Does it really matter? I would say finding a problem to solve is the first step. From there you can assess if your target users are more prone to use a desktop

Does it really matter? I would say finding a problem to solve is the first step. From there you can assess if your target users are more prone to use a desktop website, a web app, or a native mobile app. I guess that some industries will only rely on installed native apps and frown upon browser web apps, and vice versa. Also, if your solution save people time, they would probably take the extra step of downloading something if they have to. I work on a rather popular dev tool, and it's a desktop app. Nothing else would really make sense for what it does (maybe a web app, but it would be more limited in capabilities). But people are willing to download it because it's useful. That being said, I think mobile apps come with different challenges regarding distribution, conversion, etc. Also, the learning curve is definitely steeper in my opinion. Learn the web techs (JS, HTML, CSS) and you will be able to build nearly anything you want, front-end (websites, web apps, hybrid mobile apps) and back-end.