Decide whether an app is justified
People do not install an app to buy once. New customers find you through search, social media and adverts, and they land on your website. An app earns its place on the phones of customers who come back: weekly grocery shoppers, fashion followers, trade buyers, members of a loyalty scheme. Start by counting them.
An app is likely to pay
- A solid share of orders comes from repeat customers
- Customers buy frequently, or from a familiar list
- Notifications would be welcome: order progress, restocks, offers they chose
- A loyalty scheme or membership is central
- Features such as barcode scanning or saved lists would save real effort
An app is unlikely to pay
- Most customers buy once or rarely
- The mobile website is slow or awkward and has not been fixed
- The only reason is that competitors have one
- No budget or person for ongoing releases
- No plan for persuading customers to install it
Fix the mobile website first in any case, since every app customer starts there. A progressive web app can cover part of the ground at lower cost, and our comparison of a PWA and a mobile app sets out where each fits.
Features that justify the app
- Staying signed in. Customers open the app and are recognized, with biometric confirmation for payment. This alone removes the main friction of mobile checkout.
- Notifications. Order progress, back in stock, price drops on saved items and offers the customer opted into. Plan them carefully, with consent and preferences.
- Reordering and lists. Buy again, regular baskets and saved lists placed one tap from the home screen.
- Camera. Scanning a barcode to find or reorder a product, or visual search where the catalogue suits it.
- Wallet and loyalty. A membership card, points balance and vouchers shown in store from the phone.
- Speed and offline browsing. Recently viewed products and the basket available at once, even on a weak connection.
- Links that open the app. A link in an email or advert should open the right product inside the app when it is installed, and the website when it is not.
Choose two or three of these as the reason for the app, and build them well. An app that merely repeats the website gives customers no reason to keep it. Our guide to planning push notifications and customer permissions covers the notification design.
Payments, accounts and app store rules
- 01
Which payment methods?
What to settleCards, Apple Pay and Google Pay, and the local methods your customers use, through a payment provider that supports them in apps.
- 02
Physical goods or digital content?
What to settleThe stores' rules treat them differently. Physical goods and services are paid for with ordinary payment methods. Digital content used in the app generally falls under the stores' own billing systems. Check the current policies for what you sell.
- 03
Can customers delete their account in the app?
What to settleApple's guidelines require apps that let people create an account to let them start deleting it from within the app. Plan the process and what happens to order records you must keep.
- 04
What is declared about data?
What to settleBoth stores require a declaration of what the app collects and why, and they review permission requests.
- 05
Is guest checkout offered?
What to settleForcing registration before a first purchase loses orders. Decide what a guest can do.
- 06
Who holds the developer accounts?
What to settleYour business, enrolled as an organization, with the developers invited as users.
Store policies change, and reviewers apply them to each submission. Read the current guidelines for your category before finalizing payment and account requirements, and leave time in the plan for a first review to raise questions.
A worked example
Requirements for a pet supplies retailer's app
This example is invented to show an app built around repeat buying. A pet supplies retailer sees that a large share of its orders are the same food, bought every few weeks.
- ReasonTwo thirds of orders come from returning customers, most of them reordering. The app's purpose is stated in one line: make the regular order take ten seconds.
- SharedCatalogue, prices, stock, basket and accounts come from the existing store through its interfaces. Nothing is duplicated. Loyalty points are read from the loyalty system.
- App featuresA buy again screen as the home page, reminders when a regular item is probably running low, barcode scan of an empty bag to reorder, and the loyalty card for use in the shops.
- PaymentsSaved cards, Apple Pay and Google Pay. Everything sold is physical. Account deletion is a button in settings that starts the process.
- UpkeepHome screen banners and featured products are managed in the store's content tools, so marketing can change them without a release. A release is planned each quarter, and after each new operating system version.
The first version leaves out browsing features the website already does well. Customers who install it reorder noticeably faster, which is the only measure the retailer set.
Content, measurement and upkeep
- 01
Let marketing change content without a release
Banners, featured collections, promotions and help text should come from the store or a content system. Store review takes time, so nothing urgent should need a new version.
- 02
Measure the app as its own channel
Installs, sign-ins, orders, repeat rate and the journey to purchase, with sales attributed correctly between app and website.
- 03
Plan how customers will hear about it
Prompts on the mobile website for signed-in customers, order emails, packaging and the shops. An app nobody installs returns nothing.
- 04
Budget for releases
Operating systems change yearly and stores expect apps to keep up. Plan regular updates and someone to read reviews and crash reports.
- 05
Require an update path
The app must be able to tell users that a new version is needed, for the day an old one stops working with the store.
- 06
Decide on support
How a customer asks for help in the app, and how staff can see what that customer sees.
The requirements checklist
Ecommerce app requirements checklist
0 of 11 checkedWhat to do next
Find out what share of last year's orders came from customers who had bought before, and how often they returned. If the figure is high, write the single task the app should make effortless. Our mobile app scope planner will help turn that into a first release.
Our mobile app development page explains how we build apps on top of existing stores. Tell us what platform you sell on and what your regular customers buy.
Common questions
Do we need an app if our website works well on phones?
Not necessarily. A fast mobile website serves new and occasional customers better than any app. An app adds value for customers who return often, through staying signed in, notifications, quick reordering and features that use the phone itself. If repeat purchasing is a small part of your business, improving the website is usually the better investment.
Can the app use the same backend as our online store?
It should. The app is best built as another front end to the existing store, reading catalogue, prices and stock and placing orders through the platform's interfaces. That keeps one source of truth. What is possible depends on the platform and plan you use, so confirm the available interfaces before design begins.
Do app stores take a share of sales made in the app?
For physical goods and services, payment is taken through ordinary methods such as cards and digital wallets, outside the stores' billing systems. Digital content consumed in the app is treated differently and generally must use the stores' billing, which carries a fee. Policies change, so check the current rules for exactly what you sell.
Should we build for iPhone and Android at once?
If your customers are spread across both, launching on one platform leaves a large group unserved, and cross-platform frameworks make a joint release practical. Check your website analytics for the split. Where one platform dominates strongly, starting there to learn and following with the other is reasonable.
How do we get customers to install the app?
Offer it at moments when it is clearly useful: after a second order, in order confirmation emails, on packaging and at the till. Give a concrete reason, such as faster reordering or delivery updates. A banner shown to every first-time visitor mostly annoys. Measure installs by source, and whether those customers then order through the app.


