Generally, good product management processes should be able to identify through customer valiation early on, whether a feature is a need and should therefore be part of a product, or not. But, this can be sort of a dark art, and in most cases, difficult to get 100% right, depending on whether you have asked the right questions, whether you have asked the right number of people, and in fact, the right people, that represents your app constituency.
But in an agile world, it is not enough to prove your hypotheses at the product-feature discovery stage, but for product managers continue to validate their product features constantly. Understanding the pulse or health of your feature post-launch is crucial to understanding how it is being used, how often it is being used, and subsequently how important it is to users.
“It's really hard to design products by focus groups. A lot of times, people don't know what they want until you show it to them.” (Steve Jobs)
This is where metrics comes in, and as the title of this article hinted, metrics is master, being able to get useful and measurable metrics on how your users are using your features help re-calibrate the prioritization of the feature over time, as well as the success of the features.
I wrote an article a short time ago, titled How to do a Retention Analysis, which is a powerful tool in helping uncover insights into not the acquisition aspect, but the retentative aspects over time. It produces a relevancy inisght over time, to prove that a certain feature is just as interesting now as it was when it was first launched.
As a bare minimum, you need to track user engagement, time spent, conversions. You need to know what metrics matter, rather than use any statistic to give you an excuse to hold on to a feature. Have some metrics that are considered big picture measurements, that align with the bigger picture goal of your organization, but also have measures that track the feature vital signs, usage and usability metrics to find feature, as well as UX pain points.
It is also important that you continue to collect customer feedback, and iterate and fix up your feature, rinse and repeat. A product manager's role is not just to identify the problem but also identify how to build the right solution to the problem. I like to use A/B Testing or Multivariate Testing, which I wrote about a while back, as ways of gauging user feedback, not just through questions, but providing different UI to different users, to see what their reactions are.