A well known saying in software development is that users do not know what they want. While this may sound harsh, there are a few reasons why direct user feedback from discussions, interviews, and focus groups does not necessarily lead to good product feedback:

• One reason for this is that everyone wants to be liked. When conducting an interview, or talking to a group of users, there is a chance that they will only say what they believe the interviewer wants to hear.
• It has a high turn-around time. Scheduling interviews and focus groups takes time and finding a time that everyone can attend can easily take days or even weeks.
• It is hard to keep asking the same group of users for feedback every few weeks. This is especially important when trying to determine whether the quality of a feature is improving with the newest updates or not.

For these reasons, it can be worthwhile to cut back on asking for feedback, but instead, measure how users are interacting with an application on a functional level and whether they are satisfied with the value they receive from an application.

One way to do this is by measuring user behavior in an application and emitting metrics based on that. In Chapter 10, Application Monitoring, Application Insights was introduced for gathering application-level metrics. While metrics are traditionally used for emitting metrics regarding application performance, metrics can also be used to emit metrics regarding application usage. Some examples are the following:

• How frequently is every page visited?
• How many times are specific operations performed?
• How long it takes to complete a certain view?
• How many users open a specific form, only to never complete it?

Gathering these metrics can deliver important insights into how users are interacting with an application and which parts they use or do not use.

Besides usage, another indicator of user satisfaction can be Twitter sentiment or the number of support requests.