Start with the job, not the format
A mobile site is for finding you. A web app is for doing something in the browser, a booking flow, a portal, a checkout. An installed app is for a relationship: repeat visits, notifications, an icon on the home screen. Most small businesses already have the first and have not asked whether they need the third.
The honest question is not 'web or mobile' but 'what do my best customers do repeatedly'. If the answer is 'check the menu, book, or reorder', you have a relationship use case. If it is 'find us and call', you do not need an app yet.
The case for staying web
If the core value is content and findability, money spent on an app is money spent on distribution you already have. Improve the mobile site, fix the load time, make the call and booking actions unmissable, and measure. That is often the entire play.
Web apps also avoid the store layer entirely: no fees, no review, no two-platform builds. For internal tools, staff portals, and single-purpose utilities, a web app is the right answer almost every time.
The case for an installed app
The moment repeat engagement matters, the installed app wins: push notifications that arrive, a home-screen presence that survives Safari's anti-install friction, and a store listing that reads as legitimacy. Loyalty, ordering, and appointment businesses see this every day.
The cost of the app is no longer the reason to say no. A builder that turns your existing site into signed native apps removes the two-platform pipeline, and you keep the site as the single source of truth.