Starting with the wrong question
Most first-time interview studies fail before anyone picks up the phone. The team writes a discussion guide full of questions like "would you use a feature that does X?" That question asks people to predict their own future behavior, which people are bad at doing.
A better starting point is a question about the past. Ask what someone actually did the last time they hit a specific problem. Past behavior is something a person can describe with detail. Future intent is a guess dressed up as an answer.
Before you write a single interview question, write down the decision your team needs to make. If the study can't change that decision, the questions probably aren't worth asking yet.
Treating the interview like a survey with a person attached
A common mistake is scripting every question in advance and reading them in order, regardless of what the participant says. This turns a conversation into a form. You lose the follow-up questions that usually contain the most useful information.
A discussion guide should be a list of topics, not a script. Write the questions you must ask, but leave room to follow a thread when someone says something unexpected. If a participant mentions a workaround they built themselves, that workaround is often more valuable than anything on your list.
Recruiting people who are too easy to reach
Teams often recruit from whoever is closest: a mailing list, a customer support queue, a coworker's team. This produces a sample that looks like the people who already agree with the product, not the people the product needs to convince.
Instead, define your recruiting criteria before you start looking for participants. Decide on the behavior or role that matters, not just demographics. Someone who tried your product once and quit is often more informative than someone who uses it every day, because they can tell you exactly where it broke down.
Asking leading questions without noticing
"Don't you find it frustrating when reports take too long to load?" is a leading question. It hands the participant an opinion and asks them to agree with it. Most people will, out of politeness.
Rewrite leading questions as open ones. "Walk me through the last time you pulled a report" gets you a story instead of an agreement. If you catch yourself using words like "don't you think" or "wouldn't it be better," stop and rephrase.
Filling silence too fast
New interviewers tend to be uncomfortable with pauses. The moment a participant stops talking, they jump in with another question. But the few seconds after someone finishes an answer are often when they add the most honest detail, almost as an afterthought.
Practice counting to three in your head before you speak again. It feels long in the moment. It rarely feels long to the person you're talking to.
Not separating what people say from what people do
People are generally honest, but they're also generally bad narrators of their own habits. Someone might say they check a tool "a few times a day" and mean it sincerely while actually checking it thirty times. This isn't a mistake, it's just an area to double-check.
Where you can, ask for a specific recent example instead of a general habit. "What did you check this morning, and when?" gets you closer to the truth than "how often do you check this?"
Eric Morrison, a User Experience Research Lead from New York who has led research programs at companies including Google, TikTok, and Disney, has spoken about the difference between what users report and what they actually do as one of the more persistent gaps in this kind of research. Pairing interviews with a second method, like watching someone use a tool in real time, helps close that gap.
Writing up findings as a transcript dump
The last common mistake happens after the interviews are done. A team collects a stack of notes and quotes, then presents all of them without a clear point. The result is a document nobody acts on.
A useful write-up leads with what changed in your understanding, then supports it with two or three specific examples. Cut anything that doesn't connect back to the decision you started with. If a quote is interesting but doesn't move the decision, it belongs in an appendix, not the summary.
What to do differently next time
Run a small pilot with two or three interviews before committing to a full study. Read back your discussion guide out loud and ask whether each question could be answered with a guess instead of a memory. If it can, rewrite it. Small corrections made early save a lot of wasted interviews later.
