Why does someone switch from one product to another? It’s rarely the first reason they’ll offer. You have to dig deeper to find out, and that's where Jobs-to-be-Done comes in.
Bob Moesta pioneered the Jobs-to-be-Done (JTBD) framework in the mid 90’s, alongside Harvard Business School Professor Clayton Christensen. In short, JTBD is a research process that helps uncover a customer’s motivation for buying your product – the “job” your product is“hired” to complete.
Today, Bob is President and CEO at The Re-Wired Group, a consultancy that’s helped develop more than 3,500 products and services. They helped us uncover the exact jobs our products were used for in Intercom’s early years, a process that resonated with us so much that we wrote a book about it.
Bob’s appeared on our podcast to talk JTBD previously, and I welcomed him back to talk about how we can continue getting better at unpacking customer motivations. If you enjoy the conversation check out more episodes. You can subscribe on iTunes, stream on Spotify or grab the RSS feed in your player of choice.
What follows is a lightly edited transcript of the conversation. Short on time? Here are five quick takeaways:
- Customers don’t randomly switch products. Either a better option comes along or something disrupts their routine, and understanding these forces not only helps you acquire users, but retain them as well.
- For larger companies to best utilize JTBD, you have to filter buyer behavior into quantitative data sets.
- The way you package and market your product will change over time, but the jobs your product is hired to do should be timeless.
- The progress your users are trying to make is the same; however, they’ll use very different language to describe it. This is where personas come into play.
- The greatest single step you can make with JTBD is to talk to somebody who recently purchased your product and somebody who recently quit. What drove their decision?
Bob Moesta: The basic premise of Jobs-to-be-Done (JTBD) is that people buy things to help them make progress. There’s nothing random about somebody buying an Intercom product today, for example. There’s something that caused them to do that: an outcome they seek or some better state they want. JTBD exists to understand both the context a customer is in and the dominoes that have to fall for them to say, “Today’s the day I’m going to buy.” What’s the outcome that sets their expectations? When are they actually switching? It’s the notion of really focusing in on those moments through space and time. What can you do to help people understand when it’s time to sign up for Intercom and to make sure you satisfy them so they’ll stay?
Des: It sounds slightly magical. How do you begin to learn what these moments are?
Bob: I usually start with interviews to understand the difference between what they say and what they do, and to understand the things that happen to them or the things they want to happen – what we call the “causal mechanisms.” From qualitative interviews, you’re able to actually see behaviors, and you can look into the data analytics and actually see what’s happening. It’s almost like theory development, to understand how to look at big data. That’s how we end up being able to see it and size it, and then we can decide how to actually build better experiences to actually deliver on the jobs.
Finding underlying causality
Des: It sounds like people will say why they buy something, but that’s not really why they bought it. Can you give an example?
Bob: I have a great story: the other day I was at dinner, and one of the gentlemen had the new Google phone. I asked what made him decide to buy it, and he said, “I looked at the iPhone X, and I looked at the notch, and that notch is the thing that caused me not to buy the iPhone and to buy the Google one.” And I said, “I don’t think so.”
You have to get past the surface crap people tell themselves.
All of a sudden, we start to talk through what’s wrong with the notch, and then talk about what else was going on, and I learn that he actually needed to build an Android app. He had to buy an Android phone to begin with. The notch had nothing to do with the story, but it’s an example of the easy thing, or the lie we that we can tell everybody else and that everybody will accept. It’s really not the underlying causality, so you have to get past the surface crap that people tell themselves.
Des: One of the techniques I’ve seen in your JTBD interviews is that you sort of freeze time. If someone tells you they just bought a mattress, you say, “You don’t just wake up out of bed saying ‘I need to buy a mattress’ or ‘I need to install Intercom.’” When you press people enough, they say, “Well actually, the spring dug into my back,” or “I got this shitty email report that wasn’t good enough, and I needed something better.”
