AFFWORLD
Start a project
2026-08-20 · 5 min read

Do you need an app, or is a great mobile website enough?

A plain guide for founders: when an app is worth building, when a mobile-first website wins, and the signals that tell you it is time to add the app.

AFFWORLD Insights

"We need an app" is the most expensive sentence in a discovery call. Sometimes it is true. Often a great mobile website does the job for a fraction of the cost and reaches people who would never install anything. Here is how we decide — and how you can, before you talk to anyone.

Start with what the user does, not what you want

Write down the three things a customer does with your product most often. Book a stay. Answer a lead. Check a balance. Now ask of each: does it happen daily? Does it need to work offline? Does it need the camera, Bluetooth, or a notification that interrupts? Does the person expect to pay with one tap?

A mobile website wins when…

  • People arrive from a search, an ad or a shared link and need to act immediately — no install step.
  • Usage is occasional: a few times a month or less.
  • Search visibility matters — app screens are invisible to Google; web pages are not.
  • You are still learning what the product is. Websites change in an afternoon; apps change through a store review.

An app wins when…

  • The same people return daily — agents, guests during a stay, members, device owners.
  • You need push notifications that actually get read.
  • It must work with no signal: at a viewing, in a basement car park, on a plane.
  • It talks to hardware — Bluetooth, NFC, the camera as a scanner.
  • Payments and sign-in should be one tap, every time.
The usual answer: start with a mobile-first website, instrument it, and build the app when the numbers show daily returning users. Several of our clients did exactly this.

The signals that say "now build the app"

Returning visitors on mobile above roughly 30%. Users adding the site to their home screen. Support requests asking for notifications. Agents screenshotting the CRM instead of using it. When two of these show up, the app pays for itself.

What it costs to be wrong

Building an app nobody installs costs the app, the store accounts, the yearly OS updates and the marketing to get installs. Building a website when you needed an app costs a redesign later — usually less. When in doubt, the website first is the cheaper mistake.

Build the app when your users ask for it with their behaviour, not when a vendor asks for it with a quote.

We build both — see web platforms and mobile apps — which is why we can give you a straight answer on the discovery call.

More insights

Keep reading.

Start a project

Read it, then ask us.

Every guide ends where a conversation should start. Send the question the article did not answer.