Book Summary — Talking to Humans by Frank Rimalovski and Giff Constable

Talking to Humans teaches you how to get out of the building and run effective customer discovery interviews.

Book summary

Talking to Humans is a practical how-to guide to running effective user interviews. It teaches you how to reach out to the right interviewees, prepare for and run the interviews, and make sense of the results.

How to prepare for an interview

• Pick the right user to interview. Every opinion is not of the same importance to you. For example, in a B2B environment, the different buyer types (economic, technical, strategic, user) have very different needs. Who do you want to learn from?
• Invite a colleague with you. You will have two points of view to the discussion, and you can split the work — one focuses on taking notes and observing the non-verbal cues while the other focuses on the conversation.
• Decide what you want to learn from the interview. You want to validate your most critical and uncertain assumptions first. Which of your assumptions might end-up killing your product? If the critical assumptions are still unvalidated, don’t focus on the details yet.
• Prepare a list of questions. Select questions that will help you validate your assumptions. Chances are you won’t be asking these exact questions, but having a guide helps you get the most out of the interview.

What kind of questions to ask during an interview

• Start with warm-up questions. The user is often nervous, so break the ice and get them to talk by starting with some easy questions.
• Ask open questions. Open questions start with words like who, what, why, and how. They typically can’t be answered with a yes or a no. Closed questions start with words like is, are, would, and do you.
• Ask about actual situations, don’t ask the user to speculate. “How would you use this product” is a bad question. “Walk me through the last time you solved this problem” is a better one.

How to run an effective user interview

• Do interviews in person. To understand what the user is really saying, you want to read the body language and other non-verbal cues. If only online interviews are possible, have the camera on.
• Talk to one person at a time. People might have a hard time expressing themselves in groups, so it’s best to talk to one person at a time.
• Be aware of confirmation bias. You probably believe in your idea — otherwise, you wouldn’t be running the interviews. People have a subconscious bias towards evidence that supports their beliefs, and it’s good to be aware of this bias when interpreting what the interviewee is saying.
• Get them to tell a story. Ask the interviewee to describe actual situations and be on the lookout for solution hacks. One of the strongest indications of a need is when the user has somehow already tried to solve the problem themselves.
• Drill down on the answers and parrot back. If something the user said sounds interesting, don’t be afraid to ask follow-up questions to understand the whys behind the whats. When an interviewee says something important, you can make sure you understand them correctly by repeating what you heard. They might correct you or give more details on the matter.
• Show prototypes only after the storytelling part. Otherwise, you might influence the stories. When you are showing a prototype, disarm the politeness training. Let the user know the prototype is an early version and that you want them to criticize it. The user will nevertheless be polite to you, so interpret everything they said with skepticism. Try to look for information that might kill your product rather than just looking for supporting evidence.
• Wrap things up with a summary. Summarize the main findings at the end of the interview. The user can confirm that you got everything correct or provide more details on the subject.

How to make sense of what you learned?

• Take good notes and keep a scorecard of what you learn. You want to share the findings with your team. Quantifying the interview findings can be useful. Target metrics to follow could be, for example, the share of people who recognize the problem you are trying to solve or that are spending money or time to solve it currently.
• Treat everything with skepticism. The interviewees want to be nice to you and tell you what you want to hear. Read the body language and try to understand the agendas and biases of the interviewees.
• Look for patterns and apply judgment. You don’t want to trust one interviewee’s opinion too much. Still, at the same time, you need to realize you will not be getting statistically significant results from the qualitative interviews. In the end, you will have to accept a certain level of uncertainty and commit to a decision.

My impressions

One of the top quotes for me was this:

Ultimately, you are better off moving fast and making decisions from credible patterns than dithering about in analysis paralysis.

It’s a controversial quote. In general, product development organizations should put more effort into understanding and validating the key assumptions they are making before starting to build features or products.

At the same time, it’s good to keep in mind that no amount of user research and testing can ever remove all the risks around a decision. In the end, you will have to accept some level of uncertainty in your decision-making.

Thinking of product development as betting can help you deal with the uncertainty. When betting, the general rule is that you should make a bet when the bet's expected value is positive and when the bet’s size is not too large to hurt you severely in case you lose.

The expected value of a bet depends on

• How much you win if you are right,
• How much money you need to bet, and
• How certain you are of the results.

To increase an initiative’s expected value, you want to make the bets smaller by building only what is needed to create value. To decrease the uncertainty, you do user research and verify your assumptions.

Thinking of product development as betting, you accept that you will occasionally lose. Still, you can be sure that as long as your process is solid and you keep making bets with a positive expected value, you will be profitable in the long run. This mindset can help you avoid analysis paralysis and allow you to move forward faster.

Top quotes

Ultimately, you are better off moving fast and making decisions from credible patterns than dithering about in analysis paralysis.