Do you need a mobile app, or a website that works well on phones?
When a small business genuinely needs a native app, when a mobile-first web app is the better buy, and why React Native changes the maths for building both.
By the Wovekit team · · 6 min read
'We need an app' is one of the most common opening lines on our discovery calls, and about half the time the honest answer is 'you need a website that works properly on a phone'. The other half, the app is exactly right and a website would fail. The trick is knowing which half you are in before spending the money.
What a mobile-first web app does well
A modern web app, designed phone-first and served fast, covers most of what customers of a small business need to do: find information, book, pay, check status, upload a document, receive a confirmation. It needs no installation, works from a link in an email or text, is found by search engines and updates instantly for everyone. It is also cheaper to build and far cheaper to change.
If your users are occasional (a customer who books you twice a year) a web app almost always wins. Nobody installs an app for something they do twice a year.
When you genuinely need an app
- Daily or weekly use by the same people. Staff in the field, members, patients, regular customers. Frequency is the strongest predictor of app success.
- The camera, the location or offline use are central, not nice-to-have. Photographing a job, scanning a code, navigating to a site, working in a basement with no signal.
- Notifications are the product. If the whole value is 'tell me the moment something happens', push notifications beat email and text comprehensively.
- One-thumb, one-handed use in the real world. Van cabs, workshops, hospital corridors.
- You already have a web product and users keep asking for the app. Listen to them.
The cost difference, honestly
An app costs more than a web app for the same journeys, for three reasons: two stores with their own rules and review processes, the backend still has to exist (accounts, notifications, admin), and every release goes through review rather than being deployed in minutes. In our bands, a first app with its backend usually sits at £15k to £40k and takes eight to sixteen weeks to reach both stores.
What has changed the maths is building both platforms from one codebase. With React Native and Expo, iPhone and Android ship from the same code with native modules where needed, so one team builds it once and every future change costs half what it did when iOS and Android were separate projects.
A practical way to decide
- 1.Write down who uses it and how often. Weekly or more: lean app. Monthly or less: lean web.
- 2.List the phone features it depends on. Camera, offline and push are strong app signals. A form and a payment are not.
- 3.Ask whether the people who matter will install it. Staff will because you ask them. Customers only will if the value is obvious and repeated.
- 4.Check whether the system behind it exists. Often the real project is the backend and the admin, and the app is the last mile.
Starting with the web and growing into the app
For many businesses the right sequence is a mobile-first web app first, the native app second once usage proves the case. That only works if the backend is built as a proper API from day one so the app can plug into the same accounts, data and notifications. We design them together even when we build them a year apart.
If you are unsure which side you are on, a discovery call is the quickest way to find out. Bring the list of who uses it and how often; that usually settles it in the first ten minutes.
Short answers.
How long does an app take to reach the stores?
Eight to sixteen weeks from one codebase to both the App Store and Google Play, including the backend behind it and the store review back-and-forth.
Is React Native good enough for a serious app?
Yes for the overwhelming majority of business apps. It is a native app, not a website in a wrapper, and we drop to native code for a specific module when camera, maps or performance demand it.
Can we start with the web and add an app later?
That is often the right order. If the backend is designed properly from the start, the app becomes another client of the same system rather than a rebuild.