How Much Does Mobile App Development Cost, and Do You Need an App at All?
On our price list an app starts at $30,000, but the first question we ask has nothing to do with price. Most people who come to us for an app could do without one, and the ones who really need it often budget nothing for what happens after launch. Here is what the price is made of, where the hidden costs sit, what AI has changed, and how to tell whether you should start at all.
SHORT ANSWER
The short answer
A mobile app can easily cost $30,000 or more, and it can also cost much less, especially now that AI knows Swift and Kotlin well. Our price list for mobile app development starts at $30,000. That is an entry point: by Clutch's figures, the average app project in the US costs around $90,000. The exact figure depends on the screens, flows and integrations, and we discuss it openly once those are clear.
An honest answer to "how much does an app cost" starts with a different question, though: do you need one? In our experience, the people who ask for an app most often are the ones who do not need it yet. Some have a budget for the build and nothing after it. Some have already put something together with AI and have no plan for getting it in front of people, and no clear idea of how secure it is. Others want an app because a competitor has one, while their website is not bringing in customers yet.
| What you are building | What to expect |
|---|---|
| First version of a product (MVP) | MVP development from $20,000 on our price list, with only what tests the idea |
| A full app | From $30,000 on our price list: iOS, Android, server, admin panel, store publishing |
| Developers for your own team | Mobile app developers at $55 to $100 an hour depending on level |
| After launch | Support, updates for new iOS and Android versions, the server and advertising are paid separately every month |
The figures come from our price list as of November 2026. They will change, while the logic stays: the build is the entry ticket, and an app costs money every month for as long as it is alive.
WHO ASKS
Who does not actually need an app
Most requests reach us with one of three stories, and in none of them is an app the first step.
"We want what the big brands have." A well-known chain has an app, and an app seems to make a business look serious. But people install that chain's app because they order there every week. If customers come to you once a year, or once in a lifetime, nobody will give your icon space on their home screen. That business needs a fast website that works well on a phone and shows up in search.
"We have enough for the build." The budget runs up to launch day and is empty after that. The app appears in the App Store and Google Play, and nobody finds it among the millions of others. With no advertising, no customer base to tell about it, and no money to fix what the first reviews point out, it goes stale within a year.
"We already built it with AI." A founder put together a working prototype over a few evenings and asks how to get it into the stores. That is a good start, and we will come back to it. In most cases, though, nobody has thought about how people will find the app or who is responsible for the user data flowing into it.
An app earns its place when people use you often and the phone gives them something a website cannot: order status notifications, the camera, location, offline use, a repeat order in two taps. Food delivery, restaurant chains with regular guests, a store people buy from every week, a service used daily. We covered this in detail in when a business actually needs a mobile app.
People install an app when they come back every week.
WHAT YOU PAY FOR
What the price of an app is made of
The client sees screens on a phone, while most of the price goes into what the phone does not show. A typical app has these parts:
| Part | What goes into it |
|---|---|
| Design | User flows, screens for different sizes, error and empty states, icons, preparation for Apple and Google requirements |
| iOS | The app for iPhone and iPad, support for several system versions, signing and publishing |
| Android | The same for Android, across far more devices and versions |
| Server (back end) | The API the app gets its data from, the database, payments, notifications, integrations with your business systems |
| Admin panel | Where your team changes menus, prices, promotions and orders and sees customers |
| Testing | Checks on real phones, on a weak connection, under load |
| Publishing | Developer accounts, store listings, screenshots, a privacy policy, passing review |
The price moves most with the number of screens and flows. An app with a catalog, a cart and checkout differs several times over from one with a personal account, a loyalty program, live courier tracking on a map and a support chat, even though both are "just an ordering app."
The second big variable is technology. Native apps, written separately in Swift for iPhone and in Kotlin for Android, give the best speed and full access to the phone. Cross-platform frameworks such as Flutter or React Native let you write most of the code once for both systems. Which one fits, and where each approach runs out, is covered in PWA, native app, React Native or Flutter.
ANDROID
Why Android often takes more of the budget
Here is the paradox that surprises clients: the biggest share of the budget often goes to Android, not to iPhone. Android phones are cheaper, so it seems an app for them should cost less too.
The reason is variety. Apple releases a few models a year, and most iPhone owners update their system quickly. Android runs on thousands of models from dozens of manufacturers, with different screen sizes, different memory, their own layers on top of the system, and old Android versions that stay in people's hands for years. Each of these combinations can draw a screen differently, treat background tasks or notifications in its own way, and some manufacturers shut apps down aggressively to save battery.
So on Android more time goes into layouts for different screens, testing on real devices and supporting older versions. In a delivery app this is critical: if a courier on an old phone does not get the new order notification, the business loses money every day. Even where most customers carry iPhones, couriers, drivers and staff often work on inexpensive Android phones, so dropping Android or giving it less care is not an option.
THE SERVER
An app does not live without a server
The app on the phone is only the shop window. Menus, prices, stock, orders, payments, rewards and purchase history live on a server, and the app talks to it every second. That is why every app that does something useful has a back end, and the back end is often more complex than the app itself.
