written by

How to kick-start a new start-up with customer interviews

Our process 5 min read


Do you have an idea for a new product? Do you want to begin a startup and become world class? Then you need to know what your customers want.

As 0smsosis, we have been in this situation often. We have a new idea for a venture and want to launch it into the world. Hopefully, it launches and everybody loves it. But more likely, people won't care. You didn't solve a problem they had, you don't improve their lives with your new product. Take Uber for example. They've provided a cheap way to get around for people, in a way that no one was thinking of. They solved a real problem, in a new way. On the other hand, there is Juicero. They saw that people wanted fresh juices. But they just provided a fancy way, not a way to be healthier.

Getting to know what people want is the first step to launch your startup. In this first step, 0smosis focuses on three important things which I will explain in this article:

  1. Take a step back: the Jobs-to-be-Done framework
  2. Finding the right people to interview
  3. Prepare and process the interview well

Take a step back: the Jobs-to-be-Done framework

You might be thinking, I'll just ask people: "What do you want?". Then you build your solution based on the answers, and you're done. Unfortunately, it doesn't work that way. As Henry Ford said: "If I asked what people wanted, they would have said faster horses." People don't know exactly what solutions they want. But we do believe people know what they want to be solved, and how they want to improve.

We use the Jobs-to-be-Done (JTBD) framework for this. JTBD says you don't need to look at what people do, but to what people want to achieve. People have goals in mind. People want progress. So look for problems instead of solutions. And build your own solution on the real, underlying problem. Two good (and free 😀) books on JTBD are When Kale and Coffee Compete and Jobs to be Done.

The Jobs to be Done for one of our ventures, Yuomi

Next, structure your own process. If you want to have consistent success, you need to have a consistent process. For example, we use our variation of the Job Map to fill in our job map. Is it perfect? No. Does it help us structure and keep the whole picture in mind? Yes it does.

Thirdly, don't keep the framework too rigid, a framework is just a framework. Certain steps can be more and less important for specific projects. Make up your own mind, mix the parts of each framework that work for you and your current problem.

Finding the right people to interview: B2B and B2C

Ok, you know your problem. Then it's time to interview people. And to interview people, you need to find those people. I've split B2B and B2C up, as they have different strategies to find the right person.


  • Appeal to people's expertise. Everyone likes to hear they're an expert. You don't need to find the world-class speaker on credits to learn what creditors are struggling with. As Dale Carnegie said in How to win friends and influence people: "Make the other person feel important – and do it sincerely"
  • Provide something up front. If you provide something at the start, no questions asked, people are in your debt. So it will be easier to ask some follow up questions. And if they say no, don't keep bugging them.
  • Contact enough people. If you think no one will answer you, still try. And keep trying! Use growth hacking techniques to connect with people, share stories and get feedback. We will cover growth hacking in Part 8, so stay tuned! We recently came across clarity.fm, a peer to peer platform to contact experts. It's a great way to find an expert in an unfamiliar field.


  • Post in online communities. There's a community for everything, from knitting, to swimming, to My Little Pony: Friendship is Magic. Engage in the community and ask for feedback.
  • Talk to people in the street. If you can't find people online, talk to people on the street. Even though 80% will walk past you, some people are happy to help.
  • Provide an incentive. And of course, providing a gift card or some cash will always help out. But don't go overboard. You want genuine feedback, not leeches that are only there for the 50€.

Prepare and process well

Awesome, you have around 5 interviews planned. Now, how do I go from those to a good view on what my customer wants? With some prep and after processing, you'll know exactly which problem to focus on.

First of all, know who you are interviewing. Understanding their context, their likes, their life. This is especially true in B2B. A banker from Goldman Sachs has very different problems that one at Argenta, a small bank in Belgium.

For the interview itself, have a (semi-)fixed list of questions. The fixed list will help with your consistency. You can also compare your interviews much better in the processing stage. But at the same time, everyone is different and every interview is unique. Allow yourself to focus on specific questions that your interviewee thinks are important. You can find our interview list here.

And lastly, document everything now and filter later. It might be tempting to only write down the important bits. But after a month, you might need to learn about that other thing. That thing you didn't write down. You can always filter your notes after the interview. Making a video or a recording will help you off later a lot.


With just five interviews, you'll find 80 to 90% of the relevant problems. But don't assume you'll know which problems to solve. A group of five respondents is not statistically relevant. Next week we'll help you how to get that statistical relevance, and be sure you tackle the right problem. Stay tuned!

jobs-to-be-done JTBD interview B2B B2C