How Do You Price a Subscription App?

Pricing an app is the decision founders spend the least time on and revisit the most. It gets made early, usually by looking at two competitors and picking a number between them, and then it sits untouched while the team optimises everything downstream of it.

The published data is unusually clear on this, partly because the App Store and Google Play constrain price points and partly because the category has been running long enough to converge. Most apps are charging roughly the same amounts, and the decisions that separate them are about billing period, conversion and offer structure rather than about the number itself.

What do subscription apps actually charge in 2026?

Subscription app pricing has settled into three narrow bands. RevenueCat’s State of Subscription Apps 2026, built on 115,000 apps, $16 billion in revenue and more than a billion transactions, puts weekly plans at $4.99 to $6.99, monthly at $7.99 to $9.99 and annual at $29.99 to $39.99. The median annual price is $34.80, up from $31.60 the year before.

Billing period	Market band	Day-60 revenue per install, apps led by this plan	Median one-year retention
Weekly	$4.99-$6.99	$0.32	3%
Monthly	$7.99-$9.99	$0.29	11%
Annual	$29.99-$39.99 (median $34.80)	$0.46	28%

Two things are worth reading out of that convergence. The first is that a price far outside the band is a decision requiring a visible justification, because a user comparing your paywall with the last three they saw will notice.

The second is that the annual band is not twelve times the monthly band. At the $34.80 median against the middle of the monthly band, per RevenueCat’s figures, the annual plan is priced at roughly a third of twelve months of monthly billing, which is the discount the market has standardised on for the commitment.

The upward drift in the median, from $31.60 to $34.80 according to RevenueCat’s 2026 report, is about 10% and lags general inflation. That suggests the ceiling is set by store price points and user expectation rather than by anyone’s costs, which is a useful thing to know before building a pricing model from the cost side.

What is the average app subscription price in 2026?

The average app subscription price in 2026 is $11.88 a week, $17.53 a month and $103.76 a year on the US Apple App Store, according to Business of Apps’ 2026 app pricing benchmarks. Those are means, and every one of them is roughly double the median.

That gap is the most useful thing on this page, because the average is the number most founders look up and the median is the number their buyers are comparing them against.

The gap exists because the distribution has a long right tail and a hard floor. Store price points stop prices falling far below a dollar, while a thin band of expensive apps sits well above the middle and pulls the mean towards itself.

Dating apps carry the highest average price of any category across all three billing periods, per the same Business of Apps data, and that one category is doing a lot of the pulling. Pricing against the mean means pricing against a distribution you are probably not in.

The medians are what to anchor on. Airbridge’s 2026 category benchmarks, drawn from more than 75,000 subscription apps and over $10 billion in annual revenue, put the weekly median at $4.99, the monthly median at $6.68 and the annual median at $29.99.

Note what that does to $9.99, the most common monthly price in the same Airbridge dataset: per those figures half the market charges under $6.68, so $9.99 is a top-quartile price rather than a normal one. Plenty of apps should still charge it.

Few of them realise they are making that choice.

Category	Weekly median	Monthly median	Annual median
All subscription apps	$4.99	$6.68	$29.99
Health and fitness	$4.99	$9.70	$39.99
Education	$5.99	$8.38	$44.99
Gaming	$4.99	$4.99	$20.55

Per Airbridge’s category figures, education carries the highest annual median at $44.99, more than double gaming’s $20.55, while health and fitness monthly at $9.70 runs close to twice gaming’s $4.99. Read the category column rather than the top row.

Per the table above, a fitness app that prices off the all-apps median is charging about 31% less than its own category median, and a game that prices off it is asking about a third more than its own.

One caveat on reconciling the numbers, because the medians here are slightly below the bands in the section above.

Airbridge’s benchmark covers more than 75,000 apps according to its own stated base, while the RevenueCat report used elsewhere on this page covers 115,000 apps and $16 billion in revenue, so the two are measuring overlapping but different populations.

Treat them as bracketing the same band rather than as rival point estimates, and use the category row for a decision rather than any single market-wide figure.

Geography moves it again, which matters if your spend is not all in the US.

