Effective Customer Interviews
Interviewing customers is an important part of Product Management. We talk to customers to identify needs, understand problems, empathize with our users, and get feedback on current products and processes. The way we interview for each of these, though, varies slightly. We wouldn't use the same interview tactics to find out a user's current problems as we would to conduct a usability test. This lesson will teach you how to interview users in order to identify pain points, problems, and needs. Watch the video below for an introduction.
When we interview people about their frustrations and pain points, we are trying to identify the needs we could fill with our software and features. It's important to focus on the problem here. We don't want to pitch users and try to upsell them on products. We want to understand and empathize with their current situation.
While conducting user problem interviews, remember to:
- Qualify the User: Make sure the user is in your target market by asking questions that reflect this as quickly as possible. You can also use a quick survey to screen participants before the interview.
- Ask Open Ended Questions: It's really easy to slip into a pattern of asking yes and no questions and forgetting to dive deeper. "Do you like this?" or "Would you use this?" are examples of close-ended questions. You would need to follow up with “why?” in order to understand the context and get anything out of their response. Instead, ask questions that result in the user talking and explaining more.
- Don't Ask the User to Speculate: Here's the dirty truth of interviewing users - they lie. People like to lead you to believe that they do things, when often times they actually don’t. If you asked me, "Do you usually go the gym?" I would answer "Yes, all the time!" But, if you asked me when the last time I went to the gym was, I’d tell you that it was actually two weeks ago. In order to get more accurate stories and have users tap into examples, start off your questions with "Tell me about a time when...".
- Never Ask "What Do you Want Me to Build?": Many Product Managers get frustrated with users because they will build them exactly what they requested, and then the user doesn’t use it. This happens when you ask too much of the user. It's not the user's job to solve their own problems, it’s YOUR job. Instead, focus on their problems. That way you’ll come up with the best solution that will meet the needs of many users, not just one. Get your user talking about their problems, their frustrations, and their alternative solutions they’ve tried but failed at.
- Don’t Rely On Surveys: At this point, many people ask, "Why can't we just send out a survey to get feedback?" Surveys have a time and a place- and this is not one of them. You can use surveys to see if your ideas apply to a wide range of people, but this only works after you know what you should be asking. At this stage of product development, we are trying to understand our users better. Only a nuanced discussion will lead us to the answers we need. If you try to give your users a survey at this point, you are essentially locking them into a series of close-ended questions. You might not even be asking the right questions. For example, if I ask the survey question, "Do you want dashboards?" I won't understand why people think they want dashboards. Interview people first to understand their pain points and context. After you interview 6-8 people who have the same patterns or problems, and you understand their context well, then you can use surveys to identify and screen more people to talk to.
Tips for the Interview:
When it comes time to actually conduct the interview, here are a few helpful tips to get you through successfully:
- Start with an objective. What are you trying to learn from these sessions?
- Brainstorm questions early but leave some room to go with the flow. You want to be able to start off easily, but then follow where the user's story takes you. You'll notice in my video I was looking at some notes to remind myself what to ask. This is good, but don't worry about hitting every question on the list.
- Bring someone else to the interview to take notes. It's easy to misinterpret someone's responses. Having another person there helps you digest the information and talk about what you learned. I love taking developers to my user interviews. They appreciate hearing and learning about our customers, which helps them do their jobs better.
- If you can record the interview (just audio) that's usually very helpful. Make sure you get your user's permission first. I typically just use the standard iPhone voice recorder.
- Circulate the results of the interview. Make sure you share what you learned with your team. I will write up a one-page bulleted list of things I learned, things I confirmed, and things I disproved.
How to Find Users
One of the hardest parts of customer interviews is actually finding people to interview. Lean Startup teaches that you can go into any coffee shop and shove a prototype in front of someone to get feedback. You'll get feedback, yes, but not very useful feedback.
Remember, we want to qualify our users first. The best way to do this is find people who are experiencing the problem you are solving. Think about the context in which your users would experience the problem, and then go to the source.
For example, if you were working for a restaurant point of sale system, go to a restaurant. Watch people use the old system. Ask the waitresses what they think. If you're making a grocery application, go to a grocery store. Ask people for five minutes of their time and identify their problems.
