Applying the JTBD Framework to Build User Personas that Stick

Marine Palamutyan, Megan Mathews at UXDX USA. Video: https://youtu.be/ZnU_n-FDxio

Readable transcript: edited from the recording's captions for readability (fillers and false starts removed, punctuation and section headings added). Wording is the speaker's own. Timestamps are positions in the video. Names marked [?] could not be verified against the audio.

A warm-up: picture a dog

[00:00:08] Host: Hi. They're from DocuSign, if you can't tell.

[00:00:14] Hello everyone. It's great to be here with you all today, and I thought we'd get started with just a little bit of exercise. A little warm up. Okay, not quite that kind of exercise. All right, if you will picture in your mind a dog. Can you see it in your mind's eye? Close your eyes if you need to. Okay, next I want you to picture a street where you're walking that dog. Got it?

[00:00:44] Okay, let's play a little game. How many people out there imagined a labrador? Let's hear some noise. Where are my lab people? I see you out there. What about my poodle people? Oh, thank you, poodle people. How many of you pictured your own dogs? Oh, okay, crowd's going crazy here. Okay, let's talk about the street. How many of you imagined a New York City street, like the ones outdoors? Okay, seeing a lot of hands go up here. What about a nice Midwestern suburb, quiet? Yep. Maybe something by the water, an ocean or beach? Oh, I've got a few there. How many of you pictured yourselves walking your own dogs on your own street? Okay, yes.

[00:01:29] I love this example because it illustrates how we're all different. We all come from our own experiences and we have our own mindsets, and those mindsets shape our behaviors and our decision making, which also carry over to our day-to-day job. And product development is no different. Many of us here are part of a product development team in some capacity, and when we think of our customers we likely have a different understanding of them and their needs. Without a shared understanding of our customers, we will build and design products from our own experiences and fragmented views, which is why companies set out to establish a shared language through personas.

Sixty personas, and a game of whack-a-mole

[00:02:08] But what if you end up with too many personas? Uh-oh. Time to air some dirty laundry. DocuSign had over 60 personas. I said over 60 personas circulating about. How did we get here? Well, the truth is it didn't happen overnight. In fact, DocuSign turned 20 years old this year, and as our company grew, and we grew to over a billion users, so grew our number of personas. We had personas by segment, we had them by industry, we had them by job titles. You name it, we probably had it. And despite having over 60 personas, there was this mentality circulating around the teams that we needed more personas. It was like a never-ending game of persona whack-a-mole. There's another, there's another.

[00:03:11] But what if there are 60 personas, what's the big problem really? Well, for us at DocuSign we really had an unmanageable set of personas at a certain point, where our assets started to become a liability for us, because they were driving inconsistent narratives, creating fragmented points of view. And also teams were becoming more and more siloed. That siloed thinking translated into the types of vision projects that were underway, and we were developing and building products that just didn't scale. We also had a lot of duplicated effort, five ways to maybe do a thing. And most importantly, our customers really bore the brunt of this cost, with a fragmented user experience. We were in a tough spot.

[00:04:07] So what the heck do we do at over 60 personas? How do we make them usable? How do we make them actionable? Which at that point begged the question, how many personas do we really need to get to a good understanding of our core users? Do we need 5, 10, 20, more? Well, at that point we made a bet. This thing's not working. We made a bet that we can accomplish this with less than 10 personas, to take control of the customer narrative and help teams collectively focus on the customer needs rather than product outputs.

Two competing frameworks

[00:04:41] But wait, there's actually more. We had two competing frameworks that had gained momentum at the company. For those who may not be familiar with jobs to be done, it's a customer-centric approach to product and service development that essentially helps us understand what are the customers trying to do in a specific context, so that we can build products to help them in that situation. In DocuSign's case it's in the context of agreements. So we had a team called jobs to be done innovators, which I was a part of, that built customer job maps that captured all the pain points and needs along the agreement journey. But it was too complex and actually intimidating. In fact, it was so massive we kind of broke Miro. It takes forever to load still. And it was also criticized for lacking the human element. They're like, well, who's doing this? It was too focused on the job rather than the people.

[00:05:40] Now, running in parallel we had another framework called the Persona Guild, which I was a part of. The guild was really focused on creating and establishing a template for personas for product development in particular, so that we could identify, well, what are the actual components that we need to take action and make it memorable and build that empathy. So the Persona Guild did an awesome job at identifying those contextual elements that are so needed, and it served as an excellent guide for future development of personas. But it still didn't solve our problem. There was still going to be more personas created. So neither of these efforts in and of themselves were able to create a unifying story and to solve our issue.

[00:06:28] And then we had an aha moment. What if we could merge these things together? It sounds so great. You've got jobs to be done, our functional components, we've got our personas bringing it to life, the contextual components. If we can merge these, we could actually have a 360 view of our customer. Sounds great, but wait, how do we do that, and do it in less than 10 personas?

