Custom Software & Integration

E-Commerce Mobile App or PWA? How to Decide

Share
E-Commerce Mobile App or PWA? How to Decide

Mobile app, PWA or mobile site? Costs, store rules, Shopify options, five decision questions, a hypothetical break-even calculation and a staged roadmap.

"Should we build a mobile app for our customers?" is usually answered first by budget and then by the reflex that competitors have one. Yet an app is a very different investment from a mobile-friendly site or a PWA. This article compares the three options on cost and benefit and gives you a framework to decide for your own business.

Three Options, Three Different Investments

  • Mobile-friendly site: A site that is fast and comfortable on a phone and opens in the browser. Every business needs this as a foundation, and none of the other options replaces it.
  • PWA (Progressive Web App): A layer that gives the same site app-like abilities such as adding it to the home screen and sending notifications. No store approval is needed and there is one codebase.
  • Mobile app: A product published in the iOS and Android stores, built and maintained separately. It can be native (separate per platform) or hybrid (one codebase).

The right question is not "app or site?" but "is our site really good on mobile, and what would an app add on top?" On a site with a weak mobile experience, an app does not fix the problem; it moves it to a second channel. We suggest checking your mobile conversion gap and your page speed first.

What a PWA Gives You, and What It Does Not

In Chromium-based browsers, a site is installable when it is served over HTTPS and its web app manifest defines a name, 192px and 512px icons, a start URL and a display mode (MDN). A service worker is not required for installability; it is used for abilities such as an offline experience.

On iOS the picture is more limited. From iOS 16.4, PWAs can be added to the home screen from Chrome, Edge and Firefox as well as Safari. According to Apple's WebKit team, web push notifications are supported only for web apps added to the Home Screen, and the permission request must follow a direct user action such as tapping a button. In other words, the visitor must first add the site to the home screen, which is an invisible barrier for most customers.

Strengths of a PWA: low cost, one codebase, no store approval wait, and organic search traffic that turns into an app-like experience. Weaknesses: no store presence, installation and notification friction on iOS, and limited access to device features.

In practice a PWA is the cheap way to test the question "do we need an app?" What share of your customers add the site to their home screen, what share allow notifications, and do those users order more often? Those three data points ground an app investment's assumptions in measurement rather than guesswork.

The Real Cost of a Mobile App

App cost is more than the development fee. Break it down:

  • Developer accounts: The Apple Developer Program costs 99 USD per year (shown in local currency by region). Google Play Console charges a one-time 25 USD registration fee. Google also requires personal accounts created after 13 November 2023 to meet certain testing requirements before publishing, so plan the account type up front.
  • Store review: Apple's review guidelines (4.2) expect your app to offer features, content and UI beyond a repackaged website. Putting your site inside a shell may not be enough.
  • Maintenance: iOS and Android versions change every year, and each update means testing, review and release. Two platforms and many devices call for an ongoing maintenance budget.
  • User acquisition: An app does not get downloaded on its own. Moving customers from the site to the app needs an incentive and communication plan.

On payments there is good news: under Apple's rules, physical goods sold in the app but consumed outside it do not require in-app purchase, and you may use methods such as Apple Pay or card payments (guideline 3.1.3(e)). So a retail app selling physical products avoids the store-commission problem.

Your Options If You Use Shopify

Shopify offers tools that carry the heavy lifting of an app. According to Shopify's developer documentation, the Storefront API can power product browsing, search and cart, and Checkout Kit embeds Shopify's own checkout in your iOS, Android or React Native app. Your checkout customisations and accelerated payment methods then work in the app too, and you do not have to build payment security from scratch.

Alternatively, you can move the store to a headless setup such as Hydrogen and feed web and mobile from the same back end, but that is a separate architecture decision. You can also weigh whether you really need an app with the lens in our Shopify app development article, and frame the custom software choice with our off-the-shelf or custom software guide.

Five Questions to Decide

  1. Is repeat purchase frequency high? In frequently bought categories such as groceries, cosmetics or pet supplies, an app can build a habit. For products bought once or twice a year, customers tend to delete the app.
  2. Do you have a loyal customer base? The people who download your app are mostly those who already know you. Without brand awareness, downloads may fall short of expectations.
  3. Is there something the app can do that the site cannot? Device features such as barcode scanning, camera product search, a loyalty card or store stock by location also help the app pass store review.
  4. Does your budget cover maintenance? Finishing the first version and abandoning upkeep means an app that gets removed from the store or runs with bugs.
  5. Is your mobile site ready? If mobile conversion trails desktop by a wide margin in your analytics, close that gap first.

When Does an App Pay for Itself?

Hypothetical example

Say your store gets 2,000 orders a month, with an average basket of 1,000 TL and a 25% contribution margin per order (250 TL). Assume the app's first-year total cost (development, accounts, maintenance) is 600,000 TL, or 50,000 TL a month. The extra orders needed to cover that cost: 50,000 ÷ 250 = 200 extra orders a month.

Now two more assumptions: 15% of orders come from app users, and the app raises those users' order frequency by 20%. At the current volume, app users place 2,000 × 15% = 300 orders; a 20% increase is 60 extra orders, or 60 × 250 = 15,000 TL a month of extra contribution. That falls short of the 50,000 TL monthly cost. To break even, app users would need to place 200 ÷ 0.20 = 1,000 orders, which implies total orders of 1,000 ÷ 0.15 ≈ 6,700 a month.

These figures are entirely hypothetical; recalculate with your own cost, margin and app usage rate. What matters is not a memorised threshold but writing down the app's extra-order target up front and measuring it.

A Staged Roadmap

For most businesses the lowest-risk order is:

  1. Fix the mobile site: Speed, the checkout step, form fields and measurement. This work is the foundation of every option.
  2. Add the PWA layer: Manifest, icons and an add-to-home-screen flow. You can see results within weeks and track how many users add the site to their home screen.
  3. Measure demand: Look at the share of users who add to home screen, allow notifications and buy again. Before sending messages through notifications, review your consent and permission processes; reflect the consent rules that apply to commercial messages such as email and SMS in your notification strategy as well.
  4. Decide on the app: If you can answer yes to the five questions and the break-even calculation above, start with a minimal first version.

Four common mistakes

  • Putting the site in a shell as it is: An app wrapped around a web view can run into store review problems and gives users no value beyond the site. Define a reason specific to the app.
  • Launching without measurement: If you cannot track app users' order frequency, basket value and overlap with the site separately, you cannot see the return on the investment. Plan measurement before launch.
  • Building stock, price and order flows separately: The app and the site should use the same product, stock and order data. Two data sources mean overselling and price mismatches; the fix is usually a shared API layer.
  • Disbanding the team after launch: Once the first version ships, assign a team responsible for bug fixes, OS-version compatibility and user feedback.

Conclusion

A mobile app is not a goal; it is a tool for a specific volume and behaviour. Strengthen the mobile site first, test demand with a PWA, and build the app only when you can show in numbers that it will bring extra orders.

Sources: Apple Developer Program (developer.apple.com/programs/enroll); Apple App Store Review Guidelines, 4.2 and 3.1.3(e); Google Play Console Help, registration fee and testing requirements; MDN, Making PWAs installable; WebKit blog, Web Push for Web Apps on iOS and iPadOS; Shopify.dev mobile and headless documentation (2026). The worked example is hypothetical.

More

Related posts