Before building Markfolder, the Twitter bookmaking app, I’d like to validate the problem that I'm trying to solve.
There are tonnes of ways of doing this, but it essentially comes down to either a) talking to people, or b) find evidence of the problem online.
With Markfolder, I decided to do the latter by searching on Twitter for people's comments about the Twitter bookmarks feature. Very quickly I found lots of complaints.
I saved links to about a dozen of those complaints so that I can go back to them later. This could be useful for getting back to those users to introduce Markfolder.
I found this research is also really useful for discovering the language people use to describe their problem. In my initial idea, I used the word "bookmark categories" or "tags" in my own notes, but this research showed that almost everyone used the word "folder" instead.
Next I did a quick survey of popular Twitter apps to see how they implemented bookmarking. Interestingly none of them have bookmarking. I found out that this is mainly due to Twitter's API limitation - bookmarking isn't supported by the API.
There was one app called twlikes that did something similar to my idea, but it hasn't been updated in a year, and I found the UI very difficult to use, so there's still room for building something useful.
So while the above is no guarantee that the product is going to be successful, at least the info gathered gives me confidence that there is a potential problem that people want solving, and it's not being solved by current popular apps.