The average mobile app earns somewhere around $1.40 per download. That number is quoted constantly and it is close to useless, because the spread underneath it runs from about $0.10 to $24.00 depending on nothing more than which category you shipped into.
Here is the more uncomfortable version. Roughly four in five apps never clear $1,000 a month, and it is rarely the install count that decides which side of that line you land on. It is whether the revenue model matches what you actually built.
This is what each model pays, what it costs, and how to tell which one is yours.
A note on where these numbers come from
Every figure below is an industry benchmark aggregated from public reporting — app-store analytics vendors, ad-network rate cards, and the subscription analytics firms that publish annual reports. Treat them as ranges that tell you the shape of a model, not as forecasts for your app.
They vary by year, by geography, and by who is doing the counting, and anyone quoting one to three decimal places is quoting a spreadsheet, not a fact. Your own first thousand users will beat all of it.
Why "per download" is the wrong unit
A download is not a user, and a user is not revenue.
The average active user generates around $0.04 per month across all categories. Games do better at roughly $0.32 in daily revenue per active user. Neither number moves at all when you get more installs — they move when the people who installed keep opening the app.
Which is why apps with millions of installs and a broken retention curve routinely earn less than apps with a tenth of the installs and users who show up on Tuesday. If you only track one thing, track the second one.
Category sets your ceiling before you write a line of code:
| Category | Typical revenue per download |
|---|---|
| Social and messaging | $0.30 – $1.00 |
| Gaming | $0.10 – $1.50 |
| Education | $1.00 – $3.00 |
| Entertainment and streaming | $1.50 – $4.00 |
| Productivity | $2.00 – $4.00 |
| Health and fitness | $2.00 – $5.00 |
| E-commerce | $15.00 – $24.00 |
An e-commerce app earns fifteen to eighty times what a social app earns per download. No amount of paywall optimization closes that gap, because the gap is not about monetization at all — it is about whether money was already changing hands inside the product.
Freemium and in-app purchases
Free to install, revenue from the small minority who buy something.
Median conversion from download to paying user sits around 2%, with most apps landing between 2% and 5%. So the model only works if the 95% who never pay are cheap to serve and stay long enough to make the 5% worth having.
Pays: a lot, from very few people. Global in-app purchase spend averages about $0.50 per user per month, and roughly twice that on iOS as on Android. Costs: you are now designing two products — the one people use and the one that sells.
IAP does not generalise. It is excellent for a game with a progression loop, and close to useless for a VPN, a photo tool, or a TV remote. The mechanic has to already exist in the product; bolting a store onto something with nothing to sell produces a store nobody visits.
If you are weighing it against the alternatives, what each model costs your users is the comparison that matters more than the revenue table.
Subscriptions
The best revenue per user on this list and by some distance the hardest to earn.
Fewer than 1 in 50 downloads becomes a paying subscriber within the first month, and under 0.1% of people actively using a free app ever convert. More than half of new subscription apps make under $1,000 in their first year.
What separates the ones that work is retention, not acquisition. Monthly plans keep about 20% of subscribers after a year; annual plans keep about 41%. Nearly a third of annual subscriptions cancel within the first month, which means your onboarding is doing more revenue work than your pricing page.
Subscriptions fit apps that are a habit and where the alternative is worse — streaming, coaching, anything with a library that grows. For a utility opened twice a month, the conversion rate will not clear the support burden you just signed up for.
Advertising
The default for everything free, because 95% or more of apps on both stores are free and something has to pay for them.
Formats are not interchangeable. Rewarded video earns roughly $10.50 eCPM globally against about $2.80 for a banner — close to four times the revenue for an ad the user chose to watch. Interstitials sit near $4.80.
Geography moves it again: North American audiences run around $6.50 eCPM, Europe near $5.00, Asia-Pacific near $4.50. The same app with the same daily actives earns materially different money depending on where those users live.
Pays: highly variable, and it is somebody else's number. Fill rate and eCPM move without warning, and one OS privacy change can remove a third of it overnight. Costs: attention, session length, and a permanent argument between your designers and your revenue.
Practical ceiling: a US casual game running a sensible mix earns roughly $0.05 to $0.20 per daily active user. Cap rewarded video at three to five views a session and keep interstitials to one every few minutes — past that you are trading retention for eCPM, and retention is the thing that compounds.
Paid downloads
About 3% of Google Play apps and 6% of App Store apps charge upfront, and that ratio is the finding.