Per the same Business of Apps benchmarks, Poland carries the highest average weekly subscription price, the Netherlands the highest monthly and the UK the highest annual, and the page is explicit that its headline figures come from the US App Store and will differ by country and by platform.

So a price copied from a US competitor is a price set for a market you may not be buying installs in.

Should a novelty app charge upfront or sell a subscription?

A novelty app should charge once. The deciding question is whether the app still owes the user something next month. A subscription is a promise of continuing value, and a product that gets used three times and then shown to a friend cannot keep that promise.

On the App Store that is a rule rather than a preference.

Guideline 3.1.2 of Apple’s App Store Review Guidelines requires an auto-renewing subscription to provide ongoing value to the customer, and the examples the section gives are new game levels, episodic content, multi-player support and cloud support, according to 9to5Mac’s coverage of the guidelines when that language was introduced.

A later copy of the same section, recorded by the policy tracker Conduct Atlas, puts it more bluntly: Apple may reject or remove an app if subscribers are not getting ongoing value.

The two quotations differ because the text has been revised more than once, so check the live guideline rather than either version of it quoted second hand.

Developers do get caught by it. Threads on Apple’s own developer forums describe 3.1.2 rejections in which App Review asked for dynamic value delivered over an extended period and was not satisfied by cloud backup and sync, which is more product than a novelty app has.

The guideline quoted in those threads also sets a floor of seven days on the subscription period and requires the subscription to work across all of the user’s devices, per the same forum quotations.

Charging upfront has a cost of its own, and it is the one founders underestimate: the free install goes away.

Around 95.37% of iOS apps and 96.99% of Google Play apps were free to download as of September 2025, according to Statista’s distribution figures, and catalogue counts attributed to 42matters in late 2025 land within a few hundredths of a point of both.

A paid download gives you no trial and no second attempt at the sale, so the whole conversion problem moves into the store listing and the screenshots.

The structure that fits most novelty apps is neither of the two the question offers. It is a free download with a one-time unlock, which the subscription tooling calls a lifetime purchase: RevenueCat’s guide to lifetime offers describes it as a one-time non-consumable in-app purchase granting ongoing access rather than an App Store subscription.

That keeps the install free and keeps a paywall you can test, while selling something a one-trick app can honestly deliver.

It is also a normal thing to do rather than a workaround: per RevenueCat’s State of Subscription Apps 2025, 35% of apps already mix subscriptions with consumables or lifetime purchases, rising to 61.7% in gaming, which is the same pattern behind charging credits instead of a flat subscription.

The retention table earlier on this page is the rest of the argument. Weekly plans keep a median 3% of subscribers at one year, per RevenueCat’s published one-year retention benchmarks.

For an app with real ongoing value that figure is a problem worth fixing. For a novelty app it is an accurate description of the product, and billing weekly for it mostly buys refund requests and a review page that tells every later visitor what happened.

What is the lowest price you can charge on the App Store?

$0.29 in the US store. Apple expanded the App Store to 900 price points running from $0.29 to $10,000, announced in December 2022 and applied to subscriptions first, with other apps and in-app purchases following in spring 2023, according to contemporaneous reports from iMore, BGR and TechRadar.

Per that same coverage, the increments matter more than the floor: prices move in $0.10 steps up to $9.99, $0.50 steps up to $49.99 and $1 steps up to $199.99, the highest 100 price points have to be requested from Apple and are reviewed, and rounded endings such as $1.00 became selectable for the first time.

The floor is almost never the constraint that binds. The commission usually is.

The App Store Small Business Program charges 15% on paid apps and in-app purchases for developers with up to $1 million in proceeds in the prior calendar year, per Apple’s own program page, against the standard 30%.

A $0.99 unlock therefore returns about $0.84 inside the program and about $0.69 outside it, and the $5.84 iOS cost per install reported by AppsFlyer makes both numbers unworkable on bought traffic.

Per Apple’s program page, the threshold is measured on proceeds after commission, and crossing $1 million moves you to the standard rate for the rest of that year.

So a cheap price is a decision about where the installs come from rather than about the price tag. Under a dollar works when the store itself supplies the traffic and the app has no ongoing server cost to carry.

