Leave a comment
Get the GH Bookmarklet

Ask GH

Hi everyone, I am working on the launch of a new product. And as a personal goal I am working on having one on one conversations with my first 100 users and to collect their feedback to build a awesome product. My question is most of the time when i ask people, they say that i have a great idea. But it could very well be that they are being polite. Can someone guide me how to interview users so that i can land on actionable insights. I know i should focus on retention or other metrics to get a feel of what my users actually think. i am more focused on how to structure a user interview FYI: I am in touch with most of my users through facebook or email. No face to face interviews are possible.

Based on this thread and my own experience. I have put together this knowledge map for other who want to learn about conducting user interviews. It has all the resources you need to study and my own insights. Here is the link http://knowledgemaps.org/map.php?m_id=169

  • PT

    pavneet tiwana

    over 5 years ago #

    Hi guys, I did some research on my own and here is just a brief summary of all the links i found helpful and what i learned from them.

    1, http://leanstack.com/how-to-interview-your-users-and-get-useful-feedback/- Tips on how to conduct a user Interview. Drive the interview, it is not a conversation. Focus on figuring out the why behind user behavior and don't add to much from your side. Also silence is not a bad thing in an interview.

    2.http://www.slideshare.net/evanish/how-to-do-customer-development-interviews-to-validate-your-startup-idea - Slide presentation on costumer development interviews. Also includes a series of questions which can be used.

    3. http://mfishbein.com/the-ultimate-list-of-customer-development-questions/- A list of question for conducting costumer development interviews

    4. http://giffconstable.com/2012/12/12-tips-for-early-customer-development-interviews-revision-3/ - More tips on conducting user interviews. Focus on listening, divide the behavior and product into two different sections. Try to guess the user behavior by asking open ended questions, focusing on behavior and drilling down when ever there is a insight to be found.

    5. https://medium.com/user-research/never-ask-what-they-want-3-better-questions-to-ask-in-user-interviews-aeddd2a2101e- Talk about what is the user trying to get done, how he is doing it currently and how can the process be better

  • JG

    Jim Gray

    over 5 years ago #

    This question comes up a lot, so I wrote out my most common answers.


    Face to face is nice when you can get it. It's often difficult to do so. The important thing is to have an in-depth dialogue with relevant people, as determined by closeness of fit to "people who will actually be paying you".

    Google Hangouts could be a good option, and has simple screen sharing if you need it. Slack is also good. Have a dialogue. And learn, don't sell, don't propose solutions. The benefit to you is to learn, and to get people to open up to you. Stick with that one goal & don't inject other priorities which sacrifice it.

    And what's the simplest form of your plan which you can possibly ship & start charging people money for? Do that, do that now. The idea-action barrier is one which most people fail to overcome. Get it out of the way, start accumulating actual customers, then improve your delivered value over time.

  • AA

    Anuj Adhiya

    over 5 years ago #

    I highly reccomend picking up a copy of @ashmaurya's Running Lean book. It will walk you through the exact process of problem and solution interviews among other things you need to do to raise your odds of success.

    Another resource you can refer to is the Lean Startup Circle Wiki: https://growthhackers.com/the-lean-startup-circle-wiki-customer-interview-templates-and-resources/

  • NA

    Nate Allen

    over 5 years ago #


    You're asking the right question. You are absolutely looking for ACTIONABLE information, which can be difficult from an interview. Here's a couple changes I made to my customer interview experiments that made them much more actionable:

    1. Develop a hypothesis. A hypothesis should be based on a core assumption about your user. Early on, a good hypothesis format is "X has a problem with Y." X being a description of your user, and Y being a description of a problem. Both should be as specific as you can get. That will make your results more actionable.
    2. Write your questions before the interview. Make sure they help you determine whether your hypothesis is correct.
    3. Ask the same questions to AT LEAST 10 people. After 10, you can decide if you need to modify your questions, or if you have proven/disproven your hypothesis.

    We have written a couple blog posts specifically on customer interview questions and structure over at http://blog.devcopilot.com/. Take a look, and I'm happy to chat more about the specifics.

    • PT

      pavneet tiwana

      over 5 years ago #

      Thanks nate, i will take a look. I will be also willing to share with you my experiences of customer interviews

      • NA

        Nate Allen

        over 5 years ago #

        Pavneet, have you put any of your research into action? I'm curious how your interviewing or your question writing is going.