This is where the problems clients do not expect show up. We have seen all of them: the server cannot take the load at peak hours, a payment hangs on the payment provider's side, notifications arrive late because the job queue is full. A restaurant app runs perfectly on an ordinary day, and then at seven on a Friday evening, with orders coming one after another, it slows down exactly where the guest is ready to pay.
So a server for an app is designed for peak load, not for an average day, and its hosting and monitoring go into the budget. If you already have a website with a product and order database, the app is often built on the same back end. We most often use Laravel for this, and how that work is priced is shown in how much Laravel development costs. One system for the website, the app and the admin panel is cheaper to maintain than three separate ones.
An app is not as simple as it looks from the screen. We regularly see projects where the budget was counted by screens, and nobody remembered that behind every "pay" button there is a server that has to answer within a second and never lose an order.
AI AND VIBE CODING
What AI changed and where vibe coding breaks
AI really has changed mobile development. Models know Swift, Kotlin and the popular frameworks well and write standard screens, forms and server requests quickly. An experienced developer working with AI gets through the routine part noticeably faster, which is why an app can cost less today than it did two years ago.
It is a different story when an app is put together with AI by someone who does not understand architecture. The first version looks like it works: screens open, buttons respond. The trouble shows up later. Each new feature is added as a patch on top of the last one, and code made of patches eventually stops taking changes: fix one thing and another breaks. We wrote separately about what to do with an MVP built with AI: keep it, fix it or rebuild it.
A mobile app carries one more risk founders rarely think about. It is installed on someone else's phone, and anyone can unpack it and look inside. Keys to the payment provider, the database or the email service, which AI happily writes straight into the code, become available to everyone. Improper credential usage is first on the OWASP list of top mobile app security risks, and insecure data storage on the phone is on the same list. An app can work beautifully and at the same time hand your customers' data to anyone who knows where to look.
A working prototype and a secure app are different things.
That is why, with a prototype like this, we start with an MVP review (from $3,000 on our price list): what in the code can stay, where the security holes are, and what happens to the architecture when there are a hundred times more users.
THE STORES
App Store and Google Play: what they charge and what they reject
To appear in the stores, an app needs developer accounts in the company's name. Apple charges $99 a year for the Apple Developer Program, and Google charges a one-time $25 to register in Google Play Console. These are small sums next to the build, but the accounts must belong to your company, not to the developer. Otherwise you do not own your own app.
Then come the commissions. If the app sells digital goods or subscriptions, Apple and Google take 15% to 30% of every payment. A small business earning under a million dollars a year can pay Apple 15% through the App Store Small Business Program, and Google takes 15% on the first million each year. For most of our clients this matters less than it sounds: Apple's rules say plainly that physical goods and services consumed outside the app are paid for by other means. Food, store products or a doctor's appointment go through your own payment provider with no store commission.
The rule that filters out weak app candidates most sharply sits in section 4.2 of the App Store Review Guidelines: an app should offer features, content and an interface that go beyond a repackaged website. If your app is a website in a wrapper, Apple is entitled to reject it. Put simply, if you have nothing to add to your website, the store will notice too.
Review takes anywhere from a few hours to a few days, and it repeats with every update. A first submission almost always comes back with questions: there is no account deletion option, it is unclear why the app needs location, the privacy policy does not match what the app actually collects. Plan the timeline with that in mind.
PWA
What about a PWA?
A PWA, or progressive web app, is a website that can be added to a phone's home screen and behaves almost like an app. On paper it is the perfect compromise: one build, no stores and no commissions.
In practice a PWA rarely works as a replacement for an app, especially on iPhone. Apple shows no install button: a person has to find "Add to Home Screen" in the Share menu on their own, and almost nobody does. Notifications on iPhone work only after that manual install. In early 2024 Apple announced that in the EU home screen web apps would no longer open as standalone apps, and backed down only after a wave of criticism. The technology is not dead, but Apple is clearly in no hurry to develop it.
So we treat a PWA as an addition to a good mobile website: for a catalog, an internal team tool or an account area people reach from a link. If you need notifications that reach every customer on an iPhone, a PWA will not do that job. The full comparison is in our article on PWA, native and cross-platform apps.
PROMOTION
Without advertising, nobody finds the app
This gets talked about least, and it matters most. An app in a store does not bring customers by itself. To get installs, people have to hear about it: install ads, an email to your customer base, banners on the website, a courier telling the customer about a discount on their first in-app order. All of that is money that belongs in the budget next to the build, not in "someday later."
And the fight for attention has become much harder. Sensor Tower estimates that about 560,000 new apps were released on the App Store in the first half of 2026, twice as many as a year earlier, while downloads over the same period grew by only 2% (as reported by 9to5Mac). Much of that flood is linked to vibe coding. The money is even less forgiving: according to RevenueCat's report, which covers tens of thousands of subscription apps, only about 20% of newly launched apps reach $1,000 in revenue within their first two years, and only 5% reach $10,000. Failure here is the majority outcome, not the exception.
