In my last post I was talking about why knowing your customer’s Jobs to be Done is crucial for making products they’ll love. Excited to start discovering the Jobs to be Done that your customers have? Here’s a quick guide with steps to get your Jobs to be Done journey kickstarted. Mind you, there’s much more to be discovered but that’s for later posts.

The process of getting from Jobs to be Done research to product improvements can be divided into roughly 3 steps:

  1. Gathering data
  2. Analysing data to extract Jobs to be Done and their context
  3. Using discovered Jobs to be Done as a blueprint to design your product and customer journey

Here I’ll describe the first two steps in more detail and give you concrete tips for executing them.

Preparing your JTBD research

As a known saying goes: ‘good preparation is half the battle’. Before starting your interviews you first have to know what you are looking for. In this case Jobs to be Done your customers have for which they would use your product. Start with noting down the signals that are important for your research.

A Job to be Done describes the progress a person is trying to make in a given circumstance. It consists out of the progress someone is trying to make, actions that someone is taking and factors that aid and hinder progress. Jobs to be Done can be summarised in many ways, but this is the format I normally use:

  • Progress: What progress is a person trying to make?
  • Actions: What is someone doing to achieve this progress?
  • Pushes: Factors that make someone look for a new solution
  • Pulls: Factors that make someone adopt a new solution

To show an example:

  • Progress: Getting to work on time
  • Actions: Walking to the train station, buying a ticket, finding a seat
  • Pushes: “I need to get to work on time to finish everything”, “The train is often delayed”, “It’s so crammed there are no seats available”, “I cannot work from home”
  • Pulls: “I can do work while sitting in the train”, “It’s better for the environment than taking the car”, “My work is close to the station”, “I won’t have to pay any parking fees”

Collecting your JTBD data by interviewing

Interviewing is one of the best ways to uncover Jobs to be Done as it allows you to go in-depth and uncover the ‘why’ of actions. You want to find out why people use the products they do. Uncovering their Jobs to be Done and all the contextual factors that influence their decisions is key.

Here a quick guide for setting up your JTBD interview:

  • Use a semi-structured set-up. Determine the areas you want to cover and formulate questions as a guideline, rather than a strict script to follow. By creating the freedom to deviate from your questions you can dive deeper into unexpected findings that might pop up and discover the unknown unknowns, the “why” you else would’ve never found out about.
  • Focus on past actions and experiences rather than general evaluations and future predictions. People are notoriously bad at predicting their own behaviour. What they have chosen to do in the past says a lot more about their future behaviour.
  • Construct a timeline surrounding the ‘hiring’ of a service. You can focus on the ‘big hire’: buying the product, or the ‘small hire’: using the product. When did the interviewee buy the product? Where? Who else was involved? Was there anything else happening at the moment? For a nice example of this check out Bob Moesta, Chris Spiek and Jason Fried’s interview on mattresses.

The structure I usually stick to is:

  1. Introduction: Explain why you’re conducting the interview. Ask whether you can record it, if applicable the signing of a non-disclosure agreement. Mentioning that there are no wrong or right answers.
  2. Introductory questions: Asking the participant about general details like where someone lives, what someone’s job and hobbies are, etc. help to familiarise yourself with their lives, to which you can refer later on in the interviewee, and it also eases the interviewee into the conversation (interviewing can be pretty exciting, also for the one being interviewed!).
  3. Move on to the main topic(s). If you want to ask personal questions try to do this a bit later as the interview progresses and you’ve built a certain trust.
  4. End the interview: Ask whether the interviewee has any questions left, anything to add. Visibly turning off the recorder at this moment can make the interviewee feel more at ease to mention something he/she otherwise wouldn’t have. Thank the interviewee for his/her time.

While interviewing it’s best to have at least one person devoting his attention entirely to the interview. So no notes taking, rely on memory and a recording instead. But with shorter on the spot interviews this can be hard so if you need to resort to note-taking write down the ‘signals’ you’re looking for (progress, actions, pushes & pulls), preferably in the exact words someone said it.

If you have more time write the interview out as a story, sticking to the timeline surrounding the ‘hire’ you focussed on. Again include the progress, actions, pushes & pulls. And one big advice: Especially if it’s a long interview, plan at least half an hour directly after the interview to summarise it. At this moment everything is still fresh in your mind, also non-verbal cues. Seriously, it will save you a lot of time having to listen back to the interview to remember everything again.

Analysing your data to uncover Jobs to be Done

Now that you have your data the next step is to analyse it to extract important Jobs to be Done. The key to doing this is detecting patterns in what interviewees said, especially across several interviews. A way of doing this is cluster analysis, which is a way of detecting patterns of similar reactions in qualitative data.

To conduct a rough version of cluster analysis execute these steps:

  • Write down important findings and quotes on sticky notes. Focus here on indications of Jobs to be Done, actions, pushes, pulls, and other things that influenced the service the interviewee decided to hire.
  • Start clustering these based on similarity. You’ll probably be able to group the sticky notes together in many different ways. This is the big reason we’re using sticky notes, as it allows you to see new connections in your data. Scribble the connection you’re observing on a sticky note, place it near the cluster, take a picture, and you can start re-clustering again.
  • You can add other sticky notes describing the relationship between different points of data and clusters.

One of my cluster analyses ended up looking like this (snippets from interviews on yellow notes, the green ones representing JTBDs, the blue ones pulls or pushes, the pink ones describing relationships, etc.):

This process might take a while but leads to valuable insights. If you’ve discovered the Jobs to be Done that are most important for your product or service start summarising these in Jobs to be Done descriptions (see the description at the start of this article).

This tweet by Ryan Singer (product strategist at Basecamp) shows a more methodological approach to JTBD clustering.

Using JTBD insights to make product and customer journey improvements / changes

If you’ve found Jobs to be Done you can use these as a blueprint to not only guide your product and customer journey design process, but also to identify promising markets and choose your product/service design strategy in the first place. I’ll expand more on this in blog posts to come.

Start your JTBD research today

JTBD research is no rocket science. With these steps you can start uncovering important JTBDs of your (potential) customers today. The most important key take-aways: When interviewing, focus on the timeline surrounding the ‘hires’ (big or small). When analysing data focus on the similarities between different interviews. And remember what a complete JTBD consists of: the progress someone is trying to make, actions for getting there, pushes that move her to look for a new solution, and pulls that draw her towards a certain solution.

Now let’s get JTBD discovering!