(2025-07-09) Chin Putting The Jobs To Be Done Interview To Practice

Cedric Chin: Putting the Jobs to be Done Interview to Practice. Much has already been written about the interview method, but I thought it would be useful to describe what we’ve learnt from doing it. This includes things that we struggled with, things we found oddly tricky, and things that we wished someone had told us earlier.

The first thing you’ll discover about the JTBD interview is that it’s really hard to do. I think this is obvious from reading between the lines of all the major JTBD texts: Demand Side Sales 101 from Bob Moesta is filled with sidebars about how things can go wrong with the interview process; the team over at the Re-Wired saw it necessary to release The Jobs to be Done Handbook — a 66 page booklet designed to be skimmed right before an interview — which gives you a hint that the interview process is not as straightforward as it seems.

I think the main reason is that the JTBD is a form of ethnographic interview, and if you’ve never done ethnographic research before, you’d have to pick up the basics as you execute.

If we take a step back, the JTBD interview is actually really simple. It is just a way of getting your customers to tell you the story of how they bought your product.

The problem with it is two-fold. First, most customers can’t remember how they bought your thing. Second, the human mind is not designed for self knowledge. Humans will make up reasons for why they do things, and you have to catch them when they do. Skill at the JTBD interview is about getting better at both things separately.

Of the two, the first problem is easier to solve. The JTBD texts have a list of tricks you can use to get your customer to remember. The most basic one is to never interview a customer who bought too long ago

This leads to a counter-intuitive recommendation: at the beginning of your JTBD journey, limit your interviews to customers who have purchased less than six months ago.

handful of tips for dealing with customers who can’t remember:

Take the pressure off — Open every JTBD interview with “Imagine that I’m filming a documentary. I’m just trying to understand

Work from the purchase backwards — The JTBD framework is built around the JTBD ‘timeline’ — that is, the series of events that led to the purchase.

Ask for specifics — Always ask for the name of the dog. If they mention that they purchased at the kitchen table, ask them if they did it in the morning, or at night. If they tell you they’re married, ask for their spouse’s name.

Asking for concrete detail helps with recall.

Use environmental cues to hammer down the timeline. If they say “I think I bought it last year” ask them: “What was the weather like when you bought it? Was it hot out?

The second problem — that humans are not made for self knowledge — is the harder problem to tackle

A strong assumption behind the JTBD framework is that “things don’t happen randomly” and “there is causality in human behaviour.”

for the specific context of this interview, it is helpful to believe that this is the case, and therefore it is helpful to act as if it were true. (All Models Are Wrong, But Some Are Useful)

Pushing for causal factors is tricky, because you don’t want to push too hard.

Noticing that something they’ve said doesn’t make sense, or doesn’t tally with other pieces of information

Pointing that out, gently, so you can dig deeper

Knowing when to change topics so they don’t feel pushed and become defensive.

I don’t have a good answer for how to get better at this, beyond “put the reps in, and get feedback from your team

The authors of JTBD have another standard recommendation that I didn’t take seriously, but should’ve: they recommend that you conduct the interview with one other person, as a pair

running the JTBD interview is very cognitively demanding, and it helps to have a second brain catch you in case you miss something.

This complicates the interview process, though: it now means that you have to find a partner, schedule them for each and every customer interview, and take the time to debrief, reflect, and improve as a team

we ran practice interviews with each other before executing it with our customers. I highly recommend this.

Practice Before Each Interview Project

Pick a friend or a family member, and then pick a purchase that they made not too long ago. (Make sure the purchase is not a gift, nor an impulse buy

Unlike Sales Safari, JTBD interview analysis is comparatively easy to do

You are expected to fill in the following format:

For each of the points in the timeline, you are expected to jot down the four forces that acted on them in that moment. This is no more difficult than a listening comprehension test — assuming you’ve done the interview correctly.

if your interview was badly done, you’ll find that you can’t fill in the template

It’s worth asking why qualitatively customer analysis is so difficult

Sales Safari is difficult because everything that a customer says has two kinds of informational value:

Content value — the information value of what is being said. Positioning value — the information value of why something is being said, and how it is being said.