Bob: Someone might tell you: “I went to give a report, and all of a sudden my bosses went crazy because it wasn’t the right data, and they made me look bad. So I have to find something better.” It’s usually the things they blame themselves for. They don’t say it’s about the product; it’s a separation between their experiences and product. You have to dig deeper than that: it’s really about seeing how products fit into people’s lives.
Trying to look at your customer through your product is like looking through a peephole in a fence. You can only see the little interactions they have, as opposed to getting above it all, looking at their life, and seeing how you actually fit in. That’s where the interview takes a turn, because most people always think you’re going to talk about the product. Instead, you’re talking about them.
When a lot of companies first start using JTBD they think, “I need you to ask about this feature and that feature.” I don’t care about any of those things. I only care about what was going on in their life that made them say, “Today’s the day.” Those are the pylons and the foundations by which people do things. They don’t think it’s part of your world as a product person, but they are the actual foundations by which you get pulled into their world.
Des: For business owners, startup founders, and anyone from designers to product folks, JTBD is a way to understand the situations and motivations that would trigger your would-be customers to evaluate your product. If you learn those things, you can lean into them. You can try to trigger that motivation earlier, right?
Bob: There are two flavors of jobs out there. One is the outcome driven-innovation. This is Tony Ulwick’s stuff. To me context and outcome go together, as pairs, as sets of data. I focus on the the progress people are trying to make, as opposed to trying to make the best product in the world. People might say they want this great outcome, but the fact is sometimes you just have to give them a half step to make the progress. As David Heinemeier Hansson said, “You’re better off with a kick-ass half than a half-assed whole.”
Adding quantitative data to the JTBD mix
Des: As communities and theories expand, there can be a range of misunderstandings and misapplications. There’s also positive progress as more people get used to using it. What have the past few years been like for Jobs-to-be-Done?
Bob: A couple things have happened. We’ve gotten some tighter processes. For example, we now we have a very formalized JTBD interview debrief process to extract the data. Each interview is about a terabyte of data, and we pull out a range of things: pushes, pulls, anxiety, habits, hiring, firing, trade-offs. We put these into a piece of software that enables us to codify and see the patterns through math. It allows us to move from a 60-day process of doing all the interviews and analyzing them to a one-week process. This also allows fairly large companies to actually build their own training programs and scale them.
Des: On the math, what’s the process for quantifying a statement you hear in an interview?
Bob: There’s the qualitative side of how we actually find the jobs, and then there’s quantitative side of which jobs are happening when and where. You can actually start to codify a big data set by the theories you have. There’s a really cool way to take a current data set and lay the jobs theory over it. For example, one of the jobs when we were working with Basecamp was “Help me think it through.” You see people putting a lot of tasks in and inviting a lot of people to collaborate, lots of conversations, but never checking anything off. They were trying to make sure they were thinking about the project correctly, but then they wouldn’t actually do anything else after that. The product can do a lot more, but the reality is that’s what they use the tool for and that’s what helps them make progress. Now you can see it in the data.
Des: So the job provides a lens, which then can reasonably approximate the data you should see. One example is that Intercom is more popular on a Monday. Why is that? Well, you have a backlog with customer support, but also our marketing products are more popular on a Monday because people are checking reports to fill in their spreadsheets. We can see that in the behavior.
Bob: The trick is manifesting that back into the data. Another example on the quantitative side is a book I have coming out with Michael Horn about choosing a college. It’s about the struggling moment when an 18-year-old and their parents figure out which school they should go to, and there’s often a mismatch between the jobs the students want and what the institutions are actually delivering.
We did a quantitative study where we took all the pushes that people had talked about, and we asked, “What caused you to say, ‘Today’s the day I want to go to college?’” They’re not your usual set of answers. We used a MaxDiff analysis of what had the most influence versus the least influence. Then we asked, “What are the two or three things you’re actually hoping for when you go through college?” We had the same result, and out of it you can actually start to see the jobs.
