How do you prioritize what feature you should build?

We turn our list of features into problem statements first. If the feature is "Add data export", that becomes "It's frustrating that I can't export my data." Once we have a list of the problems, we send them out to a bunch of our active users to vote on using a special voting method called pair-choice comparison to surface the problems that are most important to our core user base. Then we outline a number of product team sprints for the following 3 months to tackle those most important problems, before repeating the process again :) We're a 4 person startup, so this helps us from straying away from the right thing to work on. Overall the approach is called Customer Problem Stack Ranking: https://www.opinionx.co/blog/cus...