Plain source, for humans in a hurry and machines in general.
Most app ideas arrive as “we need an app”. Some of them do. Many need a website that is fast on a phone, and would be better off without the store, the review queue and the yearly maintenance an app brings with it. Five questions usually settle it.
How often will people open it?
An app has to earn a place on the home screen, and frequency is what earns it. Something people use several times a week, like booking classes at their gym, checking in for shifts or logging work in the field, justifies an install. Something they use twice a year, like booking a dentist or checking an invoice, does not.
Every install is also a step where people drop off. A website works from a link in an email, a QR code on a poster or a search result, the first time, with nothing to download.
Does it need the phone itself?
Websites can do more than people assume. They can take photos, read your location while the page is open, play video, take payments and, since iOS 16.4, send push notifications once the site has been added to the home screen.
What they cannot do reliably is anything in the background or close to the hardware: tracking location while the app is closed, talking to Bluetooth accessories, reading NFC tags on an iPhone, adding widgets, or working with health data. If the product depends on any of those, it is an app.
Does it have to work without signal?
A website can cache pages and keep working offline for a while, but the browser decides how long that data is kept. Safari can clear a website’s stored data after seven days without a visit, unless the site has been added to the home screen.
For a field team doing inspections in a basement or a car park, that is not good enough. Work that has to survive a dead zone and sync later belongs in an app, where the data stays on the device until it is safely on the server.
How will people find it?
Search engines index websites. They do not index what is inside an app, only its store listing. If people find you by searching for what you do, the website does that work, and an app cannot replace it.
An app helps once people already know you and come back often: it is one tap away, and a notification can bring them back. That is a retention tool, not an acquisition tool, and it works best on top of a website that does the finding.
What does it cost to keep?
The build is the visible cost. The rest arrives every year:
- Developer accounts. Apple charges $99 a year for its developer program. Google Play charges a one-off $25.
- Operating system updates. iOS and Android each ship a major release every year, and Google Play requires updates to target a recent Android version.
- Review on every release. Each update, including a one-line fix, waits for store review before users get it.
- Commission on digital goods. Selling digital content or subscriptions inside the app costs 15 to 30 percent in store commission. Physical goods and real-world services, such as a class booking, can use your own payment provider.
A website changes the moment you publish, with no review, no commission and no yearly fee beyond hosting.
Is there a middle ground?
Yes: a fast website that can be installed. Added to the home screen, it opens without the browser around it, can store what it needs for offline use and, on iPhone and Android, can send push notifications. For many products it is the right first version.
Built on a proper API, it is not a dead end either. If usage proves an app is worth it, the native app becomes a second client of the same back end, with the same accounts and data, instead of a new product. How we choose between native and shared code for that app is in the companion post.
A quick test
Build the app if two or more of these are true:
- People will use it at least once a week.
- It needs the camera, location in the background, Bluetooth, NFC or health data.
- It has to work without a connection.
- Notifications are central to what it does, not a nice extra.
- Your customers already look for you in the App Store or Google Play.
If none or one is true, build a fast website first. It is cheaper to make and cheaper to keep, and if it succeeds, it is the foundation the app will need anyway. When the answer is an app, this is how we build them.
Questions people also ask
Can a website send push notifications to an iPhone?
Is an app better for SEO?
Can we start with a website and add an app later?
How long does app store review take?
Who wrote this
Alexander (Sander) van Hooff
Alexander (Sander) van Hooff has been building software since 2015: websites, web applications in Laravel, and native apps for the phone and the desktop. Vuewer is his studio, run from the Alicante region of Spain for clients across Europe.
Work with VuewerRelated reading
Native or cross-platform: how we decide
Native code for each platform or one shared codebase? The four questions that settle it, what each option costs to maintain, and where the back end fits.
Multilingual Laravel: translated routes done properly
Why /nl/pricing loses to /nl/tarieven, how to build translated slugs in Laravel, and the hreflang mistakes that quietly cost you the other languages.
SEO, AEO and GEO: what actually changed
Three acronyms, one shift: search engines now answer instead of link. What that changes about how pages should be written, and what it does not change at all.