then spike to 9 to 14% in month 12 as renewal approaches. Only about 5% of cancelled annual subscribers come back within a year [2].

So a health app has two moments that decide its revenue: the first week, when it has the least data, and the twelfth month, when it should have the most. Most products are designed for neither.

Four ways to put a personal result inside the trial

1. Start from the history that is already on the phone. An iPhone counts steps by default, so most iPhone users arrive with months of step data in Apple Health, often with sleep and heart rate from a watch. On Android, history depends on which apps have been writing to Health Connect. HealthKit lets an app read the history the user grants. Health Connect gives an app 30 days of other apps’ history by default, and more with the separate history permission [3]. Read it on day 0 and the first screen can be about the user, not a placeholder. We covered the pattern in the cold-start problem in health apps.

2. Give users without a wearable something on day one. In most markets, smartphone owners far outnumber wearable owners; we gathered the figures in smartphone vs wearable adoption. A product whose first insight needs a watch has written off most of its installs before the paywall. Steps, movement and sleep timing from the phone are enough for a first, honest read.

3. Schedule the first trend before the trial ends. Work backwards from the trial length. With seven days, decide which comparison the user will see on day five, such as this week’s sleep timing against their last month from history, and make sure the data to support it exists. A trend the user sees on day five is worth more than a better one on day ten.

4. Treat onboarding as the sales page. With 82.1% of trials starting on the day of install, the paywall sits inside onboarding for most health apps [1]. Showing a personal preview before the paywall, built from history, is the one piece of evidence the user has that the product understands them.

Then design for month twelve

The second cliff is predictable, which makes it manageable. Before the renewal date, the product should show what a year of data has produced: how the user’s sleep, activity or recovery has moved since they joined, measured against their own baseline. That is the one thing a new competitor cannot offer them, because it does not have the year. And because so few cancelled annual subscribers return, the year-in-review has to land before the renewal decision, not in a win-back email after it [2].

Monthly plans behave differently: cancelled monthly subscribers reactivate at about four times the annual rate, 20% against 5% [2]. For a health app whose value builds slowly, a monthly plan is a cheaper way to stay reachable while the data accumulates, at the cost of more month-to-month churn.

Health and fitness apps convert better and earn faster than almost any other kind of app [1]. That is the opportunity and the trap: the revenue arrives in the first two weeks, while the value of health data takes longer to build. The products that win close that gap from both ends, with history and phone data on day 0, a trend before the trial ends, and a year in review before the renewal.