Web app or mobile app? Start with the situation, not the icon
By Pixit
“We need an app” is a reasonable starting point. “It must be installed” is a decision that still needs a reason. How often will people use it, what will they do and where will they be? An icon on the home screen is satisfying. It does not create a reason to open the product.
A link can be the best entry point
A browser-based application may suit a client portal, an internal tool or a service used occasionally. Someone follows a link from a message and reaches the product without first completing an installation.
Consider a customer reviewing a proposal once a month. Adding a download and onboarding step needs a benefit that outweighs that effort. Test the journey on a phone from the start. A web app should not simply be a desktop layout squeezed into a smaller rectangle.
Mobile earns its place through specific needs
An installed mobile app may be appropriate for frequent use, deeper device interaction or an experience closely tied to the operating system. List the actual requirements: taking photos, location, notifications or working without a connection.
A field worker recording an intervention and synchronising later has different needs from someone making an occasional enquiry. Do not add device features simply to make the specification feel complete. Describe the situations that make them necessary, including what should happen when a permission is refused.
Where does a PWA fit?
A progressive web app can provide some app-like capabilities. Support depends on the feature, browser and device, so the critical behaviour must be checked on the environments you intend to support.
Choose a short list of real devices and browsers used by your audience. Prototype the decisive function before committing to the whole build. A successful demonstration on the developer's own phone is useful evidence, but it is not the same as validating the target audience's conditions.
Include operation and distribution in the budget
Interfaces are only part of the work. You also need data handling, server behaviour, permissions, testing and maintenance. A mobile distribution plan adds publication preparation and continued attention to the chosen platforms.
Clarify who owns the publishing accounts and how corrections reach users. Ask what happens when the network disappears halfway through an action. Sending the same booking twice is not merely a visual inconvenience; the product needs a deliberate answer.
Can the project begin on the web?
Yes, when that is a sound way to test the service. Clear data structures and business rules can help future development, but they do not make a later mobile interface free or automatic.
At Pixit, we prefer deciding with three usage scenarios, including one that goes wrong: no connection, a denied permission or an older device. If the proposed approach still supports the essential task, the choice has substance. Your product deserves more than a format selected out of habit.