The moment you are buying installs, the arithmetic in the acquisition section above applies and the price has to clear a cost per paying user that starts in the hundreds of dollars rather than the cents.

Should you sell monthly, annual or weekly?

Sell annual as the default and monthly as the alternative, and stop treating weekly as an embarrassment. Per RevenueCat’s 2026 figures, apps whose revenue is led by annual plans post a median day-60 revenue per install of $0.46, and monetise installs roughly twice as well as other groups at day 14.

Retention is where annual really separates: RevenueCat’s published one-year retention benchmarks put the median at 28% for annual subscribers against 11% for monthly and 3% for weekly.

Weekly is the part of this most people get wrong, and the published numbers are not close to the folk wisdom. Weekly-led apps post a $0.32 median day-60 revenue per install, which is slightly ahead of monthly-led apps at $0.29, not a fifth of annual.

Weekly plans convert installs to trials at up to 5.4 times the annual rate, 9.8% against 1.8%, and weekly has grown to 55.5% of all app subscription revenue from 43.3% two years earlier, per the same report.

The problem with weekly is not what it earns, it is that almost nobody is still there a year later.

Category norms override the general case. Per RevenueCat’s 2026 data, gaming sells 82% weekly, productivity 77% monthly and health and fitness leads annual adoption at 68%. Both ends make sense from the inside: a game’s value is immediate and episodic, a fitness plan’s value accrues over months and the annual commitment is part of the product’s promise.

One operational detail before defaulting to annual. Month one accounts for 35% of all annual cancellations in RevenueCat’s 2026 data, and 95% of annual subscribers who cancel never come back. The headline annual conversion number therefore overstates what you keep, and the first four weeks after purchase are worth more attention than the eleven months that follow them.

What does a paying subscriber actually cost to acquire?

A paying subscriber costs an order of magnitude more than an install, and that gap is what decides whether a price is viable. Work it in two steps: what you pay for an install, then what fraction of installs ever pay.

Both halves are published. AppsFlyer data compiled by Digital Applied puts global iOS cost per install at $5.84 in Q1 2026, up 19% year on year, against $1.92 on Android, and IdeaEquity’s 2026 analysis puts iOS at 2.0 to 3.5 times Android across verticals.

Median install-to-paid conversion runs at 2.0% for mid-priced apps, per RevenueCat’s 2026 report. Divide one by the other: a $5.84 iOS install at a 2.0% conversion rate is about $292 spent to acquire one paying subscriber, against an annual plan with a $34.80 median price.

That arithmetic does not resolve by charging $39.99 instead. It is also why, per RevenueCat, only 4.6% of newly launched subscription apps reach $10,000 in monthly revenue within two years. Most apps that work are not buying their installs at median conversion rates, and the ones that are buying installs have either a much higher conversion rate or a much higher price than the median.

Which points at the levers, in order of force. Conversion rate moves payback most, because doubling install-to-paid halves the cost per paying subscriber without touching the price. Billing period moves it next, because annual collects the whole year up front instead of hoping for twelve renewals. The headline price moves it least, since the band is narrow and users compare.

Pricing and acquisition therefore cannot be decided separately. A price that works on organic installs can be unviable the week paid social turns on, and a creative strategy that produces cheap installs of the wrong people makes any price look broken. We run both halves of that as one loop, and the pricing decision sits inside the app launch work rather than beside it.

Does charging less get you more subscribers?

Charging less does not get you more subscribers, and this is the most expensive assumption in app pricing. Per RevenueCat’s 2026 report, median install-to-paid conversion runs at 2.8% for high-priced apps, 2.0% for mid-priced and 1.4% for low-priced ones. The expensive apps convert downloads roughly twice as well as the cheap ones.

Revenue per subscriber then multiplies the gap rather than offsetting it. The same report puts median monthly revenue per subscriber at $35.89 for high-priced apps against $6.67 for low-priced ones, a factor of 5.4, with the high-priced top quartile above $56. A cheap price is not buying volume in exchange for margin. It is producing less of both.

The mechanism is not mysterious. Price is read as a quality signal on a paywall the user looks at for four seconds, and a low price attracts people who were browsing rather than people with the problem your app solves. It also feeds the acquisition algorithm a noisier definition of a valuable user, which makes every campaign after it worse.

