Two ways to fail, compared

Native iOS App or Web App: A Decision Table for a Business With No IT Department

A native app has to be approved by Apple. A web app has to be remembered by your customer. Those are the two ways this decision goes wrong, and for a business your size the second is usually the cheaper mistake to make and the easier one to fix.

November 9, 2026 · 9 min read

published figures · November 2026
The same idea, priced two ways
Native iOS beside a web app, at the starting line
Published
DataCram web app, starting priceFrom $575
Fee to Apple for a web app$0
DataCram iOS app, starting priceFrom $3,750
Apple Developer Program, per year$99
Automatic install prompt on iOS, web appNot supported
Gap between the two starting builds$3,175
published prices and MDN documentation, as of November 2026

The short answer

For most small businesses a web app is the better first build. It starts at $575 with DataCram, works on iPhone and Android from a link, and needs no App Review. Native iOS starts at $3,750 plus Apple's $99 yearly fee, as of November 2026. It earns that when people use it weekly and need the deepest access to the phone.

On this page
  1. Two products with the same icon
  2. The table to hold your idea against
  3. Getting a web app onto an iPhone, as the documentation describes it
  4. Push notifications, the feature that starts this argument
  5. Where native still earns the extra money
  6. Selling a subscription changes the arithmetic
  7. Three made-up companies run through the table
  8. The route that lets you change your mind

Two products with the same icon

A native iOS app is software written for iPhones, delivered through the App Store, and reviewed by Apple before anyone sees it. A web app is software that runs in the phone's browser, reached through a link. Add a small file and it can sit on the Home Screen with its own icon and open without browser bars. People call that version a Progressive Web App, or PWA. Your customer may not be able to tell which one she is holding, and that is the point.

The differences show up in the bills, the rules and the Tuesday morning fixes. A web app can change the minute you decide it should. A native change goes through a new submission. A web app runs on Android with no second build. A native iPhone app does not. And Apple's yearly fee, $99 as of November 2026, applies to only one of them. No, sole proprietors do not get a waiver, by Apple's own account.

The table to hold your idea against

Native iOS against a web app, with vendor facts as of November 2026
Native iOS appWeb app (PWA)
Starting build with DataCramFrom $3,750From $575, depending on complexity
Yearly fee to Apple$99None
Who approves itApple's App Review, with no promise of a yesNobody
How someone gets itApp Store listing or direct linkA link, then Add to Home Screen if they choose
Install prompt on iPhoneThe App Store buttonNone automatic, so you have to show them how
Push notifications on iPhoneYes, with permissionOnly once on the Home Screen, iOS 16.4 or later
AndroidA separate build, plus Google's one-time US$25Same build
Pushing a fixNew version, new reviewPublish it and everyone has it
Charging for a real-world serviceCard or Apple Pay, no Apple commissionYour own checkout
Charging for digital subscriptionsIn-app purchase, 15% under Apple's Small Business ProgramYour own checkout
Best fitWeekly habit, heavy device useEverything else, to start

The commission rows come from Guidelines 3.1.1 and 3.1.3(e) in Apple's App Store Review Guidelines and from its Small Business Program. Google's fee is on its Play registration page. If you sell Android subscriptions, Google's service fee table needs its own careful read, since Google itself says there is no single fee.

Getting a web app onto an iPhone, as the documentation describes it

The old complaint was that iPhones hated web apps. Part of it is out of date. According to MDN's guide on making web apps installable, read in November 2026, iOS 16.4 and later can install them from the Share menu in Safari, Chrome, Edge, Firefox and Orion. On iOS 16.3 and earlier, only Safari could. The technical bar is modest too: a manifest file and a secure address. A service worker, the script behind offline behavior, is not required just to install.

The part still true is that nobody will find it by accident. MDN says the automatic install prompt, known as beforeinstallprompt, is not supported on iOS. So your customer sees no banner offering to install your app. If you want an icon on her phone, you must teach her: tap Share, tap Add to Home Screen. Put that on the page with a picture. Then expect plenty of people to skip it and keep using the link, which is fine. For a monthly customer, a link in a text message is the whole product.

Glowing cyan and blue lines forking from one point toward two phone outlines on a deep blue and violet field, thin data grid, no text
One decision splits into two kinds of bill, and only one of them comes with a yearly fee to Apple.

Push notifications, the feature that starts this argument

The usual reason to want an app is the urge to buzz a customer's pocket. That is possible on the web now, with conditions. WebKit announced on February 16, 2023 that iOS and iPadOS 16.4 added Web Push for Home Screen web apps. Three things have to be true, and a small business should read them as a list of ways to lose people.

  1. 1.The customer has already added your site to her Home Screen, which is the step most people skip.
  2. 2.The site's manifest sets display to standalone or fullscreen, which is your developer's job and takes minutes.
  3. 3.Permission is requested in response to a direct action, such as tapping a button, so a pop-up on page load is out.

Native has a smoother path, but it still needs a yes from the customer, and it still needs her to have installed your app. For the jobs a small business mostly wants to do, such as reminders, confirmations and on-my-way notes, the message that reaches the most people is often a text. That has its own registration and consent rules, covered in automating customer follow-ups. Do not assume push saves you from that work.

