PWA or native app in 2026: what can a PWA do on iOS, and when is native worth it?

Below you will find a dated capability matrix, the fee structure of both stores, and a seven-question test. All platform facts are as of September 30, 2026. Each one links to the vendor's own documentation.

Short answer: seven questions for an online store

Answer yes or no. Count the yeses in the native column.

1. Must push reach iPhone users who never install anything? Web push on iOS needs a Home Screen install first. A hard yes leans native.
2. Is the App Store or Google Play a discovery channel for you? Some categories get found in store search. Yes leans native or a wrapped PWA.
3. Must the app work fully offline? Browsing a cached catalog is fine for a PWA. Offline checkout with later sync is harder on iOS.
4. Do you need Bluetooth, NFC or background location? Yes means native. Safari supports none of these for web apps.
5. Do you ship changes weekly? A PWA updates on the next load. No store review sits in between.
6. Can you fund two codebases for years? Native usually means iOS plus Android. That is two builds and two release trains.
7. Is your mobile site already fast? Then a PWA is a small step. If the site is slow, fix that first.

Zero or one yes in the native column: build the PWA. Three or more: plan native, or a PWA now and native later.

What a PWA can do on iOS in 2026

Apple's rules changed three times in three years. First came push. Then an EU scare. Then open installs.

In February 2023 WebKit announced Web Push for Home Screen web apps in iOS and iPadOS 16.4. The same release added the Badging API. The catch is in the name: push works only after the shopper adds your site to the Home Screen. A Safari tab gets nothing.

In early 2024 Apple announced it would remove Home Screen web apps in the EU under the Digital Markets Act. It then reversed course and restored the feature with iOS 17.4, per Apple's EU developer page as archived on March 9, 2024.

With Safari 26, every website added to the Home Screen opens as a web app by default, and a manifest is no longer required. There is still no install prompt on iOS. Users go through the Share menu, as web.dev's installation guide explains. Expect that manual step to cost you installs.

Capability matrix, as of September 30, 2026

Two rows matter most for retail. Push on iOS depends on the install. Storage on iOS is best-effort, so treat the cache as a cache. Keep the cart and order history on the server. WebKit says a standalone Home Screen app gets the same quota as the browser, and it may be granted persistent mode on request.

Where native still wins

Native still wins in four places.

Hardware. Bluetooth and NFC have no web API on Safari. If your store pairs with a scale, a printer or a loyalty tag, the web will not reach it on iPhone. Android Chrome covers part of this, which splits your audience.

Background work. Native apps can track location in the background with permission, per Apple's Core Location guide. A PWA gets location only while it is open. Order-ready alerts for curbside pickup work better native.

Widgets and system surfaces. Home Screen and Lock Screen widgets come from WidgetKit. A web app has none. For a daily-deal store, a widget is free real estate on the phone.

The store as a channel. Some shoppers search the App Store first. A listing also carries ratings. If your category lives there, being absent costs installs you never see.

These four points are real, and they are narrow. A catalog, a cart and a checkout need none of them. If your roadmap needs all four, go with native iOS and Android apps from the start.

The e-commerce view: speed, conversion, discovery

Speed is where a PWA pays back for a store. It is also the claim most often inflated. Skip vendor slides with big round conversion lifts. Look for measured studies.

One measured source exists. Google commissioned 55 and Deloitte to track retail and travel sites for 30 days in late 2019. A 0.1 s mobile speed gain was linked to an 8.4% lift in retail conversion and 9.2% higher average order value, per web.dev's "Milliseconds make millions" write-up of June 24, 2020. Note the scope. The study measured speed, not the PWA label. A PWA helps only if it makes pages faster.

Conversion also depends on friction. A PWA opens from a link. No install stands between an ad click and the product page. A native app asks for a store visit before the first view. That step loses people, and every lost session is paid traffic.

Discovery cuts the other way. A PWA is indexed like any website, so your product pages rank in search. A native app's screens do not. The store listing is its search page. For most online stores, organic search brings more buyers than store search. Check your own analytics before you assume otherwise.

Repeat buyers are the exception. Loyal customers who order weekly accept an install. For them, native push and a stored login are worth the effort. Serve both groups if the numbers justify it.

What drives the cost

We give no price ranges here. The scope decides the number, and we calculate it individually after a short brief. The structure of the spend, however, is easy to compare.

Fees listed on apple.com and support.google.com on September 30, 2026.

The fee rows surprise many retailers. If you sell sneakers, the store commission does not apply. Apple's Guideline 3.1.3(e) says physical goods must be paid by other means, such as Apple Pay or card entry. Commission bites only on digital goods: gift subscriptions, paid memberships, in-app content. Know which side your revenue sits on.

The biggest line is invisible in the table. It is people. Two native apps need iOS and Android skills, two QA passes and two release trains. A PWA needs one web team. That gap compounds every year you run the product.

On timelines, our PWA page lists an MVP in 4–8 weeks. The final plan comes from the brief.

Hybrid routes: a PWA first, a native app later

You do not have to choose once. Two hybrid routes work well for stores.

PWA first, native later. Launch the PWA and watch the data. How many shoppers install it? How many allow push? How often do they come back? If a loyal segment emerges, build native for them. The PWA keeps serving search and ad traffic. Your APIs, catalog and checkout logic carry over, so the second step costs less.

A PWA wrapped for the stores. On Android this is officially supported. Trusted Web Activity opens your PWA full screen inside an Android app, verified through Digital Asset Links. It needs Chrome 72 or later. You get a Play listing without a second codebase. New personal developer accounts must pass a testing phase first. Read the Google Play closed testing rules before you set a launch date.

iOS is stricter. Guideline 4.2 says an app must offer more than a repackaged website. It adds that apps should not primarily be web clippings. A thin wrapper around your store will likely be rejected. To pass, add native value: native push, a barcode scanner, a wallet pass, a widget. At that point you are building an iOS app. Budget for it as one.

A practical order for most stores:

1. Make the mobile site fast.
2. Turn it into a PWA with an offline catalog and push.
3. Wrap it for Google Play with Trusted Web Activity.
4. Build a native iOS app only if data shows a loyal, repeat segment.