[00:06:52] Well, we needed more brains and perspectives. So we advocated for dedicated time and resources to accelerate the progress of merging these two bodies of work, so that we could create a core set of personas that would be the ones to rule them all. And then we can socialize them far and wide to help with establishing that shared language, and develop a curriculum that would help teams and enable them to start using these new personas. After all, it's not good to build personas if no one's going to use them, right?

[00:07:23] We also knew that in order to be successful, nothing about this process could exist in a vacuum. So we engaged leadership at each step along the way and gained that buy-in. We also established a solutions team that was cross-functional, made up of researchers, designers, subject matter experts, content writers, and they were really targeted and honed in on, in a holistic and balanced way, creating a new solution for us. So we've got our mission, we've got our resources, our people are stacked and ready to go. It's go time, baby. Pressure was on. I'm not going to lie, I was sweating a little bit. I think we all were, quite a bit.

Starting with jobs, and a hierarchy of jobs

[00:08:03] So the first thing we did was we started with jobs. We realized that we can really lean into the jobs to be done framework to structure this new merged framework around, because what jobs to be done offered us was an end-to-end view of the agreement, core jobs that were agnostic of segment, company size and product. So it allowed us to talk from a very customer-centric point of view and hone into the key moments that matter to our customers. But out of the box it was not perfect.

[00:08:39] To simplify it, or make it work for us, we actually adapted the jobs to be done framework and created this hierarchy of jobs, illustrated in this pyramid. And to help with having that shared language, we locked in the language on the top two layers, the journey phases and the core jobs, and then let teams actually uncover micro jobs and user tasks as they continue to look through nuanced behaviors and interactions.

[00:09:07] Now that we had the core jobs, we could anchor our synthesis of the personas to core jobs. If you've watched It's Always Sunny in Philadelphia, the mess of papers behind Charlie is exactly how we felt trying to take inventory of all the personas, remember, the 60 we mentioned, and then trying to merge it with our jobs to be done framework. We searched a lot for industry examples and playbooks where a company successfully merged these two bodies of work, but there was a glaring gap. So the pressure was on to innovate our own and make our own that would work for the company. A lot of long nights in there, a lot of discussions, possibly some drinks or five. We're not talking about that.

Synthesizing 60 personas down

[00:09:52] But let's talk a little bit about what we did with the synthesis. With over 60 personas we had this sea of data that we really didn't want to lose. So the first thing we did is we evaluated each legacy persona and we binned them by the core job that they accomplished. Now, what we're trying to look for is not every pain point. We're not trying to detail every goal, every motivation, because if we did that, instead of a one pager, which is what we're aiming for, we would have more of a persona book that no one would want to touch. So instead we synthesized just for the commonalities. What are the key trends that we can extract? So again, we synthesized pain points, goals and motivations across those clustered sets. We won't be able to actually share our personas with you, but we're going to share with you a blueprint that you all can take to your own companies if you choose to take on this challenge.

[00:10:46] But we weren't done yet. The next thing that we had to solve for was how to go from these clustered jobs to be done groupings to actually a persona that folks could use. And you might wonder, well, why not just turn the work that you did for the core jobs right into the personas? Easy. Well, if you're a smaller organization, or if you have a more narrow job map, something like that could really work for you. At DocuSign it wasn't quite so simple, for a few reasons. One is that we had multiple job maps that our personas needed to account for. Secondly, if we took a step back and looked at each core job, not all of them were weighted the same in terms of the importance to our business. So we had to get creative and innovate again.

[00:11:33] We actually grouped some of the jobs so that we had them linked to a single persona, and that way, no matter what step the customer was at, we had a persona linked to it. You can see in this example, this will be available, but feel free to take a screenshot if you want to use this at your company as well.

Personas based on jobs, not roles

[00:11:54] So now that we've shared some of the process and the blueprint, one thing we wanted to really hone in on and be the key takeaway, if nothing else from this talk, is this: our new personas are based on the jobs that people do, not the roles that people have. What does that mean? Well, these personas are designed to be product agnostic, segment agnostic, industry agnostic. So you're no longer going to find your persona by sales, or persona by legal teams. And as a user you don't just wear one persona hat. You actually put on the hat of the persona that you need, based on the job that you need to do in a moment. This critical shift in our framework, away from traditional personas to a jobs to be done informed persona, has allowed us to create a shared narrative across a billion of our users.

[00:12:53] To illustrate how this works, let's use an example we can all relate to. Families, right? As a part of a family you might take on many jobs, like being a parent, or a financial planner, or caregiver, or a cleaner, because somebody's got to do those dishes, right? But every family is structured differently. So if you're part of a smaller family then you would have multiple of those responsibilities, but as the family grows then you may distribute those roles. So then somebody might be dedicated to doing the financial planning of the family, other ones might be just on caregiving duty and pick up and drop off if you have a kid.