A simple test: if your website is not being promoted today and does not bring orders, an app will not bring them either. An app does not create demand. It serves the customers you already have more conveniently. A restaurant chain with thousands of regular guests gets more repeat orders and pays less to delivery aggregators. A new business without a customer base gets an icon nobody installed.
So if marketing is not planned and your website is not promoted, you may not need an app yet. The money set aside for it will come back faster through the website, advertising and search promotion, and the app is worth revisiting once you have a base of customers who would find it more convenient.
OUR EXPERIENCE
What we built and what it taught us
We have built a delivery app, a shopping app for a large online store and apps for well-known restaurants. We cannot show them: all of them are under non-disclosure agreements, and the clients prefer not to name their contractors publicly. But they work, and people order through them every day.
Those projects left us with a few rules we now apply from the first conversation.
- Server and processes first, screens second. A delivery app does not break at the button. It breaks where the order moves to the kitchen, to the courier and into the accounts.
- Android gets a margin. Testing on old and cheap phones is its own line in the estimate, not something to "check at the end."
- Accounts, keys and code belong to the client's company from day one. An app published from the contractor's account does not belong to the business.
- Advertising is planned together with the launch. A launch date without a plan for how customers will hear about it is just a date.
After launch an app needs attention every month: Apple and Google release new system versions every year, and an app has to support them to stay in the stores. We showed how to price that kind of work in Laravel support and maintenance, and the logic is the same for mobile apps.
HOW TO START
How to price your own app
An exact price cannot be named until there is a list of screens, flows and integrations. Any figure from a website calculator without that list is a guess. This is how we go about it.
First, a short conversation about the business: who your customers are, how often they buy, what you have now, how much you are ready to put into promotion. In a strategic session ($500, credited toward the project within 30 days) we will tell you honestly whether you need an app or whether the money works harder in your website. We talk people out of apps more often than we take them on, and we consider that part of the job.
If an app is the right call, the next step is a discovery phase (from $4,000): flows, screens, server architecture and an estimate by part. After it you know the price and the timeline rather than a range, and you can choose where to start. Often the right start is an MVP with one main flow, not a full app with a dozen features. We explained why an MVP is the cost of entry, not a minimal product.
If you already have your own team and lack a mobile developer, you can bring in mobile app developers by the hour. And if a prototype has already been built with AI, start with a review of what exists: it costs less than building on top of code nobody has checked.
Building the app is only half of the budget. How much to set aside so people find it, install it and come back, and why ads at a few dollars a day do not work, is in what it costs to market a mobile app.
FAQ
Frequently asked questions
How much does it cost to develop a mobile app?
On our price list a full app starts at $30,000 and a first product version at $20,000. The price depends on the number of screens, flows and integrations, and on whether a server already exists. An exact figure comes after the flows are described, not from a calculator.
Why can Android development cost more than iPhone?
Because of device variety. Android runs on thousands of models with different screens, system layers and versions, so layouts, testing and support for older phones take more time. Couriers and staff often work on Android phones, so it is not the place to save.
Does my business need a mobile app?
It does if customers use you often and the phone gives them something the website cannot: notifications, the camera, location, a quick repeat order. If customers come rarely and your website is not bringing them in yet, start with the website and promotion.
Can I build an app myself with AI?
A working prototype, yes. Without architecture, though, code made of patches stops taking new features, and keys and user data often turn out to be unprotected. Before publishing such an app and letting customers in, have developers look at it.
How much do the App Store and Google Play charge?
An Apple developer account costs $99 a year, and Google Play costs $25 once. On sales of digital goods and subscriptions the stores take 15% to 30%. Physical goods and services consumed outside the app are paid through your own payment provider with no store commission.
Can a PWA replace a mobile app?
Mostly no, especially on iPhone. Apple shows no install button, and notifications work only after a manual add to the home screen, which very few people do. A PWA fits a catalog or an internal tool, not an app whose notifications the business relies on.
What does an app cost after launch?
Every month you pay for the server, support, updates for new iOS and Android versions, developer accounts and advertising. Support on our price list starts at $1,000 a month, and the promotion budget depends on the market and should be planned before launch.
Sources
- Apple: App Store Review Guidelines (sections 3.1.3 and 4.2)
- Apple Developer Program: membership and fee
- Google Play Console: service fees
- RevenueCat: State of Subscription Apps 2025, share of new apps reaching $1,000 and $10,000 in revenue
- OWASP Mobile Top 10: top security risks for mobile apps
Not sure whether your business needs an app, or what it would really cost? Let's look at it together.
One notification a week with the new PERETZ articles. Nothing else.
Related Reading
-
17. 09. 2026
When Does a Business Actually Need a Mobile App?
-
02. 11. 2026
PWA, Native App, React Native or Flutter: Which Fits in 2026?
-
22. 10. 2026
You Built an MVP With AI. Keep It, Fix It or Rebuild It?
-
15. 11. 2026
Laravel Support and Maintenance: What It Costs per Month and What Is Included
-
15. 11. 2026
Mobile App Marketing: What It Costs to Get People to Use Your App