You make money from an app by choosing one revenue model that fits how people already use it (subscriptions, in-app purchases, ads or paid downloads), then building, marketing and charging in that order. Most founders start with the charging step and wonder why nothing happens.
Enjoy Movies Your Way, a US family streaming filters app, reached 38K+ monthly active users on 50K+ downloads by December 2023. That is 76% of downloads still active monthly (ApsteQ case study). I look at that number before I look at any revenue chart. Revenue is a multiplication: active users, times the share who pay, times the price. If the first term is weak, no pricing trick rescues the other two.
Founders usually ask me about the paywall first. I ask about week-four retention first. After 20+ years in growth marketing across 300+ brands, I have seen the same order of operations separate apps that earn from apps that just exist. This post lays out that order, with the store fees, the models and the mistakes that cost the most.
Key takeaways
- Consumers spent roughly $150 billion on in-app purchases in 2024 across about 137 billion downloads (Sensor Tower, State of Mobile 2025), so money is there; it concentrates in apps that keep users.
- Store fees shape your model: Google Play charges 15% on the first $1 million of yearly revenue (Google Play Console Help, 2026) and Apple's Small Business Program cuts its commission to 15% for developers under $1 million in proceeds (Apple Developer, 2026).
- Duolingo ended 2024 with about 9.5 million paid subscribers against roughly 116 million monthly users (Duolingo Q4 2024 shareholder letter, 2025). Even a category leader converts under 10% of its active base to paid.
- Do the work in this order: build it, get it found, get it paid. Skipping the middle step is the most common reason an app earns nothing.
How do apps actually make money?
Apps earn money in four ways: subscriptions, in-app purchases, advertising and paid downloads. Freemium is the most common setup: the app is free to install, and revenue comes from a minority who pay or from ads shown to everyone else. Your category and usage pattern decide which of the four fits, and picking the wrong one is expensive to undo.
Here is how the four compare, using the store rules published by Apple and Google.
| Model | How it earns | Store commission | Fits best when |
|---|---|---|---|
| Subscription | Recurring weekly, monthly or yearly fee | Google Play: 15% from day one. Apple: 30% in year one per subscriber, 15% after 12 months, or 15% flat under the Small Business Program | Users return weekly and get ongoing value (tools, content, learning) |
| In-app purchases | One-off items, credits, unlocks | Standard 30%; 15% on the first $1M (Google) or under the Small Business Program (Apple) | Games, creator tools, apps with consumable value |
| Advertising | Impressions, clicks, rewarded views | No store cut; ad networks take their own share | High session frequency and a large free audience |
| Paid download | One-time price before install | Same as in-app purchases | Niche professional utilities with a clear, easily described payoff |
Fee sources: Google Play service fees and Apple's Small Business Program. Rules change, so check both pages before you set prices.
A real conversion example: Duolingo
Subscription conversion rate is the share of active users who pay. Duolingo is a public company, so its numbers are checkable: about 9.5 million paid subscribers and roughly 116 million monthly active users at the end of 2024 (Duolingo investor relations, 2025). That works out to around 8% of monthly users paying. The free tier is not charity; it feeds the paid tier and an ad business at the same time.
The lesson is the ratio, not the brand. If a market leader with years of tuning converts under one in ten, plan your first model on a low single-digit conversion rate.
A worked revenue formula
For illustration: imagine an app with 10,000 monthly active users. If 3% subscribe at $6 a month, gross revenue is $1,800 a month. Under a 15% store fee you keep $1,530. Run the same 10,000 users through ads at 30 impressions each per month and a $5 payout per 1,000 impressions, and you earn $1,500 before network costs. The two models land close together here. Subscription revenue grows with price and retention. Ad revenue grows only with session volume. Which one you can move depends on your product.
My rule of thumb: if users open the app less than three times a week, ads will not pay the bills. Choose a model where one engaged user is worth more than a few cents a month.
What is the right order to build, get found and get paid?
The order is build, get found, then get paid. Each step has one job: the build proves the app solves a problem, discovery brings real users in, and the payment layer converts the ones who stay. Reverse any two and you spend money measuring noise.
