On Monday, I gave a short attendee talk at MicroConf Growth on how to conduct user interviews and usability sessions. It was only twelve minutes, though, so I had to focus it on the key point:

Getting useful user feedback is all about how you act. To find your user’s pain points (and in turn opportunities for your business), you need to make them feel comfortable being open and vulnerable with you. This means you cannot negate them in any way, otherwise they’ll shut down or try to impress you. It is not a conversation — you are there to absorb whatever they tell you.

Or, as fellow speaker Hanne Vervaeck so succinctly put it, “Shut up and listen.”

In talking to people after, I realized there were a lot of little things you can do to make your interviews go smoothly and help you get useful feedback. This blog post attempts to pull them all together. (This post is a bit long, so you may want to bookmark it and come back to it later when you’ve knocked out the 398 other things on your to-do list and are finally ready to hit the ground with interviews.)

Interviews as marketing

• Interviews can also be a powerful fuel for word-of-mouth marketing. Especially for a small B2B company with the founder doing interviews, people knowing that a company has founders who will listen is MASSIVE for word of mouth. We are all so used to being ignored by companies that it’s a rare novelty when they do, and it makes you feel special. We’ve had people on our free tier become significant promoters and lead to lots of business because they tell people they know that we’re a company that actually cares about their users, offer testimonials, and otherwise promote us like crazy without us asking.
• Interviews should never be used to save, sell, or cross-sell. Ever! Sorry. It will ruin the trust you’ve built up and they will tell people how you deceived them. This is especially true for cancelation interviews. Resist the urge.

How many and how often

• You don’t have to do a ton of interviews. I personally tend to do 1–2 interviews a week to build general understanding of our customers, but you don’t have to do that many. 1–2 a month is a good goal to start with. Do them when you have a specific question in mind, like why a landing page isn’t working or why customers in the {x] industry are such sticky customers for you.
• For any specific research question, you really only need to talk to 5 people. The Nielsen rule of thumb for usability is that you’ll get 80% of the feedback you need from 5 people. The informal rule of thumb is “stop interviewing when you start hearing the same things over and over again.” The most I’ve ever interviewed for a discrete question is 12, and that’s usually a sign the question was too broad. After 5 people, you can do some tweaks and tests, and then revisit in a few months if things still aren’t performing as hoped. Or never. Up to you.

Who and when to interview

• If you interview people too soon after they first sign up, they may confuse it as an onboarding call. I would suggest waiting at least a month after their first purchase before doing an interview. Onboarding calls are still worth doing, they’re just different and generally less process/pain-point focused.
• Interview happy people so you can find more happy people. This means reaching out to people who’ve used your product for a long time. This goes against some conventional advice that you should focus on new people and people who canceled. Focusing on people who cancel is a symptom of the natural human tendency towards loss aversion (see the research of Daniel Kahneman/Amos Tversky). If you look at your data and notice that customers in {industry} have a particularly high LTV/low churn, reach out to them so you can understand them better and find more people in that industry.
• Interview a cross section of customers, not just the highest paying ones. You should be in regular contact with your highest LTV customers in general to keep them happy, but some incredible insights can come from people who are only paying you a small amount per year.
• You don’t have to interview each person for a specific customer. Let’s say 5 people from a company use your product. While it would be helpful to hear why the buyer eventually bought it vs why the team member put it in front of them in the first place, don’t feel like you need to interview everyone.
• I like to automatically trigger interview emails after someone has made a purchase. It means the product is relatively fresh on their mind and seeing the charge go through didn’t cause them to cancel, which is always a good sign.
• Cancelation interviews are Expert Level and hard. People are usually upset. Handle with care. If you’re just starting out, I’d suggest sending plain-text emails to cancels and get your cancel feedback via email instead.
• Cancel interviews and surveys are great clues for usability issues that help you save other people. If people cancel because they wanted a feature they didn’t think you offered, run usability testing on your site/landing pages/order forms/etc.
• Send out plain text emails from the founder. People are more likely to reply to emails from the founder. (Response rate is still pretty low — currently, our new user feedback emails have a 4% reply rate.)
• Subreddits are a great place to recruit non-users. This is my favorite tactic for recruiting non-customers for usability screenshares. Find a subreddit in the niche you need and provide an incentive. I got 70 replies from a post I put in a focused subreddit that only had 5,000 subscribers. For example: “I’m looking for {description of person} to help me test a new website we’re building. $25 gift card to Amazon if you participate!”

