User research in three days
Ask about yesterday, not tomorrow
People are bad at predicting what they'll do and good at describing what they did. Instead of “would you use a tool like this?” ask “tell me about the last time you ran into this. What did you do?” The first answer is politeness; the second is data.
Ask to see. A person who shows you the spreadsheet they maintain reveals in two minutes what they wouldn't tell you in an hour.
A conversation structure that works
Five minutes of background, twenty on the last case in detail, five on what was most annoying. Don't present your idea until the end — the moment you do, the person switches to being polite.
Five to seven people in the same role are enough for the patterns to repeat. If every conversation surprises you completely, you're talking to groups that are too different.
What to do with it
After each conversation write three lines: one quote, one pain, and what it changes in the plan. By the end of the week you have a page that settles arguments better than any presentation.
Going deeper
Improving an existing product also starts with watching. Sit a user in front of the screen, give a real task, and stay silent. Five minutes of silence while someone hunts for a button is worth more than any internal design debate.