On this page
A leading question hints at the answer you're hoping for, so participants go along with you instead of describing what actually happened. The fix is straightforward: ask only about things that happened, don't assume conclusions, avoid loaded words, and ask one thing at a time. This article covers the seven most common types of leading questions in user interviews, shows before-and-after rewrites, and includes a checklist you can use when writing your interview guide and after interviews.
Why leading questions do so much damage
Participants usually want to be helpful and don't want to disappoint the researcher. Once a question reveals what you hope to hear, many people go along with it or give an answer that sounds reasonable. The researcher gets an echo of their own hypothesis and mistakes it for validation.
Worse, leading is often hard to spot. When researchers write questions, they already have a view in mind, so their own questions read as neutral to them. When pulling quotes after the interview, it's also hard to see that an answer was led by the question. That's why you need checks before and reviews after, rather than relying on instinct in the moment. In The Mom Test (2013), Rob Fitzpatrick sets out the basic principle: ask about specific things people have done in the past, not what they think of your idea.
Seven common types of leading questions
1. Assuming the answer
The question already contains a premise, and participants can only answer within it. For example, "What difficulties did you run into at checkout?" assumes checkout was difficult.
2. Double-barreled questions
Ask about two things in one question, and participants often answer only one, or give a vague "it's fine", so you can't tell which one they're judging.
3. Loaded words
The question contains slanted words like "convenient", "annoying", "expensive", or "innovative", and participants will follow that word's lead.
4. Hypothetical future questions
Questions like "Would you use it?" or "Would you buy it?" almost always get overly optimistic answers. People are bad at predicting their own future behavior, and there's a wide gap between saying they'll use something and actually using it.
5. Asking whether they like it
"Do you like this design?" is a closed question that also carries social pressure, so most participants will say yes. Even if they say they like it, you don't know what they like or whether that will turn into action.
6. Offering the options for them
"Did you stop because of the price or the features?" limits the answer to the options you thought of. If the real reason isn't among them, you miss it.
7. Pressure from others
"A lot of users say this feature is really useful. What do you think?" puts the majority view up front, and participants feel awkward disagreeing.
Before and after rewrites
The general approach to rewriting: replace opinion questions with experience questions, replace closed questions with open ones, split one question into two, and remove loaded words. The product scenarios in the table below are examples.
| Type | Before | After |
|---|---|---|
| Assuming the answer | What difficulties did you run into at checkout? | Walk me through the last time you checked out. |
| Double-barreled question | Do you think this feature is easy to use and useful? | What were you doing the last time you used this feature? (Then ask separately how it went.) |
| Loaded words | The new home page is a lot more convenient, isn't it? | After the redesign, what did you do the first time you opened the home page? |
| Hypothetical future questions | If we built this feature, would you use it? | When you need something like this, how do you handle it today? |
| Asking whether they like it | Do you like this design? | When you look at this page, what's the first thing you notice? What would you do next? |
| Offering the options for them | Did you stop using it because of the price or the features? | When was the last time you used it? What happened after that? |
| Pressure from others | A lot of users say this feature is really useful. What do you think? | Have you used this feature? What was it like the time you used it? |
| Asking about willingness to pay and price | This price works for you, right? | When was the last time you paid for a similar service? How much did you spend? |
If a rewritten question feels a little flat, it's probably right. A good question takes no position; the information comes out in follow-up.
What to do when you ask a leading question mid-interview
Leading phrasing will slip out in the moment. When it does, don't pretend it didn't happen. Follow up right away with a neutral question to bring things back, flag that part of the answer in your notes, and give it less weight in analysis.
The first answer, "pretty convenient", was led by the question. Only after asking about specific experience did it become clear the participant had never used the feature. If you'd only written down the first answer, your finding would be the opposite of the facts.
How to check systematically across a study
- Once your interview guide is written, go through it against the checklist below, mark any problem questions, and rewrite them.
- Ask a colleague who doesn't know the project's hypotheses to read the guide and see whether they can guess the answers you want.
- Run a pilot interview and record it, then listen back specifically for leading lines you said on the spot.
- When working through interview notes, mark answers that came right after a leading question, and don't use them as the main basis for a finding.
- Before drawing a conclusion, check for counterexamples. A finding with only supporting evidence and no counterexamples is a reason to go back and look at the questions themselves.
- Does this question ask about a specific thing that happened, or about an opinion or prediction?
- Does the question assume something already exists (for example, a difficulty or a need)?
- Does each sentence ask only one thing?
- Does it use slanted words like "convenient", "annoying", "expensive", "easy to use", or "innovative"?
- Can it be answered with just "yes" or "no"? If so, make it open-ended.
- Does it list answer options for the participant?
- Does it mention the views of other users, experts, or your team?
- After reading this question, could the participant guess what I'm hoping to hear?
- Did you finish asking how they do things today before getting to the product or concept?
- For willingness to pay, did you switch to asking about real past purchases?
Common mistakes
- Checking only when writing the guide, not during interviews: leading often happens in follow-up questions asked on the spot. Listen back to recordings specifically for these lines.
- Treating "looks good" after a concept demo as validation: it's mostly politeness. Look for real actions, such as asking about price or launch date, or offering contact details.
- Making questions vague and hard to follow to avoid leading: participants won't know how to answer. Keep questions specific; just remove the bias.
- Deleting the record once you realize you asked a leading question: the answer itself shows the participant is easily swayed. Keep it and flag it.
How to do it in heari
When heari generates an interview guide from your business question, it avoids assumed answers, double-barreled questions, and loaded words, and the research lead can edit each question while reviewing the guide. The AI-moderated interviews follow up on participants' answers, prioritizing specific experiences. In the generated insight report, each finding comes with participant quotes, counterexamples, and how many participants mentioned it, and the research lead can accept, edit, or reject each insight.
FAQ
What is a leading question?
A leading question is one whose wording, premise, or options hint at the answer the interviewer expects, nudging participants to go along with it rather than share their real thoughts and experiences. Common forms include assuming the answer, loaded words, double-barreled questions, and asking whether they like something.
Is "What do you think of this product?" a leading question?
It isn't strongly leading, but it asks for an opinion, so answers tend to be vague and polite. A better approach is to ask about specific experience, such as "What were you doing the last time you used it, and how did it go?", then follow up based on the answer.
Can you ask "Would you buy it?" in a user interview?
We don't recommend asking it directly. People's predictions about their future purchases tend to be optimistic. Ask instead about times they paid for similar products or services. If you need to quantify price acceptance, use a dedicated method such as the Van Westendorp Price Sensitivity Meter or the Gabor-Granger pricing method.
How can you tell whether an interview finding was led by the question?
Go back to the raw notes and see which question came right before the quote that supports the finding. If the question already contained the wording of the finding, or offered only yes or no, that quote should carry less weight. Ideally, confirm it with other participants' answers to open-ended questions.
