How to interview potential customers using Jobs to be Done (JTBD)

If you are an aspiring founder or a designer, you know how tricky can be doing interviews with early adopters. Fortunately, the Jobs to be Done (JTBD) framework can help us, by providing a step-by-step template to use during customer discovery. When designing and developing a new product, it is essential to never lose sight of potential consumers: to study their behavior, to talk to each other, to involve them in the design process helps to significantly reduce the risk of building something that does not correspond to their needs and that they will not buy.

There are several ways to gather information about consumers. For example, Sketchin – a swiss design firm – lists 74 tools for «analyzing and understanding the behavior and needs of users». Among these, interviews are surely one of the most effective. However, conducting an interview in the correct way requires a lot of practice and several attempts. What are the right questions to ask? How should they be placed? How should the answers be analyzed?

In recent weeks, I have worked with Alessandro Zonnino to design a process to effectively teach innovators to do research and get into the shoes of who will use the products they are designing. We started by analyzing over 200 interviews conducted by the students of the Digital Product Design course I hold at the Roma Tre University with the aim of identifying the main mistakes made by people who are interviewing for the first time. Therefore, we reviewed the literature on the subject, focusing in particular on the works of Clayton Christensen, Alexander Osterwalder, Jake Knapp and the founders of IDEO, Tom and David Kelly.

The common mistakes among non-experts

Let’s start with what we learned from my students. The most common errors are essentially two. The first consists in asking questions that solicit the answers we would like to hear. It happens to those who have an idea for a new product and unconsciously search confirmation and reassurance; they are afraid to discover that the brilliant idea they had is not so ingenious. This dynamic has been brilliantly explained by Rob Fitzpatrick in the book The Mom Test. There are many ways to get comfortable answers. For example, if I had invented an app to order coconut on the beach, I could ask for confirmations by asking: «have you ever bought coconut on the beach?» or «would you like to order coconut directly from the umbrella?» If you are thinking that only a stupid would ask such questions, I can assure you that they are much more frequent than you imagine. When I talk to aspiring entrepreneurs about their ideas, I always ask if they have talked to their potential customers. If they answer «sure and everyone told us they would buy it» it means that they (unknowingly) made interviews to be told that the idea is beautiful. Some add: «we also talked with experts and they told us they were very interested».

The second most common mistake I have seen is to use closed questions or even to construct a questionnaire, turning a qualitative interview into a quantitative survey. The two instruments have completely different objectives: the purpose of an in-depth interview is to explore a topic to discover the factors that characterize it. Conversely, a questionnaire assumes that you know the dimensions to be investigated and that you want to measure the frequencies. So, you need to learn to ask open questions that allow you to explore the interviewee’s world to gather useful information for the design activity.

Unfortunately, it is not enough. Even when one has learned to avoid comforting questions and to listen to user stories, there remains a fundamental problem: what to ask? How do you get the most useful information out? Is it possible to follow a model? We started from Clayton Christensen’s Job to Be Done theory.

Jobs to Be Done

In Competing against luck, Christensen defines a job as «the progress that a person is trying to do in a given circumstance». In other words, each of us sets goals and moves in the direction that allows him to reach them. In this path, he hires products and services that allow him to make progress in the desired direction.

The Job to Be Done model is semantically richer than the more popular approach that suggests that a person purchases a product because it represents the solution to a problem. In fact, a job is not necessarily a problem. Christensen offers many non-obvious examples to explain his theory. For example, why do fast-food customers buy a milkshake? The answer is not univocal: in the morning, commuters buy it because they need to “suck” something during a boring journey. In the afternoon, parents buy it as a reward for their children. The same product performs two completely different jobs and competes with completely different categories of alternatives. Being aware of this, it is possible to discover many opportunities to improve the characteristics of the product and its desirability towards specific segments of customers.

The transition from problems to jobs involves a substantial change of perspective. In interviews, we are no longer looking for problems to be solved (it is often a frustrating research) but for the progress that people are trying to do under certain circumstances. The latter have a decisive impact on a person’s ability to achieve the progress of the job. In fact, circumstances determine the obstacles and the problems that are involved in achieving a goal. At the same time, they set people’s expectations about the results they can achieve.

The Job to Be Done theory was also used and synthesized by Alexander Osterwalder in his Value Proposition Canvas. According to the Swiss scholar, it is useful to distinguish between functional jobs (when people try to solve a practical problem, like how to cut grass or eat healthily), social jobs (when people want to appear positively or gain power) and emotional jobs (when people look for a specific feeling like safety). At the same time, Osterwalder emphasizes that jobs are not all the same and that consumers will tend to deal with problems based on their negative (pain) or positive (gain) consequences.
Pains describe everything that disturbs or prevents you from completing a job. These may be undesirable results, material obstacles or risks to be run. On the other hand, gains describe the desired benefits and can be classified as indispensable, expected, desired and unexpected.