Where native still earns the extra money

We would not sell native builds if the case were empty. Apple builds its tools for its own hardware first, and native gets the fuller run of the phone. Support for device features in the browser moves over time, so we will not print a list that goes stale. Instead, open a test page on the oldest iPhone your users carry, then try the exact thing you need. Here is where native tends to win anyway.

A daily habit

Members or staff open it constantly. The extra polish pays off because they feel it dozens of times a week, and they already have a reason to install.

Heavy camera, scanning or background work

Repeated barcode scans, or a job that keeps running while the phone is in a pocket. Test the browser first, then decide with evidence.

Long stretches with no signal

A web app can work offline. A crew in rural areas relying on it all day is a tougher test, and worth the conversation. See the inspection app page.

Customers who expect a store listing

Some audiences may look for you in the App Store by name, such as members of a gym. Ask yours before you assume. Someone with a flooded basement is calling, not browsing.

If your customers would never search the App Store for you, a listing there is a very expensive sign in an empty hallway.

Selling a subscription changes the arithmetic

Say a one-person fitness coach, invented for this example, sells a $20 monthly plan of workouts delivered in an app. Inside a native app, Guideline 3.1.1 points to in-app purchase, and Apple's Small Business Program rate is 15%. On the web, she uses Stripe. Using the rates cited here, as of November 2026, the gap on one member per month looks like this. The figures are our arithmetic.

One made-up $20 monthly subscription, by where it is sold
In-app purchaseWeb checkout with Stripe
Platform commission$3.00 at 15%None
Card fee at Stripe's standard rateNo separate charge$0.88
Stripe Billing, pay as you goNot used$0.14 at 0.7%
She keeps, per member, per month$17.00$18.98

That $1.98 is real but not decisive on its own. Set it against an app's other benefits and the extra build. Guideline 3.1.1(a) says US storefront apps need no entitlement to include outside purchase links, and Guideline 3.1.3(b) has its own conditions for subscriptions bought elsewhere, which is worth reading if you plan on both. For a subscription business, our pages on subscription websites and adding Stripe go further.

Three made-up companies run through the table

Web app, and stop there

  • A solo mobile mechanic whose customers check job status twice a year: a portal that opens from a text
  • A six-person cleaning company whose staff see schedules on mixed phones: one web build covers everyone, and employee apps explains the private routes
  • A bakery taking weekly preorders: a saved link, and possibly a Home Screen icon for the keenest customers

Web first, native if it proves out

  • A small gym whose members book classes several times a week: frequency is there, so measure it on the web and see if members ask
  • A pest firm whose techs scan bait stations all day: test scanning in a browser on their real phones, then decide
  • A language tutor selling lesson packs: start on a web checkout, then weigh Apple's commission against the listing

Notice nobody here starts native. That is not a law. A company whose whole product is the app, with a proven audience, may go directly to native and be right. Before that, though, try the app budget ladder, and if you still want it, read the full iPhone bill.

The route that lets you change your mind

Choosing the web first does not trap you. The expensive part of many small apps is the data, the rules and the connections behind the screen. Built with a sensible API, those can later serve a native app too, though we would not promise how much carries over until we have seen the code. It is the opposite of the wrapper trap on turning a website into an app.

Our published prices, as of November 2026: web apps from $575, depending on complexity, and iOS or Android apps from $3,750. The pricing page lists the rest. Tell us what the thing has to do at (210) 346-0848, and we will say which side of the table you are on. The software we built and run ourselves, including Athena CMS, lives on the web, and we have been building there since 1999.

Questions owner-operators ask

Not as it stands. A web app lives at a link and needs no store. Wrapping one in a thin native shell to submit it runs into Guideline 4.2, which, as of November 2026, asks for features, content and interface that lift an app beyond a repackaged website. If you want a listing, plan real native features. If you only want an icon, a Home Screen install gets you there.

They can, using a service worker, though it is not required just to install one. How well depends on the job and the phone. A form that holds its data until signal returns is a reasonable ask. A full day's work in the field with large photos deserves a real test on real phones before you trust it. Native remains the safer choice for heavy offline use.

Some will, with instructions. MDN notes that iOS has no automatic install prompt, so you must show the steps: tap Share, then Add to Home Screen. Many people will simply use the link, and that is usually fine. Design as if nobody installs it, send the link by text, and treat the icon as a bonus for your keenest customers.

Usually a web app. One build from $575, depending on complexity, opens on both. Two native apps mean more scope, plus Apple's $99 a year and Google's one-time US$25 registration fee, as of November 2026. A mixed crew or a customer base you have not surveyed is one of the strongest arguments for starting in the browser.

Yes, and it is often the sensible sequence. You will learn what people really use, and the data and connections behind the screen can often serve a later native app. How much carries over depends on how it was built, so raise it at the start. Native builds with us start at $3,750, depending on scope, whenever that day comes.

Not automatically. Either one is only as careful as the login, the server and the handling of data behind it. A web app requires a secure address to be installable at all, per MDN. For cards, send payments to a provider's hosted checkout, which Stripe says reduces your obligations without ending them. Ask any builder how accounts and backups work on your project.

Keep reading

Free · No obligation

Tell us the job and we will pick a side

Describe what the app has to do, who uses it and how often. A person at DataCram replies with native or web, the reason, and the published starting price for that route.

Which of these sounds like you? (pick any)

Rather talk now? Call (210) 346-0848.

CallGet a quote