Don’t ask them what they want, ask them what they are trying to achieve.
Customer interviews should not be about asking customers what they want (they often don’t know what is possible), but rather understanding what problems they currently have, what outcome they are trying to achieve and what is difficult or frustrating with the current solutions available.
You can use the information to determine how important the problem is to them, how actively they are looking to solve it and what the product solution could be.
The beauty is that you don’t need to do that many. In most cases, after 8-10 interviews of a particular customer segment, you will start to hear the same problems or desires which means you can stop, as additional interviews don’t provide new information.
Speak to different types of customers.
Every product idea will likely appeal to a number of different types of customer.
For example, you are building an app to help people learn to play the guitar. You might split these customers into age groups. Young guitar players learning with a teacher may have a very different set of problems compared to an adult first-timer who doesn’t want to try and teach themselves alone. For building great products and getting them off the ground it’s better to pick one group first and provide an amazing experience for them while keeping an eye out for the larger market.
Before you start, brainstorm the different types of customer segments there might be and then try to speak to 8-10 people in each major bucket. Don’t worry about going too niche here in terms of types of customers.
Another example for a B2B business would be to segment by size of company. Smaller companies will have less complexity and different purchase behaviours compared to a mid or large size business.
Finding customers to talk to.
You want to be able to talk to each person for 60 mins. Ideally face to face, but over a video call is also ok.
Focus groups don’t work, as they introduce a dynamic where people try to impress each other or are influenced by what others are saying. So think about this before you go out trying to find people to talk to. It is also ok to offer a little incentive for people to give you this time.
Surveys have their place for generating higher level statistics but they won’t give you the deep insight that 1-1 interviews provide.
Here are my preferred ways to find people to talk to (in order of preference).
1. Use your network as much as you can or infiltrate a network that you can then ask. From meetups to Facebook groups, there are lots of ways that your potential customers are connecting with each other and you should try to get into those groups to help your understanding of their problems.
2. Good founders are not afraid to hustle. Find places where your potential customers are and go ask them for some of their time. You will be surprised how many people will feel happy speaking to you.
3. Lastly, if you have the budget you can always pay to speak to people. I am not a great fan of this due to the amount of money involved (not to mention it’s good practice hustling.)
If you have the cash, there are a number of agencies that will find the right type of people that fit your demographic. Expect to pay around £50+ or $60+ for each customer to the agency in the consumer space and a lot more for professionals or enterprise customers.
Create a discussion guide.
As a pre-product founder, you are in information-seeking mode. It’s not the time to sell your idea. It is important to be mindful of the questions you ask, be inquisitive, interested and always keep the questions open.
A good indicator that you are asking the right types of questions is that you are listening over 80% of the time. Even better, the format of the interviews is roughly the same regardless of the company.
Tailor the following format for your company, and it should be optimised to extract actionable information.
Very brief intro to the general theme of the company.
It's ok to say "I am working on a new idea in the music education space".
Ask for demographic data.
Get them talking with easy questions to get the conversation flowing. These short answer questions will also help you categorise the responses and create personas and user journey maps. In the “learning to play guitar” scenario, you might want to ask their age, when they started to play the guitar, etc.
Ask about triggers for them starting to look for a solution.
It’s useful to ask your customer about triggers for when they started to look for a solution to their problem if that’s relevant.
Ask about real examples of their problem and how they have solved it in the past.
The main part of the interview is to get them to run through their past experiences. E.g. “Can you tell me about your most recent guitar lesson?”. By talking about the past you are eliminating a lot of bias. You are helping them express the steps they took to solve the problem, understand their desired outcomes and the feelings they had along that journey.
Some good follow-up questions to their statements include:
- Can you tell me about that in more detail?
- What was frustrating?
- What were you trying to achieve?
- How did that make you feel?
- That’s interesting, tell me more?
- What do you feel about it now?
- What were the fun parts of this?
Share the concept.
This can give you a reaction to the initial idea. A good follow-up question might be:
Is there any reason why you wouldn’t want to use a service like that?
Follow up with a 2x2 feature card sorting exercise.
If you have lots of ideas for features to your product solution you can write them out on cards and then ask them to rank them on an x-axis of not important to important and then a y-axis if you want as well.
For example, in our learning guitar product, you might have the following ideas for features:
- Videos taught by famous guitarists
- Close-ups of finger positions as music played
- A catalogue of music of popular songs
- The ability to listen to you play and tell you where you are wrong
You can then ask how they rank each of these by importance and by how delightful they would find them.
Make sure to ask why they are ranking things a certain way.
It’s a good idea to ask “why” multiple times in order to reveal the emotional reasons for those choices rather than just the functional ones.
Ask for a referral to some more people to interview.
The people you are interviewing will very often know other people like them. So make sure to ask them for introductions.
Record the interviews.
Whether you run these interviews as a pair or by yourself, it’s preferable to also record the conversations so you have all the data. Make sure to ask permission to record them. This will allow the conversation to flow without you losing the detail.
Often, your interviewees will want to jump around but it will help you if you can capture the steps sequentially. You will find yourself pulling the interviewee back from their flow which can be uncomfortable for them but it will get you richer data and create a sequential timeline. You can always take notes of the jumps and then you can refer back to them. Transcribing services are pretty cheap these days so make sure to use a decent microphone when you record the conversation.
Stop when you start hearing the same responses.
Interviewing 8-10 customers is the minimum number to make sure you have enough material for each customer segment. If you can do more then that is great, however, you should stop when you start getting similar responses.
You are learning nothing new and it is far better to build a prototype of the product and get feedback on that as it will give you richer information on your solution.
What to do when the interviews are complete.
Once the interviews are complete you can use them to create personas and experience maps to then help you ideate on solutions to solve the problem.
Alternatively, the interviews might tell you that the problem is that big of a problem and you might need to keep looking.
