Would it help to let user research be run by someone else than the designer if possible. It seems to me that confirmation bias is very hard to avoid when you designed the solution yourself or when you’ve been involved for a long time. Briefing a researcher to look for falsification might be more effective? What do you think?