Incentives

• For non-users, give them an incentive that’s not tied to your product, like a $25 gift card to Amazon. It’s way cheaper than spending $500 on Facebook ads post-launch on something that doesn’t convert.
• For users, you can give them an incentive that is related to your product. For example, if your product is $50 a month, give them $10 off their next bill. People like this because they notice that new customers often (for any service) get coupons, and this feels like a reward for loyal customers. Happy vibes all around.
• For B2B, you may not need to incentivize existing customers because getting to offer feature suggestions directly to the founder is generally enough of an incentive. For B2C, you definitely do.

The interview itself

• Use a script. If you haven’t interviewed before, you will probably be nervous, and that’s okay! I had the good fortune to learn under and observe two incredibly talented and seasoned researchers, Betsy Bland and Dr. Helen Fake, PhD. Something they taught me was to use a script. It helps with the nerves and makes sure you stay on track. For the first 100 or so interviews I did, that’s exactly what I did. I like this one a lot, and would modify it for the product. Copy it into a document, put about 5 spaces in between each question, and print it out. Then, have it in front of you during the interview and write down notes under each question. It’ll help with the anxiety of “what do I say?” and will keep your notes organized.
• Ask the reaching-for-the-door question halfway through. I schedule this into my interviews, and it leads to incredible insights, and this is why it’s so important to act in a way that lets your customer become open with you. Doctors talk about how the patient will bring up the actual reason they came in just before the end of the appointment as they’re reaching for the door. Half way through — so if I’ve said it’s a 30 minute call, 15 minutes in — I’ll say “Thank you so much, I’ve learned so much from you today. That about covers the questions I had, is there anything else you want me to know?” Pause, pause, pause…and they will fill the pause and the floodgates will open. I’ve had interviews completely turn around at this point.
• Do interviews over the phone because people will be more open. They will be less open about their fears, pains, and struggles if they have to see your face. I suggest sending out the invite as a phone call (get their number) so there is no room for them to become anxious about you seeing them.
• Do usability sessions over screenshare. I like using Google Hangouts. It’s free and they don’t need to download software, which can be a pain. Since you’re a small organization you don’t need to record them.
• If they ask you a question about your own experience, always refocus on the user. For example, if you are the founder doing the interview, and they ask you “have you heard of X tool,” your answer is always “I have, and I’d love to hear how you use it” even if you haven’t or even if you previously worked for that company for 10 years and know everything about it.

Post-interview

• Send people a thank-you, whether it’s via email or (bonus points) handwritten note.
• Tag any feature requests/problems they bring up, and follow up with them if/when you add those features/fix those problems/address a pain point they brought up. I usually create a GitHub issue out of Intercom, but there are a ton of other ways you could do this (spreadsheets, Userfeed, etc).

Sharing and analyzing the feedback

• If you’re a small company, you don’t have to follow “best practices” invented by big companies with 50 person research teams. You don’t have to do post-its and transcripts and card sorting and journey mapping. Those things are great but you’ve got a million other things to do.
• Take notes in a central location. I like to take notes in Intercom attached to the user and tag my teammate. If you have less than 5 people in your company, just email them out after the interview with some key takeaways. If you’re taking notes on paper as suggested above, scan them in.
• If you’re at a large company, use video to share the customer’s perspective. In my old life as a corporate product manager, I tried a zillion things to get people to hear the voice of the customer. Slack posts, infographics, presentations, hand-outs, it goes on. The most effective way is to have people sit in the room with you as a silent observer, but that isn’t very scalable. I’ve found that the most effective way to get people to empathize with the user is to share a video of them saying it. If I had to rank strategies from least to most effective, I’d say text => graphical => audio => video => in the room.