User research makes sense with product decisions to be made after the talks are over. Collecting a large number of notes does not automatically tell which screen needs to change. Before starting the research, it is necessary to write the uncertainty in one sentence: Why do people stop when they register, what information cannot they find, or in what order does they do their current work?
Select the participant based on his/her behavior
The first person to try the product on a daily basis does not have the same problems. Separating new and experienced users makes it easier to interpret findings. Reach out to people who represent the actual tasks of the target users rather than recruiting participants from a group of friends who seem appropriate for the interview. Do not present the results of a small study as the view of the entire market.
Distinguish observation from interpretation
"This screen is a mixed" comment; "The participant opened three pages to find the delivery information" is observation. The second allows to investigate what information is needed. Ask the user to describe the last work he did, the documents he used, and the obstacles he encountered, rather than asking him to design solutions.
Draw the flow and trace the decision
The initial, decision points, error states, and completion step should be visible on the user flow. Note which finding supports each change. A behavior repeated with the preference of a single person should not be evaluated at the same weight.
The research delivery should include not just a transcript of interviews; priority issues, suggested trials, and questions that have not yet been answered. The next prototype test tests these decisions so that the research is not a separate report, but a justification for screen design.