A very effective outcome for user research is to have N fact based observations of users doing one activity that involve going from A to B to C. And just by reading these fact based observations, one can figure out a way to implement a change that eliminates B. Such a change is immediately and unequivocally of value (High Gain) as demonstrated by user research data.
In this particular case, there is no need for explaining or interpreting. The implementation of the release notes assistant is one example that demonstrate how that kind of user research is relevant and helpful.
Unfortunately it is not a general case because there are a number of cases where such raw data is not available for various reasons. For instance it does not help if the path chosen by a user is inherently flawed and need to be re-invented, not just optimized.
A very effective outcome for user research is to have N fact based observations of users doing **one** activity that involve going from A to B to C. And just by reading these fact based observations, one can figure out a way to implement a change that **eliminates** B. Such a change is immediately and unequivocally of value (High Gain) as demonstrated by user research data.
In this particular case, there is no need for explaining or interpreting. The [implementation of the release notes assistant](https://codeberg.org/forgejo/discussions/issues/197) is one example that demonstrate how that kind of user research is relevant and helpful.
Unfortunately it is not a general case because there are a number of cases where such raw data is not available for various reasons. For instance it does not help if the path chosen by a user is inherently flawed and need to be re-invented, not just optimized.