What about gift cards? I've seen many businesses offer people $50 Amazon gift cards, or even cash, in exchange for 30 minutes of their time. There are pros and cons to gift cards. The pro: you get lots of people to answer you. The con: they may not be the right people. You want to attract people based on the problems that they are experiencing - not the fact that they are looking for easy money. If people have a strong problem, they will answer you. You can offer incentives in the form of a free month of your product, or other credits, but don't give away free money. There are tons of people who prowl Craigslist looking for these opportunities and they are not your target users. If you have a rigorous process to make sure someone has a problem and they are your target user, then go right ahead and offer them a gift card. But remember, if users have a strong problem, they're usually happy to talk to you for free in exchange for a potential solution.
If you are working for a business-to-business company, you can't go out into the world as easily. But when you are thinking about a new product or feature it's still just as important to gather user feedback. Try to engage other customer facing internal teams (sales, account management, customer service) as those teams are often a great proxy for users or able to connect you directly with customers. If you tell Account Management / Customer Service about the problem you are working on, they may have customers that have complained about the current feature and might be open to making an introduction. Similarly, the Sales team may be able to get you in contact with a prospect who ultimately chose to work with a competitor because that feature was not available. These customers may eventually become part of a Customer Board, which guides product development moving forward. Try and talk directly with the users to ask clarifying questions and get to the heart of the issue. It can be a meeting where the AM/Sales Rep is there as well, but direct communication is ideal. Alternatively, you can gather quantitative feedback by creating a survey and asking Account Managers to send out to customers. Creating these internal relationships will be helpful in choosing beta testers and potentially customer testimonials to help with marketing when you are ready to launch.
How Many Users Should I Talk To?
This is a huge debate in most companies. How many users do I need to talk to? Many companies think that user research is a waste of time because it requires people to spend 100 hours talking to users before building something. This is a complete myth.
I love this article from Neilsen Norman Group, the brainchild of Don Norman, one of the father's of UX. It says to get 80% of the information you need to make a decision on what to do next, you just have to talk with 5-9 users. That's it! Seriously, I can tell you from experience it works.
I have done the 20 one-hour sessions, and after about 9 people you hear the same thing over and over again. You feel more confident that you understand the problem. In these sessions you are trying to look for trends. If you have 9 people who all say something different, try narrowing down your target audience and attracting people in that.
Remember that when we are talking with users to understand their problems and the context around when they experience it, we're not looking for them to tell us exactly what to do. We just need to be 80% certain that the next thing we try is in the right direction. If it isn't, we figure out where we went wrong by talking to users.
User research is supposed to point you to the right path, not give you all the answers. At the beginning of every project, take 4-5 hours and schedule 30 minute phone calls. Go out and find people so you can watch them experience the problem. Just this short time is going to give you a ton of learning to get started.
Remember to Empathize, Not Sympathize
In the video about Building Empathy, we saw the difference between empathizing and sympathizing. When you are listening to your user's responses it's important to do the former. Sympathizing leads to trying to solve their problems, but ineffectively. This ruins the interview.
For example, I was once the head of UX at a company where I had to fight to get buy-in to interview my users over many months. The management team was "afraid" to let me talk to users without a sales person present (crazy, I know). So when I finally got permission to go interview a user, I was sent out with a sales person.
While I was asking the user about their current problems and frustrations with their workflow, the salesperson tried to solve them immediately. "Oh, but you can use this feature for that problem!" It was completely disruptive to the interview, and I didn't get good information out of it.
It's important to bring more people into the research process, but you want to make sure they are well trained. Teach them the basic rules of customer interviews. Tell them you will lead the first few times, and then they can jump in after a while. Teach them the difference between empathy and sympathy, and how empathy is what you are after.
Time to talk to users! Target and find a few users to interview. Start by identifying your objective for the interviews (identify problems, empathize, get feedback). Then write out the list of questions you would like to ask. Invite a developer to go along with you!
No buy-in to interview people? How can you get around this? Consider this example. One of my clients is a large bank. One Product Manager wasn't allowed to talk to users because there was a separate department that did that. So she talked to her manager about how to get the information she needed. He suggested she sit in a bank's office, and ask questions casually of people as they came in. She only needed to do that for 20 minutes and she got all the information she needed. How can you (legally) skirt the rules to do what you need to do?
Talking to Humans - Free e-book by Giff Constable
It's Our Research - by Tomer Sharon
Validating Product Ideas through Lean User Research - by Tomer Sharon