The most revealing, most valuable bits of any customer research interview lie in the positioning component of what is said, not the actual contents of what is being said

We may get more specific about what positioning value looks like, of course. When you’re listening to an interview, ask:
Why are they saying this?
What words are they using? Why are they using those words, instead of alternatives, when describing their situation?
What psychological state are they in right now? What psychological state were they in back then?
What is the situation that they find themselves in at the time of purchase?

This is still only half the picture. The other half comes from noticing patterns across customers. Things like “wait, isn’t it odd that five customers have said the exact same thing? What’s going on here?”

JTBD is slightly easier, of course, because it merely focuses on explicating a buyer’s timeline. But to get the most out of this process, you still need to analyse a little for why customers are doing the things they do — and this analysis is where most of the cognitive costs will lie.

What did our JTBD interviews uncover? We learnt a number of things, but most of those findings will require too much context to understand. Instead, I’ll describe just one finding to you, as an example of how the JTBD framework can be useful — especially when used in combination with a data capability.

Commoncog runs a membership subscription, and — roughly speaking — has two parallel paths in its growth loop.

In the first path, a reader signs up for the newsletter, and becomes a regular subscriber for some indeterminate amount of time. At some point, they become a paying member.

In the second path, a reader discovers Commoncog, and becomes a repeat reader

They never once sign up for the newsletter. After some indeterminate amount of time, they become a paying member.

We know both paths exist

What we didn’t know — or had even bothered to find out — was how many paying members became paying members through the first path vs the second. I assumed that more members came from the first path.

During the JTBD interviews, we began to realise that members who became paying customers through the first path experienced a different timeline from readers who became paying customers through the second path.

members who had been newsletter subscribers became a member in the following manner:
After a period of months — sometimes years — Commoncog publishes a paywalled article that just happens to be about something they’re dealing with at work or in their lives. This serves as the buying trigger and gets them to take out the credit card and buy.

On the other hand, members who become paying customers through the second path:
they discover a post that belongs to a series of articles, and then they read every single article in that series that is not paywalled.

they decide that they really really want access to some of those paywalled articles that they’d skipped over in a series. They decide to purchase.

We learnt that 56.5% of members came from the second path. Only 43.5% were newsletter subscribers before purchasing.

Hien then created the following visualisation:

In the JTBD framework, purchases are made because of a buying trigger

What this graph shows us is this: in any given month, the number of paying members who bought because of a paywalled article would typically outstripped those who bought after becoming a regular newsletter reader

But in two of those months, newsletter readers who became paying members far outstripped the number of new members who were website readers. Those two months corresponded to the two times we increased prices of the membership program

One conclusion might be that Commoncog’s email list contains a list of potential customers in passive looking. Creating a time wall should nudge these individuals towards a buy decision.

How is this Knowledge?

It is not enough to have findings, in the way that I’ve described, above — we want to know how it can be predictive because we want to be able to act on it.

One piece of knowledge is clear: more people become paying Commoncog members through the second path, so we should focus on that. Notice that this may be stated in the form of a prediction: “if we focus on the second path, we should see higher conversions to paying members.”

But: how? There is, of course, an obvious hypothesis: because the majority of repeat readers paid after encountering a paywalled article (in the context of a series), we should modify the Commoncog site to increase traffic and ease discovery for published series

Notice that we still have to test this. All we have right now is a correlation

There is one other hunch that we can chase down. The price increase and subsequent spike in purchases tell us that a good number of mailing list subscribers are in the ‘passive looking’ phase of the buyer’s journey

Creating a time wall (such as sending an expiring discount to the list) should result in a burst of signups.
Of course, whether we want to do that is a different matter. Commoncog doesn’t discount right now. Are there other ways to create a time wall?

into the following series of steps:
Figure out how your customers buy, either by using data, or via qualitative interviews (Sales Safari, or JTBD, or DPIs).
Ask: “how is this knowledge?”, which forces you to come up with predictions. (“Doing X will result in Y after Z time”).
List out a number of experiments for these predictions and prioritise those experiments

Let’s pull back to talk about demand.

I recommend both Demand Side Sales 101 and The JTBD Handbook. Both are very short books — you should be able to finish them in a single weekend