Web & mobile · Mobile apps
Do you need a mobile app or a better mobile site?
An app feels like the serious option and a mobile site like the compromise. Often it is the other way around: an app earns its place only when people come back often, want to be notified, need it offline or use the phone itself — and it keeps costing money every month after launch.

Start with the site you already have
New customers almost never meet you in an app. They come from search, ads or a link a friend sent, and they land on your mobile site. Nobody installs an app to buy a sofa once or to book a single roof inspection.
The people who would install your app are your most loyal customers — the same people who already put up with your mobile site. If that site is slow, the search is weak and checkout makes them type a card number every time, an app won’t fix the first impression for everyone else. It will move a share of your best customers to a second product you now have to maintain.
So look at the mobile site first. Pull from analytics the share of traffic and revenue on phones, the mobile conversion rate against desktop, and how often a typical customer comes back. Then do the five core tasks on a phone yourself: find a product, filter, add to cart, pay, check an order. Most of what people expect an app to fix — speed, staying logged in, fast reorder, saved payment — can be fixed on the web.
Four reasons an app pays off
1. Repeat use. People need you weekly or more: reordering groceries or pet food, a coffee loyalty card, booking a gym class, a daily tool for field technicians. If a typical customer visits twice a year, the app gets deleted the next time they clear space on the phone.
2. Notifications people want. Order status, a delivery at the door, a class spot opening up, a price drop on a watched item. Push works only if customers opt in and keep it on, so it has to earn that with messages they would be annoyed to miss. If the list of planned notifications is mostly promotions, you need an email list, not an app.
3. Offline or a poor connection. A warehouse scanner, an inspection form with photos filled out in a basement, a ticket that has to open at a stadium gate with no signal. The web can cache some of this, but an app handles it more reliably.
4. Device features. Bluetooth for a connected device, background location, NFC, home-screen widgets, fine camera control, health data. Browsers have caught up on some of these — the camera and basic location work on the web — so check each feature before you assume it needs an app.
A fifth case is simple: an internal app for your own staff, where installing it is part of the job.
What an app costs after launch
The build is the visible cost. Keeping the app alive is the one that surprises people.
- Two platforms. iOS and Android, either as two native apps or as one cross-platform codebase in React Native or Flutter. Either way, two store listings and two review processes.
- Yearly OS updates. Apple and Google ship a major OS version every year, new screen sizes keep appearing, and the stores raise their minimum technical requirements from time to time. An app nobody touches for a year starts to break.
- Slower releases. Every update goes through store review, and some users never update. Your backend has to support several app versions at once, and you need a way to force an update when an old version breaks.
- More moving parts. Push infrastructure, crash reporting, app analytics, an API that the site and the app share.
- Its own marketing. Installs don’t happen by themselves: a store listing, screenshots, ratings, prompts on the site.
- An owner. Someone has to run the roadmap, read the store reviews and answer them.
Before approving a build, ask whoever quotes it what one year of maintenance costs. If nobody can answer, the project isn’t ready.
The options in between
- A progressive web app. A website that can be installed to the home screen, cache content for offline use and send web push. On iPhone, web push works only after the user adds the site to the home screen, which few people do without a prompt. One codebase, no store review.
- A better account area. Staying logged in, order history with a one-tap reorder, wallet payments. This covers much of what loyal customers want from an app.
- SMS and email for updates. For order status, they reach every customer, not only the ones who installed something.
- A small app for one job. Instead of copying the whole store, build a reorder-and-loyalty app for the segment that buys every week.
A quick decision test
Answer five questions honestly:
- How often does a typical customer need you?
- Which notifications would they be glad to get?
- Does any task have to work without a signal?
- Which phone features do you need that a browser can’t provide?
- Who will own the app for the next three years, and with what budget?
If the answers are “a few times a year,” “mostly promotions,” “no,” “none” and “not sure,” put the money into the mobile site.
What to do this week
- Pull the mobile share of traffic and revenue, and how often a typical customer returns.
- Do the five core tasks on your mobile site on a phone and note what is slow or clumsy.
- Write down the notifications you would send, and cross out the ones that are really promotions.
- List the device features you need and check which of them work in a browser.
- Get a one-year maintenance estimate before you get a build estimate.
Boson Group builds mobile sites, web apps and native mobile apps, so we have no reason to push you toward either one. If you are weighing the two, write to team@boson.group and tell us what your customers do on their phones.