Where this stops applying: in categories with a genuine free alternative one tap away, or in markets where the median income makes a $39.99 annual plan, the top of the band reported by RevenueCat, a different kind of decision entirely.

RevenueCat’s own regional splits, per the same report, show top performers reaching 10.4% install-to-paid at the 90th percentile in North America against 1.5% in Latin America, which is a difference in what the market will carry rather than in paywall craft.

Do introductory offers and discounts pay for themselves?

Introductory offers raise conversion reliably and pay for themselves conditionally. According to RevenueCat’s analysis of introductory offers, offering an introductory price increases an app’s chance of reaching the top-performer tier above 10% install-to-paid conversion, which is a meaningful movement given how far the overall median sits below it.

The condition is the arithmetic on the other side. Every dollar discounted lengthens payback unless the conversion lift more than covers the lost lifetime value, and the deeper the discount the better the acquisition creative has to perform to stay solvent. An intro offer is a decision to buy conversion with margin, so it should be evaluated on cost per retained subscriber rather than on the conversion rate it produced.

The framing lever is worth more than it looks, because it costs nothing. In a RevenueCat price test run in parts of Latin America, anchoring the default yearly plan to its monthly equivalent on the paywall lifted annual take rate by 10% and raised trial starts by 30% with no impact on trial-to-paid conversion. Nothing was discounted. The price did not change at all.

One thing to avoid: permanent introductory pricing. An offer that never expires is just a lower price presented as a concession, and it trains the acquisition channel to deliver people who will only ever pay the discounted rate.

How should you test a price change?

Test one variable at a time and break every result down by country before drawing a conclusion. RevenueCat’s guidance on paywall experiments is explicit that aggregate results routinely hide meaningful regional differences, and a global average can conceal a test that is losing badly in your largest market while winning in a smaller one.

Sequence the tests by expected effect. Framing and plan presentation first, since they are free and they move take rate.

Trial length next: according to RevenueCat’s 2026 report, trials of 17 to 32 days convert at a 42.5% median against 25.5% for trials under four days, while the share of apps using trials shorter than four days rose from 42.1% to 46.5%, so the category is moving away from the evidence.

The headline price last, because it is the most constrained by the market band and the most disruptive to reverse.

Give each test enough time to read. A subscription price test that looks decisive after four days is usually reading trial starts rather than paid conversions, and the two can move in opposite directions.

RevenueCat’s own test, cited above, found that exact pattern in the 30% lift in trial starts with no change in trial-to-paid conversion: a real improvement at one stage of the funnel that produced no additional revenue at the end of it.

A note on AI-powered apps, where the pricing question is currently least settled. RevenueCat found these apps generating 41% more revenue per customer while churning about 30% faster than traditional apps, with annual retention of 21.1% against 30.7% for non-AI apps. That combination pushes them toward shorter billing periods and credit-style packaging that the rest of the category has been moving away from.

The short version

Price inside the market band unless the paywall can justify why you are not: $7.99 to $9.99 monthly, $29.99 to $39.99 annual with a $34.80 median, per RevenueCat’s 2026 report. Ignore the market average while you do it.

According to Business of Apps’ 2026 benchmarks the average is $17.53 a month and $103.76 a year, roughly double the medians, because a thin band of expensive apps pulls the mean away from the middle. Anchor on your own category’s median instead.

Default to annual, which retains 28% of its subscribers at one year against 11% for monthly and 3% for weekly, and do not dismiss weekly, which out-earns monthly per install and now carries more than half of all subscription revenue in the category.

Then check the number against what a paying subscriber costs you to acquire, not what an install costs.

At the $5.84 iOS install reported by AppsFlyer and the 2.0% median install-to-paid rate in RevenueCat’s report, that is roughly $292 for one subscriber on a $34.80 plan, which is why conversion rate and billing period move payback further than price does.

And resist the instinct to discount your way to volume: high-priced apps convert installs to paid at roughly twice the rate of low-priced ones and earn 5.4 times the revenue per subscriber, so a price set low to be safe is usually the expensive choice.