Charging before install works for a small set of products with a reputation already built — the store listing is doing the entire job of selling. One developer's weekly downloads fell from the thousands to about forty after adding a $0.99 price. The product did not change.
Users who pay upfront do retain better and churn less. That is real. It is also a smaller audience, and both stores take 30% of it.
Bandwidth sharing
The user opts in, and a capped share of the bandwidth their device is not using gets relayed for public web data collection. No ad renders. No paywall appears.
It pays up to $0.50 per user per month. A user counts for a day once their device has been sharing for 60 minutes in that day, each qualifying day is worth one day's share of the monthly rate, and the month is the sum of its days. The arithmetic is written out here, including what happens when your install base moves mid-month.
That is not subscription money and we would not pretend otherwise. What it is: revenue that arrives without you changing a single screen, on top of whatever you already run. Devices on hosting, datacenter or VPN IPs are not counted.
It is also the model with the most to be suspicious about, because it is the only one on this list that does something on a user's device that the user cannot see. The questions worth asking are about caps, consent, and where the traffic goes; they are the same ones we would ask, and how it works is our answer to them.
Hybrid, and why most apps should be
Only about one app in ten runs a genuinely mixed model, and gaming is most of that ten.
The logic is segmentation. Subscribers want no interruptions. Free users will trade attention for a reward. A small group will buy something outright. One model cannot serve all three, and picking one means writing off the other two.
The number worth knowing: people who engage with rewarded ads are several times more likely to also make a purchase. Ad viewing and buying are not substitutes, which is the assumption that keeps developers from combining them.
The failure mode is placement. A rewarded ad at a friction point reads as punishment, an IAP that tilts competitive balance costs you goodwill you cannot buy back, and a subscription without ongoing value does not renew. Add one layer at a time and watch ARPU, lifetime value and churn together — if one stream is eating another, that trio shows it before the revenue line does.
What to actually do this week
- Work out your revenue per active user, not per install. If you have never calculated it, that is the finding.
- Look up your category's range above. If you are inside it, your problem is retention. If you are far below it, your problem is the model.
- Pick the model that matches what your product already does, not the one with the best headline number.
- Add exactly one revenue stream and give it a month before adding another.
- Whatever you add, ship it behind a flag to 1% of users first and watch your crash rate and your retention curve for a week.
Step 5 is the one people skip, and it is the only one that cannot be undone cheaply if you were wrong.
Your first 100 devices go live today.
No sales call, no approval to start. Sign up, drop the SDK in, and watch the first device appear in your dashboard.
Create your free accountFrequently asked
How much does the average app make per download?
Around $1.40 across all categories, but the useful number is your category's range rather than the global average. Social and messaging apps earn roughly $0.30 to $1.00 per download, gaming $0.10 to $1.50, productivity $2.00 to $4.00, and e-commerce $15.00 to $24.00. A download is also not the right unit — the average active user generates about $0.04 per month, and that figure moves with retention rather than with install count.
Why do apps with millions of downloads still make no money?
Because downloads measure interest and revenue measures use. An app with a million installs and a broken retention curve has very few people to monetize on any given day, while an app with a tenth of the installs and users who return weekly has a real audience. Roughly four in five apps never reach $1,000 a month, and the deciding factor is usually whether the revenue model fits the product rather than how many people tried it.
What percentage of app users actually pay?
About 2% of downloads convert to a paying customer in a freemium app, with most apps between 2% and 5%. Subscriptions are harder: fewer than one in fifty downloads becomes a subscriber within the first month, and under 0.1% of people actively using a free app ever convert. Any model that depends on purchases has to be designed around the 95%-plus who will never make one.
Which ad format pays the most in mobile apps?
Rewarded video, at roughly $10.50 eCPM globally against about $2.80 for a banner and $4.80 for an interstitial. The gap is close to four times between rewarded video and banners. Geography changes it again, with North American audiences around $6.50 eCPM compared with roughly $4.50 in Asia-Pacific. Rewarded formats also cost the least retention, because the user chose to watch.
Can I combine several monetization models in one app?
Yes, and only about one app in ten does. The reason to combine them is that different users monetize differently: some will subscribe, some will trade attention for a reward, and a few will buy something outright. People who engage with rewarded ads are several times more likely to also purchase, so the two are not substitutes. Add one stream at a time and watch ARPU, lifetime value and churn together, because a new stream that cannibalises an old one shows up in that trio before it shows up in total revenue.