We looked at 1,500 students during their first year as freshmen at University of Maryland, Baltimore. If you asked the administrators what they thought was going on, they thought everybody was there for what we called “the ideal college experience”: help me find out who I am, help me figure out a major. It turns out that the majority of people coming in were thinking, “I have a little extra time, I want to learn something new, and I want to better myself.” All of a sudden, you have very different sets of orientations.
Push vs pull
Des: That’s the classic ‘person’s not buying what the company thinks it’s selling’ scenario. Do you see the reverse, where people think they’re applying Jobs, but they’re not actually applying it? I constantly see software folks, people building startups, struggling to know where to stop. They do their JTBD interviews and they learn that Johnny wants marketing data. Why does he want it? Maybe he wants to be a good employee, so he can get promoted so he can earn more money. Then when the startup writes up the job, they say, “What does our product do? It does everything: It gets people promoted, it helps them have a happy marriage, it even exports to Microsoft Excel!” All of those things are true at different zoom levels, but they don’t know which one to anchor on. Any guidance there?
It’s the struggling moment where they can’t do something that causes them to take the leap.
Bob: I always think about it as a level of abstraction. What is the underlying energy that’s actually motivating them to do something? Nine times out of 10, it’s as much about push as it is about pull. It’s the struggling moment where they can’t do something that causes them to take the leap. This is the difference between what I call supply-side innovation, “I have this technology, what job can it do?” versus demand-side innovation: “What progress do customers wanna make, and how do they value it?”
Where does the product’s responsibility give way to their responsibility to follow through? That’s a hard point to find, and there are clearly points where people are extrapolating way beyond the consumers’ scope, focus or attention. You can try to extrapolate that, but it’s going to scare them because they haven’t thought about it. To me it’s really about what people can actually talk about or articulate – or what’s just over the edge of what they can articulate and have no language for. Making the leap of, “I’m going to hire Intercom and get my promotion…”, you can wish there were a direct connection, but there’s not. It’s the results Intercom gets you that will get you promoted.
Des: Some background here: Intercom started in 2011 as a single product. All we wanted to do at the start was to make internet business personal, but what we were building specifically was a way to make it really easy for people who run internet businesses to talk to their customers, and vice versa.
Fast-forward a couple of years, and what we had was a monolithic Intercom with a single piece of software that could do a lot of things. I first contacted Bob in 2013, because we didn’t know what to say on our marketing site. We didn’t know how best to articulate all of the weird ways Intercom was being used. So we spoke to 15 customers and came to the conclusion that people were using Intercom in four different ways, which we called Observe, Learn, Support and Engage. The customers were saying four things:
- “Help me see my users”
- “Help my users talk to me”
- “Help me provide timely responses”
- “Help me to get people to do the right thing at the right time.”
Next, we produced these job boards to split the product up notionally. We ran multiple marketing sites for each of the things we did, and we priced each of them differently. The whole project was a massive success.
Bob: But at this point you’re not actually marketing the jobs so as much as you are presenting them like use cases, right?
Des: The use cases still tie to the jobs. You’ll see that we have use cases called Support and Retain, Capture and Convert and Onboard and Engage. All of that research still holds true, because one of the core ideas about a Job-to-be-Done is that it should be timeless. It should represent a persistent need, not something that’s temporal. That has certainly been true for us. As to how we package, we still find ourselves in a situation where significant parts of our product are used for different purposes.
As an example, you can use our Inbox product to talk to sales customers – in which case, all you care about is your Salesforce integration, fast replies, lead qualification, etc. However, the whole right-hand panel of that same inbox will turn into a customer support column, which shows: “Here’s Bob. He has seven projects open. He’s struggling on project number eight. He’s paying us $49/month.” The same technology powers that, but it’s used in a different way.
If you want Capture and Convert, we will sell you our Messages product, so you can start conversations with visitors, and our Inbox product, so you can actually reply to those conversations and manage them at scale. In the case of Support and Retain, we’ll sell you a knowledge base, and an inbox. So that’s the kind of shift that we’ve made: we market solutions, but we sell products.
