Very interesting question. I considered all these too before starting the project. My own experience has been one of jumping straight into it with full implementation of all screens across Adalo. I thought that felt right for me because I had the user story in my head already based on my own ‘need based experience’. Additionally, I was thinking only of building an MVP only for one market initially, so jumping straight in worked. With this approach I got to improvise a lot when building the app. I also didn’t do any market research, so it allowed me the flexibility to modify the app as I saw fit.
I think a more detailed approach of a paper sketch, building a proper user stories, building a UML diagram et al would work if you want to build a near final version of your app as opposed to an MVP first. Presumably, there’s a solid market research report in place that tells you consumer needs and behaviours very clearly. Anyhow, had I used this detailed approach, I felt I would have to be rigid with the app build and left myself little room for manoeuvre.
So on one hand I think for building an MVP and allowing yourself flexbility, jumping straight in is better. On the other hand, if there’s a full market research study outcome and revenue plan in place, one is probably better off fleshing out user stories first, have a full UI/UX and then get on with building the app.
I also think Adalo is a great tool for jumping straight in. If you were to go through a detailed approach and you have a good few pennies to spend, then you might as well hire a professional app coding team too.