        • PT

          pavneet tiwana

          over 5 years ago #

          Hey Nate, I tried doing this. And did get some feedback. But i realized my target group was wrong. I need some savvy group of users who will be potential users of my product.

  • TL

    Tor L. Bollingmo

    over 5 years ago #

    Read up on the Jobs To Be Done framework, and do interviews around the job and not your own product. See http://jtbd.info/ and order this http://jobstobedonehandbook.com/

    Feel free to shoot me an email tor/at/torbollingmo.com if you have any questions.

  • JD

    Jesus Del Campo Salvá

    over 5 years ago #

    Nice one.

    I had this issue at the very beginning and I highly recommend you to have it done face to face.

    Personally I get a better grip when I talk to customers/users face to face, because then you can read them as a whole, even look at their body language and see if they are indeed being polite or are being honest :)

    At least thru Google Hangouts as @grayj said

    Best regards!

  • DL

    Dylan La Com

    over 5 years ago #

    Great question @pavtiwana!

  • SK

    Srini Kami

    over 5 years ago #

    Great question,

    I have been looking into this myself for my APP from last few days. Here is what I put together from various sites, books and templates and other online resources. I am getting ready for interviews, but haven't done any interviews yet.

    My APP NoAccent helps people, who speak English well, but still have non-native accent, so questions are specific.

    1st rule of validating your idea: Do not talk about your idea.
    Ask for brutal and honest feedback
    2nd rule of validating your idea: Do not ask about the future assumptions. Ask past and present
    Don’t talk much, guide, lead, but listen.
    Ask open ended questions
    Drill down, for important information
    Ask about introductions
    Occasionally repeat what they said to confirm

    1. How long have you been trying to reduce your accent?
    2. Why do you want to reduce your accent?
    3. What are your doing presently to reduce your accent?
    4. What other things or practice you do for accent reduction?
    5. What is hard about current thing you are doing?
    6. Why is it hard?
    7. What, if anything, have you done to solve that problem? this tells you how deep is the problem, if they are not using anything right now, the problem is not big enough.
    8. Tell me about last time you tried that (currently using product or process)?
    9. Where do you go and find information about accent reduction online?
    10. What don’t you love about the solutions you’ve tried?

  • LS

    Logan Stoneman

    over 5 years ago #

    My team has done face-to-face interviews with both our happy & unhappy users. In addition, we also record the conversation through Google Hangouts on Air and then give that link to the rest of our team to watch when they have free time. That way everyone from biz to dev has an idea of what our users are saying.

  • BC

    Becky Cruze

    over 5 years ago #

    Great question, and there are already a lot of insightful responses here. I just wanted to add that I found the "How to Start a Startup" lecture by Twitch cofounder Emmett Shear on this topic very helpful. Here's a summary of his lecture, which also has a link to the full video of his talk: http://thinkapps.com/blog/launch/how-start-startup-running-user-interview/.

  • SJ

    Sebastian Johansson

    about 5 years ago #

    Great advice in the comments already. Some stuff I have done in the past as well:

    1. Thrown out a survey to 200+ people on mechanical turk. Described the product idea and asked them if they would buy it. If not, why?

    They might not be in your target group but a lot of times they can still provide good, cheap and quick feedback. Based on the feedback I polish up the product idea description.

    2. Set up a landing page with a fake order button and spend some money on ppc. This rules out all the nice, emotional people that will tell you that your idea is great, even if they dont think so. If someone really likes it they will click the buy button, or at least send you an email asking for more details.

    Then I try to do more standard qualitative cust. dev. interviews too.

    But Ive seen more upsides testing business ideas in a high tempo pace, rather than going really deep with every idea. If an idea had some bearing, it always showed some traction straight away. At least for me.

    To go through the whole cust dev process in a proper way before you can decide if you should fold, might take you a month. If one on average needs to test maybe 40 ideas before you find a winner, thats 40 bloody months.

    Think a lot of entrepreneurs would benefit from a less formal, high tempo testing, pre- cust dev process. Then only taking an idea through the formal cust dev process, if it showed strong traction with actual orders. At least it has worked for me.