Researching the Researchers: User Research for Expert Scientific Tools

When I do user research for scientific and research software, the first thing I tell PMs is that we are studying experts, not novices. The people using these tools have decades of domain knowledge and very little patience for interfaces that treat them like beginners. That changes how I research. I am not measuring whether someone can find a button; I am mapping how a scientist actually reasons through a dataset, which filters they reach for first, and where the tool forces them to break flow and drop into a spreadsheet or a script. For a PM, this is where the roadmap signal lives. The pain that matters is rarely cosmetic. It is the query they cannot express, the export they run every single morning, the visualization they rebuild by hand. I lean heavily on contextual inquiry and watching real work, because senior researchers will under-report friction they have already normalized. My job is to surface the workarounds they have stopped noticing, translate them into prioritized problems, and give product a defensible reason to say no to the loud feature request and yes to the quiet, structural one.