Native app, PWA or responsive website: how to choose
Three answers to one question, and the difference between them is almost always explained in technical vocabulary that does not help anyone decide. This guide decides by consequence: what you gain, what you lose, and what becomes impossible.
What each one is, in a sentence
Responsive website. Opens in a browser and adjusts to the screen. Does not install, does not receive notifications, does not appear in any store.
PWA. A responsive site with three additions: it can be added to the home screen, it keeps what it has already loaded, and it receives push notifications. It is still reached through the web and still outside the stores.
Native app. Installed from the App Store or Google Play, appears in store search, has full access to the device, and can sell subscriptions through the store's own system.
The table
| Website | PWA | Native | |
|---|---|---|---|
| Appears in store search | no | no | yes |
| Icon on the phone | no | yes, if the person adds it | yes, on install |
| Push notifications | no | yes, with an iPhone caveat | yes |
| Works on a bad signal | no | yes, for what already loaded | yes |
| Camera, GPS, biometrics | partial | partial | full |
| Sell subscriptions through the store | no | no | yes |
| Shipping a fix | immediate | immediate | goes through review |
| Cost to start | lower | lower | developer account and store process |
The five questions that decide it
Answer these and the choice makes itself.
1. Will people look for you in the store? If your customer searches the App Store for "[your business] app", being there is the only thing that solves it. No PWA appears in that search. This is the most legitimate reason of all to go native.
2. Do you need to reach people? Notifications change a business when they are the product: an appointment reminder, an order update, an on-call alert. When they are decoration, they decide nothing.
3. Do you need to charge a subscription inside the app? Inside an app installed from a store, selling digital content outside the store's own system is a policy violation on both, with removal as the penalty. If your model is a subscription consumed inside the app, that is an architecture decision, not a preference.
4. Is the device part of the product? Scanning barcodes all day, recording video, using biometrics: the more the product leans on hardware, the more native justifies itself.
5. How often will you be fixing things? A site and a PWA ship a fix immediately. A store app goes through review, and that difference shows up on the evening you need to fix something at seven on a Friday.
The iPhone detail that decides most cases
Worth isolating, because it is the fact most missing from these comparisons and the one that most often changes the answer.
On iPhone, a PWA only receives push notifications if the person has added the app to their home screen. Opening it in the browser and allowing notifications is not enough: without that step, push does not arrive.
That sounds like a detail and is not. "Add to home screen" is an instruction most people do not understand and a large share never complete. If your strategy depends on notifying iPhone users, and you have no confident way to ask for that step, a PWA will not deliver what you planned. Installed from a store, push works with nothing to ask.
What each one costs, in process
The cost in money is the least interesting part. The cost in process is what bites:
Site and PWA have no process. You publish and it is live. That is why they are the right path for validating anything.
Native has a developer account, identity verification, a store listing, a content rating, a data safety form and a review. On Android, new personal accounts also go through a closed test with a fixed minimum duration. None of it is hard; all of it is calendar time, and most of it waits on somebody else.
The choice does not have to be permanent
We are an interested party here, so treat what follows as description.
The expensive mistake is not choosing wrong, it is choosing in a way you cannot change. People who start with a site built one way and later decide to go to the stores usually redo the work, and that is what makes the decision feel bigger than it is.
On Fabapp the same app is published as a site, works as a PWA with home-screen installation, and can be packaged as a native app submitted to the stores under the owner's own developer account. It is not three products, it is one. That does not make the decision irrelevant, it makes it deferrable, which in practice is better: you find out whether people come back before paying the process cost of being in the stores. All three paths and the rest of what the platform ships are on features.
The recommendation, without hedging
Start with the site or the PWA. It publishes immediately, fixes immediately, and answers the question that matters before any other: does anyone come back?
Go native when one of these is true: people look for you in the store, you need to notify iPhone users without relying on an instruction nobody completes, or you need to charge a subscription inside the app.
Do not go native because it feels more serious. That costs weeks of calendar and changes nothing for the person using it.
On Fabapp the Free plan does not ask for a card, so you can have the web version standing before deciding anything about stores.