The Value Proposition Canvas suggests that a desirable product should help the client accomplish his job by minimizing pain and maximizing gains. Obviously, it’s easier said than done.

Jobs theory gives us a very useful model for setting up consumer research and for conducting in-depth interviews.

The innovator’s point of view

It should be emphasized that no one deals with research without information and opinions, especially if he already has an idea for a product. Indeed, more often than not, the idea arises precisely because the innovator has experienced a problem and has started thinking about how to solve it and therefore has an opinion already formed on the opportunity space.

Having a solution in mind is probably the biggest obstacle when talking to potential customers for two reasons: 1) it’s the spring that triggers the reassuring questions we talked about earlier; 2) it leads to explore some dimensions to the detriment of others, highlighting some information and underestimating others.

Steve Portigal in Interviewing users: how to uncover compelling insights suggests to “free the mind” by adopting rituals before starting to design and conduct the interviews. The first is the brain dump: the team shares hypotheses, expectations, beliefs, perspectives on what will happen. Saying it aloud and writing it on a whiteboard, the team brings these themes in a neutral space. This exercise can be usefully integrated by asking experts what their point of view is on the topic under investigation. This is what Jake Knapp suggests in Sprint. How to solve big problems and tests new ideas in just five days who writes: «nobody knows everything. Instead, information is distributed asymmetrically».

Sharing what we known about the topic we want to investigate and gathering information from experts help us to make an assessment of what we think we know about a certain argument and to identify the issues to be investigated. In these days we are conducting a series of interviews to “fine tune” the syllabus of the Innovation School of Rome: it is a useful example to illustrate how we are building the framework. Both Alessandro and I have a certain amount of information and opinions on the topic: we have read a lot of literature, we have experimented with two different syllabuses for the course of Digital Product Design at Roma Tre University. I have several years of experience both as a small entrepreneur and as a mentor in acceleration programs. It is the classic case of a team who have identified a solution by relying on their professional and personal history. The risk of self-reference is around the corner.

All this information allowed us to identify a series of relevant topics when we talk about designing and developing a new product. These include: team, project development, mentor and coach, fundraising, and so on.

For each theme, we have identified a list of useful questions to conduct the interview. For example, in the case of the team, we were interested in knowing how it was formed, if it is complete, what the skills are, what the dynamics are among the team members and so on. It is important to make a list of possible questions even if we do not use them all: they will help us to govern the conversation.

Conduct the interview

Once the topics to be explored have been identified and after having identified the questions to be asked to explore the individual themes, we can begin our interviews. The best way to do this is to use open questions and let the interviewer talk: the goal is to build a small documentary. For example, the interviews we are doing for the Innovation School of Rome to aspiring entrepreneurs start with a simple “tell me about you” and continue with a “tell me your idea”. After these two warm-up questions, we move on to explore the history of the project and the team: «Now, I would like you to recall the steps you have made so far. Let’s start from the beginning: how did your idea come to your mind?»

As the interviewee tells his story, he also highlights his Jobs to Be Done, that is, the goals that have kept him busy, that have created concerns or that he has not yet managed to achieve. In the interviews we have conducted so far for the school, for example, there are those who stressed the need to find a technical partner, those who must launch the product, those who feels the need to get advice from experienced people, whose who needs to fundraise and so on. In other words, the general job “I have to create a startup” contains a series of smaller jobs, which are the progress that the interviewees are trying to do to reach the goal.

After having identified a certain number of important jobs for the interviewee, we can go in depth and ask more specific questions. These have the objective of analyzing what is the current approach, what are the pains and what the criteria of success. For example:

We are experimenting with different templates to explore the facets of a job. The one proposed in this table is a simplification of the Job Atlas elaborated by Stephen Wunker in the book Jobs to be done: a roadmap for customer-centered innovation.

Conclusions

Interviewing dozens of people is a long and tiring process that does not offer many shortcuts. In this article, I described a framework for designing and conducting interviews based on specialized literature and my personal experiences as a teacher. Furthermore, it is a model that, in addition to being methodologically robust, should be easy to explain to those who have never done research on users.

Obviously, this article does not exhaust the theme. In addition to the effort and time to carry out an interview, in fact, it is also necessary to consider what is needed to process the information in a structured way and to identify the most useful data. This topic will be the subject of the next post.