Step 1: Build it (one core loop, not ten features)
Vibe-coding tools now let a non-technical founder produce a working prototype in a weekend, and I think that is a good thing for validating an idea. The prototype is a doorway, though. Production apps need store-compliant payments, analytics events that tell you what users do, crash handling and a backend that survives launch day. That is the gap our app development team usually closes when a founder arrives with a prototype that works on one phone.
Build only the core loop: the single action users repeat that delivers value. Instrument it from day one, so you can answer "who came back on day 7?" with data.
Step 2: Get it found
App Store Optimization (ASO) is the practice of improving your listing (title, subtitle, keywords, screenshots, ratings) so more of the people who search the stores install your app. It is the cheapest source of installs because organic search costs nothing per click. Our ASO service starts here for that reason.
Paid user acquisition comes second. In the LunarLinc case study (a friendship app launched Android first in India), 113 video ads drove 41,420 installs, while 83 image ads drove 468 (ApsteQ case study). The app has passed 50K+ Google Play downloads since November 2023. The creative format changed the cost of finding users by a wide margin, and we only learned that because we tested both at volume.
Step 3: Get it paid
Add monetization once you can see retention. A workable sequence:
- Measure day-7 and day-30 retention for a cohort of real users.
- Pick the model from the table above that matches how often they open the app.
- Show the paywall right after the user's first moment of value, not at install.
- Offer one monthly and one annual price; test a free trial against none.
- Check payout after the store commission, not before.
EarningsCall, a US fintech app for investors, shows why this order pays off. In its 2022 to 2023 campaign, 60% of downloads were monthly active (3K+ monthly active users on 5K+ downloads), and the app now has 100K+ downloads (ApsteQ case study). An audience that returns that often can support a paid tier. An audience that installs once and vanishes cannot, whatever the price.
The data behind app revenue: why retention beats downloads
App revenue concentrates in apps that retain users, so active users, not downloads, set your ceiling. Market-wide figures show a large pool of money, while public company data shows how small a share of users pays. Put the two together and the practical target becomes clear: win a defined group of loyal users instead of chasing install counts.
Market size is large, but concentrated
Sensor Tower's State of Mobile 2025 report put 2024 consumer in-app purchase spending at about $150 billion, on roughly 137 billion downloads (Sensor Tower, State of Mobile 2025). Divide the two and you get about $1.10 of spend per download across the whole market. That average hides a steep spread. A handful of apps and games take most of the money, and most apps take almost none.
Conversion is low even for winners
Duolingo's roughly 8% paid share of monthly users (Duolingo investor relations, 2025) is a ceiling to aim at, not a floor to expect. Fees then trim what is left. At the standard rates in the Apple and Google documentation (30% or 15%), a $10 subscription pays out $7 to $8.50 before your own costs (Google Play Console Help, 2026; Apple Developer, 2026).
What our published case studies show about activity
| App (ApsteQ case study) | Downloads | Monthly active | Active share |
|---|---|---|---|
| Enjoy Movies Your Way (Dec 2023) | 50K+ | 38K+ | 76% |
| EarningsCall (2022 to 2023 campaign) | 5K+ | 3K+ | 60% |
For illustration, using the Enjoy Movies pattern with round numbers: 50,000 downloads at 76% active gives 38,000 monthly users. If 2% of those paid $5 a month, gross revenue would be $3,800 a month. Had the same 50,000 downloads kept only 20% active, the same 2% conversion would produce $1,000. Same installs, same price, same conversion rate; the revenue difference is nearly four times. That is why I spend more of a client's strategy time on retention than on the paywall.
Insight from the numbers: every point of monthly activity is worth more than every point of price. A 10% price rise lifts revenue 10%; moving activity from 40% to 60% lifts it 50%.
If you want to see how we measure this on a real launch, our app marketing work reports active users next to installs, because installs alone mislead.
What mistakes stop apps from earning?
Most apps fail to earn because of five avoidable errors: charging before value, picking a model that clashes with usage frequency, buying installs without watching retention, ignoring store fees in pricing, and testing too few creatives. None of these need a bigger budget to fix. They need a different order of work.
1. Putting the paywall before the first win
A paywall at install asks for money from a stranger. A paywall after the user's first completed action asks for money from someone who has already seen the value. The second version almost always converts better, since the user has evidence instead of a promise.