Bob: To me, jobs are use-case sets: there may be five or six use cases for the product, but the jobs are universal, like you said. Sometimes people can talk about the problem, or sometimes they actually walk in with the solution. But it might be the wrong solution, and you can help them understand why they might need two things instead of one (or a different set). Eventually, they have to be pulled into the job context so they can actually say, “That’s me. I want that. I’ll come.”
Understanding how people buy
Des: When I was working with the marketing team at Intercom, one of the big a-ha moments for me was seeing the many ways people come looking to us for what we would call our Capture and Convert solution. Some people say that they’re looking to increase website conversion, and that’s something that our product does. Other people will say that they need to replace their existing live-chat tool, so they’re searching for an Olark alternative, and we need to be able to put our hand up for that as well.
A customer might be looking for live chat for reasons like, “I have a problem,” meaning that they want to increase conversion. “I have a broken competitor,” when they mean that they’re not happy with my current offering. A third one might be, “I know the type of solution I need, and I’m here shopping.” Sometimes you have to teach the customer and draw a line for them between your product and their problem. Other times it’s as easy as them saying: “I know you guys do live chat. Where’s the buy button?” You need to be able to cater for all of these with one marketing site.
Bob: But the underlying premise is that the progress all of them are trying to make is the same, even though they’re using different language. This is where personas come in. I might have somebody really sophisticated, who knows what they’re doing and knows the vocabulary. They’ll say, “I want to increase conversions.” Someone else might say: “I’m new to this whole thing. I just want live chat. I don’t even think about conversions.”
People don’t buy jobs, they buy products.
The job is that foundational piece. I put personas and demographics underneath it. I may have to use different ways to capture them, but underlying causality is the same. In some cases, they don’t have to know about the job, but in other cases they might be really explicit and know everything about the job. That doesn’t mean you turn a job into the marketing; you have to be able to look through the job and through different persons to discover how to market better. People don’t buy jobs, they buy products.
Des: That’s a key insight: you have to understand your buyer personas, which might be of different levels of maturity, experience or naivety. You need to know who they are. Your job is a fixed thing, then your marketing site is a lens through which all personas should be able to see the solution you’re offering in their own language or with their own experiences.
Bob: Or in a range of languages – you might have different landing pages for the personas. But the fact is, whether they say “help me convert more people on my site” or “I need a chat tool”, they have the same intent.
Des: If you’re a startup founder, what’s a single step you can take with Jobs?
Bob: The greatest single step you can make is to actually talk to somebody who recently purchased you, and talk to somebody who recently quit you – or quit the competitor that you’re going after. By understanding these switching moments, you’re pulling a thread. And then once you’ve seen it, you can’t unsee it; you’ll see it over and over again.
The first step is always a set of interviews. I’m not talking about surveys. Literally get them on the phone and ask the basic question: why was today the day they signed up for this product? The thing you have to realize is that it’s not random, and you have to dig as hard as you can past the bullshit stuff they’ll tell you upfront. There’s always something deeper, because nobody really wants to switch. Habit is the strongest force of all, and people will just keep doing what they’re doing unless something gets in the way or something better comes a long. There has to be enough energy for them to stop something and start something.
Just go talk to your customers. That’s where this all began.
Des: It’s like detective work. I’ve seen you do it, when you say to someone, “You’re saying ‘cool.’ Unbundle that for me. What does this word ‘cool’ mean to you? Oh, it means ‘useful.’”
Bob: And then you ask, “What does useful mean?” You get them to tell you what’s not useful because sometimes they can articulate better in negative space than in positive space. It’s about playing with the words. The English language kind of sucks, because the same word can mean five different things – or five different words can mean the same thing. I often have to say, “I have no idea which meaning you’re using, so let’s just take it down a notch.” I can’t design product without fully understanding this first.