Web app or mobile app? Choose by how your customers already behave

Clients ask us this weekly, and most have already decided on "app" before asking. Usually that is the wrong answer, and it costs an extra few thousand dollars.

Real costs, for a baseline

Web app — runs in the browser, one codebase for every device. Done properly, it installs to a phone's home screen and runs full-screen, nearly indistinguishable from an app.

Mobile app — React Native or Flutter, one codebase for both iOS and Android. Roughly 2–2.5 times the cost of a web app for the same functionality.

The difference is not the code. It is:

• Building and signing for two platforms
• Developer accounts: Apple $99 a year, Google $25 once
• App Store review — one to seven days per update, and it can be rejected
• Supporting several OS versions at once
• Users who never update, so you maintain backwards compatibility

That last point is the most expensive long-term and the least discussed. With a web app, every user is always on the newest version. With a mobile app, six months later you are supporting three versions of your own software.

The four deciding questions

1. Does it need to work without a network?

The strongest and clearest reason to build an app.

A courier in a basement car park, stock counts in a dead zone, job records on a construction site. If the work stops when the signal drops, you need an app.

A web app can cache offline to a limited degree, but if offline is the core requirement rather than the exception, do not force it.

2. Do you need push notifications users genuinely must see?

Note "genuinely must see". Web apps receive push on Android. On iOS they do too, but only once the user has installed the web app to their home screen — and most will not.

If your operation depends on somebody seeing a notification within a minute — a driver accepting a job, a doctor paged for an emergency — you need an app.

If notifications are reminders, a Zalo message is cheaper and gets read more in Vietnam.

3. Do you need deep access to phone hardware?

Continuous barcode scanning, NFC, a camera with live image processing, Bluetooth to a printer or scale, background location.

Taking one photo and uploading it is fine on the web. Scanning two hundred barcodes in a row in a warehouse needs an app.

4. Will users open it every day?

This is an economic question, not a technical one.

An app occupies a slot on somebody's home screen. That slot is only worth it if they open it often. For internal software staff use daily, an app makes sense. For something customers use twice a year — checking a warranty, viewing a contract — nobody installs it, and you have paid for an app with no users.

A number worth keeping: in Vietnam, the rate at which people click a link and successfully install an app is typically under 20%. A web app has no step that loses 80%.

If all four answers are "no"

Build the web app. And if you later need an app, you lose nothing — the backend, the data and the business logic all carry over. An app is just another interface layer on top.

That is why we recommend a web app until there is a clear reason for an app: it is not a dead end, it is the foundation.

If one answer is "yes"

Then it is usually web app plus app, not app instead of web.

Customers browsing, viewing products and signing up — on the web, because nobody installs an app to do one thing. Staff operating daily — on the app, because they have a reason to install it.

We ship React Native and Flutter kits that share a backend with the web build, precisely because this split is the right answer most of the time.

The question asked backwards

"Does having an app look more professional?"

In Vietnam in 2026, no. Your customer already has forty apps and does not want a forty-first. A web app that opens instantly from a Zalo link gets used more than an app that requires a store download.

If you are weighing this and are not sure which case you are in, message us on Zalo with answers to the four questions. If all four are "no", we will tell you plainly that you do not need an app — even though the app is the larger contract.