In Brene Brown’s new book, Atlas of the Heart, she talks about the research for the book and how they took 550,000 comments from an online course and turned them into quantitative insights.
It’s a genius way of doing research because there are some questions you just can’t answer without talking to people. Still, you also want to know your data is reliable—and that’s where turning customer interviews into quantitative data comes in.
With this method, you’re pulling concrete data from the information people give you through in-person or online customer interviews. Here’s how it works.
4-part framework for turning customer interviews into quantitative data
Identify questions to ask
Asking good questions is an art, prompting people to open up and want to share. Being too formulaic will cause people to shut down and give formulaic answers. We’ve all seen surveys or questionnaires that are so dry, we just want to speed through and finish. Asking questions that draw people out is necessary for conducting effective customer interviews.
Asking good questions is also a science. You need to present every interviewee with the same questions in the same way. If you start to lead customers toward specific answers or fill in the blanks for them, the bias created will ruin your whole interview.
We recommend starting with a small set of questions that you know you want answered. Try those questions out in a few customer interviews, noticing which additional questions naturally come up. Then, add those questions to your interview bank for future calls. It’s okay to go back and interview the same people twice if you end up generating a bunch of new questions during the interview process.
Here’s an example from a set of interviews we’ve run at Interact a number of times. We ask 19 questions in the interviews, pegging them at about an hour. This list started with seven questions, but as we ran interviews, new ideas came up, bringing the total to 19. We ask each customer these same exact questions during live phone calls.
We recommend live phone calls instead of surveys because you can’t really control who is being interviewed in a survey. Plus, you’re less likely to dig deep into an answer without the back and forth of a live conversation.
Gather a set of customers to interview
We can talk for days about selection criteria and biases, but at its most basic, you want to select a group of people who represent the categories of customers with whom you work. You’ll need to know your customer personas, and then you can select a few customers from each persona.
You may also want to consider the following factors:
- How long has each person been a customer?
- Are they having success with your product?
- Should you talk to people who aren’t customers?
- Should you talk to people who have canceled?
You don’t have to have everything ironed out before you begin interviews. Often, we find that after a few conversations, things start to solidify, and we get clearer on who is a good fit for the research project.
Analyze customer answers
Once the interviews are done (we recommend a minimum of 15 interviews), it’s time to look for patterns in the responses, which will help you create quantitative data. In its simplest form, you’re counting instances of what people say and comparing it to the overall number of interviews you do. Then you’ll have your data.
This post on content marketing for coaches is a great example. We interviewed 30 coaches and pulled insights from their responses about what works for them in content marketing. From there, we put together a solid data-driven blog post.
Below is a screenshot of one person’s responses. We put all the responses in tabs, then compared the content of each person’s answers to develop quantitative data about the overall picture of how people are answering questions.
Draw conclusions
Fifteen customers responding similarly is a solid number to go off of. Depending on how big the decision is, some people will go up to hundreds, thousands, or more.
You’re mainly looking for patterns. As you have conversations, stick to the script and don’t lead your customers to certain answers. Listen and notice when the same answers come up over and over again.
Once information starts to pour in, you can reality check the conclusions against historical data from other companies in your industry. Look back in time at what has worked well, connect it to the present, and you have a nice gut check.
You can also check your conclusions against your intuition framework, ensuring that whatever ideas come out of your data align with what you know about your customer base already.
Putting it into practice
Next time you’re faced with a big challenge in your business—whether it’s a slow-down in growth or needing to decide on product updates—try the customer-interview-to-quantitative-data process. It will give you the benefits of qualitative research because you’re actually talking to people. It will also give you the reassurance of quantitative research and knowing you have reliable, clean data.
Here are the four steps again for reference.
- Identify questions to ask
- Gather a set of customers to interview
- Analyze customer answers
- Draw conclusions