2. Choosing ads for an app people open once a week
Ad revenue is a volume game. For illustration: an app opened four times a month, showing five ads per session, serves 20 impressions per user per month. At a $5 payout per 1,000 impressions, that user earns you $0.10 a month. A subscription at $5 needs only 2% of users to beat it by a wide margin. Match the model to the frequency.
3. Buying installs without a retention gate
Paid acquisition can fill a dashboard with downloads that never open the app again. Set a rule before spending: scale a campaign only if its users reach your target day-7 retention. The LunarLinc result above (41,420 installs from video versus 468 from images) is a reminder that the creative drives cost, and cost per active user is the figure that matters, not cost per install.
4. Forgetting that stores take a cut
Founders model $9.99 and forget the payout is $8.49 or $6.99. If your margin is thin, check the Apple Small Business Program terms and the Google Play fee schedule before you fix a price, and re-check yearly.
5. Skipping listing work
A listing with weak screenshots and an untouched keyword field wastes every ad dollar sent to it. Fix the page first. It is the cheapest conversion lift you will find, and it is why we put ASO ahead of paid campaigns on most plans.
One more pattern I see in audits: teams manually export data from the store consoles into spreadsheets each week and react late. Setting up automated reports and alerts through AI automation shortens the gap between a retention drop and a fix from weeks to days.
Where app monetization is heading in 2027
These are my predictions, not reported facts. App revenue in 2027 will favor products that combine cheap building, strong distribution and flexible payment routes. Each trend below follows a mechanism already visible in 2026.
- Building gets cheap; distribution gets expensive. When AI tools let anyone ship a prototype in days, store listings fill with lookalikes. The scarce resource becomes attention, so ASO and creative testing matter more than the code.
- Web checkout becomes a real option. Court and regulatory pressure in the US and EU is loosening how apps can link out to their own payment pages. Rules are still shifting, so read the current App Review Guidelines before you build a flow around them. If the fee on a web purchase drops, margins on subscriptions improve.
- AI features change pricing. An app that calls a language model on every request has a variable cost per user. Flat unlimited plans can lose money on heavy users, so I expect more credit-based and usage-tiered pricing.
- Hybrid models win. A free tier with ads, a paid tier without them, and a few one-off purchases on top. Duolingo's mix of subscription and advertising points the same way.
The constant across all four: you still need people who come back. Nothing on that list repeals the retention math above.
Frequently Asked Questions
How much money can a small app realistically make?
It depends on active users, conversion and price, so any single figure would be a guess. I would rather you run the formula: monthly actives times paying share times price, minus the store fee. Start with a low single-digit conversion assumption. If the result does not cover your costs, change the model before you spend on ads.
Should I charge for my app or make it free?
For most new apps I recommend free to install with a paid tier. A paid download removes the chance to prove value, and strangers rarely pay for something they cannot see. The exception is a niche professional tool where the buyer already knows they need it. Even then, test a free trial before locking in an upfront price.
How many users do I need before I add monetization?
There is no magic count. I care more about whether a cohort is still using the app at day 30. Once you can see a stable group coming back, add a paywall and test it. Adding it earlier mostly teaches you that strangers will not pay, which you already knew.
Do I need a developer to make money from an app?
You can validate an idea with AI-assisted building tools, and I encourage that. To take payments properly you need store-compliant billing, subscription handling and reliable analytics. At that point a development partner saves time and avoids rejected submissions. Think of vibe coding as the sketch and a professional build as the structure you can charge for.
Are ads or subscriptions better for a new app?
Subscriptions suit apps people open often and rely on, like tools, learning and content. Ads suit high-frequency, large-audience apps where each user is worth cents. My bias is toward subscriptions for small teams, because you can earn from a modest audience. Check your session frequency first; it usually settles the question.
Conclusion
Making money from an app is a sequence, not a trick. Build the one loop users will repeat. Get found through ASO and tested creative. Then add the paywall once you can see who stays. Fees, pricing and model choice all matter, but they multiply whatever active audience you have earned.
Three principles to keep: measure active users before revenue, match the model to usage frequency, and price from the payout after the store cut. If you want a second pair of eyes on your build, your listing or your acquisition plan, our teams handle all three through app development and app marketing. You can book a free strategy call and we will map your fastest route to the first paying users.
Want a second pair of eyes on your growth?
Book a free 30-minute strategy call. Bring your numbers, leave with two or three moves worth making. No pitch, no deck.
Book a Free Strategy Call