Is your business ripe for a mobile app in 2023?
A video embedded in the mobile app strategy article.
How to tell whether your business needs a mobile app
The video asks the question; this is how we answer it when a prospective client asks us directly. There are three tests, and a business that fails all three is usually better served by a website, a better booking flow, or an AI agent answering the phone.
The first test is repeat use. An app earns its place on a home screen only if someone opens it weekly — a fitness programme, a loyalty scheme, a field team logging jobs. If the customer interaction is annual, a website does the same work for a tenth of the money. Custom mobile app development starts at $20,000 for an MVP here, and that number should be justified by frequency before anything else.
The second is whether you need the device. Camera, GPS, offline operation, push notifications, Bluetooth, biometrics, background processing: these are the things a browser either cannot do or does badly. Needing one of them is the clearest argument for native development. Needing none of them is the clearest argument against an app at all.
The third is whether the work behind the app already runs. An app is a window onto a process. If the process is a spreadsheet three people update by hand, the app will expose that rather than fix it — and the fix is usually cheaper. This is the most common reason we tell someone not to build yet.
If you are unsure, the scoping phase is the cheap way to find out: one to two weeks, ending in a written fixed-price estimate and a straight answer about whether the thing is worth building. We would rather say no at that stage than take a build we do not believe in.
Related: MVP development and scoping · all services
Has the answer changed since 2023?
The video dates from 2023, and the three tests above have held up. What has changed is the set of alternatives. In 2023 the choice was mostly between an app and a website. Now there is a third option that did not exist in a practical form then: an AI agent that does the job the app was meant to do, without anyone installing anything.
Many app ideas we hear are really about access. Customers want to book, ask a question, check an order or reschedule, and the business imagines an app as the place they would do it. An AI receptionist or chat agent can handle most of those requests on the phone or on the website, at any hour, for a fraction of what an app costs. If the goal is answering customers faster, start there.
The case for an app is now narrower and clearer. It is strongest when the customer relationship is frequent, when the product needs the device, and when being on the home screen is itself valuable: loyalty, habit, membership, field work.
Native, cross-platform or web app
Once the answer is yes, the next question is how to build it. A native app, written separately for iOS and Android, gives the best access to the device and the smoothest performance, at the cost of two codebases. A cross-platform app built with React Native or Flutter shares most of its code across both platforms, which usually cuts cost substantially and keeps features in step. A progressive web app runs in the browser and can be added to the home screen, and it is the cheapest route if the device features you need are ones the browser supports.
For most business apps we recommend cross-platform. The SnapFit corporate fitness platform we built for Snapchat employees and the Go Run social running app are both React Native, shipping to iOS and Android from one codebase. Native makes sense when the app lives or dies on camera, sensor or graphics performance.
What an app costs and how long it takes
An MVP on one platform starts at $20,000 and ships in eight to ten weeks. A production app for iOS and Android with accounts, payments and offline support runs $50,000 to $120,000. Enterprise builds with complex integrations start at $120,000. Every phase gets a written fixed price before work begins.
The build is not the whole cost. Budget for maintenance after launch: operating system updates from Apple and Google arrive every year, and an app that is not kept current starts to break or gets flagged in the stores. Hosting, third-party services and store developer accounts add running costs as well. A realistic plan counts the first two years, not just the launch.
Questions to answer before you commit
Who will open this app every week, and why? If you cannot name the reason, the app will be installed and forgotten.
Which device capability does it need that a website cannot provide? If the answer is none, build the website or the agent instead.
Is the process behind it already working? An app amplifies whatever it sits on top of, including a manual process that does not scale.
How will you get people to install it? An app with no plan for its first thousand users is a cost with no return. Existing customers, a membership base or staff who are required to use it are the easiest starting points.
What does success look like at thirty days? Retention is the number that matters most for an app. Decide how you will measure it before launch, so you can tell whether the app is working and what to change if it is not.
If those answers are clear, an app is likely worth building. If several are not, a website, a booking flow or an AI agent will probably serve you better for now, and an app can follow once the demand is proven.
Related: all case studies · services