[00:13:32] So to accommodate for specific nuances like geography, industry, company size, we created what we call persona modifiers. You start with the base persona, then you layer on the modifiers to see if any of the pain points or motivations actually change. That way it reduces the need to create additional personas, and keeps history from repeating itself, because we didn't want to have more.

Three work streams: design, content and education

[00:14:01] With the hard part of figuring out this merged framework done, now we wanted to give you a backstage view of how we brought this work across the finish line in a two-week sprint. As many of you might know, a common risk with personas is that you put in all this work, you create them, and then nobody uses them and nobody adopts them. They sit on a pretty shelf and nobody touches it. So to mitigate that risk we actually used highly visual assets, gaming, and really got creative in how we developed and rolled out the personas so that it resonates with teams.

[00:14:38] So we actually broke into three work streams. We had design, content and education, to tackle different parts of this. For example, the design team worked on things like the layout of the persona cards, or the iconography that would be inclusive and a reusable component. So essentially we built our own mini design system for the persona work.

[00:15:02] Our content work stream was made up of researchers and content writers and subject matter experts, and they were really the crux of collecting feedback and synthesizing everything that was all the text that was shown on the personas. They helped us identify what to call the personas, and they also did an incredibly sweeping pass to make sure all of our language was unified and it was in easily understandable language and tone, so that anyone that's reading these personas could understand it. None of that crazy jargon.

[00:15:36] And our final work stream, my personal favorite, is our education work stream. This is where you go from theory into practice. We knew, like Marine said, that we couldn't just do a big drop, do a splash, and expect folks to walk away as experts. So we put together a whole rollout plan, and we put education right at the center of that rollout plan.

[00:16:00] So we asked ourselves this one question: what does it look like when personas are being effectively used on the ground? What changes about their behavior? What decisions? How does it look because they have them versus they don't have them? And the answer to that question led us to create a whole curriculum around adoption. We leaned heavily into learning theory and gamification techniques so that we could really make this enjoyable and fun and lower the bar to entry. Because after all, it's a whole new framework that we have to teach people, and it's a new way of thinking about personas that takes time to build knowledge on top of each other. And it's fair to say that this part of the project actually went from two weeks to about a six month time period.

The impact, a year in

[00:16:51] And now we're almost a year in, and what a great impact we've seen so far. I wanted to highlight a few of them for you today. One of the greatest impacts of this work has been establishing that shared language, which was one of our biggest problems when we started this project. So now teams are actually referencing personas for product demos and roadmaps and PRDs. We're all speaking that unified language about our customers, and it makes it easier to identify overlaps and common needs, and encourages teams to actually collaborate and talk to each other when they realize they're building for the same persona and the same user.

[00:17:34] And because our personas were tied directly to our jobs to be done, our core jobs, it acted as a gateway and invitation to invite teams to dig deeper into the nuanced customer behavior and really get specific about what are the pain points we're really trying to solve for. And because we had this whole end-to-end view of our customer experience, we could rally behind those specific moments. This type of scenario-based innovation strategy has begun to take off, and the core personas were a really crucial part of that, because it helped teams to anchor to those pain points. We started to see a shift away from this product-centric roadmap development work that's happening, to more of a customer-based roadmap process. And on top of that, it's helped us to consolidate effort, and we're also saving ourselves time from having a lot of duplicative work.

[00:18:34] And having a core set of user personas in a shared language, it actually enhanced how we engage and communicate with customers. So now we do recruiting of research participants leveraging our personas, we're building persona based customer panels, we're organizing our support content by that, and also curating an agenda and content for our annual conference for our customers. In addition, we started exploring how we might personalize some of these experiences based on personas.

[00:19:07] It's been incredible to watch these new jobs to be done informed personas take off at DocuSign, and actually see the narrative shift to being more customer-centric. Many teams, including our president of product and technology and the CEO, are actually aware of our user personas. So it's an exciting time. Every day is a new day with new things coming in.

Reflections on the journey

[00:19:31] But this has been a massive effort and wasn't without a lot of trial and error, so we thought we'd share some of our reflections on our journey, in hopes that it can be helpful for those of you that might be experiencing similar issues with personas.

[00:19:45] So we started in persona chaos, mind you, and we were really at a point where having so many personas became more of a liability than an asset for us. And so we made a big bet that by streamlining the narrative it would help, and that's exactly what we did. It's been incredible to be able to take back the narrative, so to speak, for our customers.

[00:20:10] And when frameworks didn't work out of the box for us, we got creative. So we want to encourage that if you find yourself in a situation when you're trying to choose a framework, don't get too attached to a framework, because they don't always work out of the box. Feel free to be creative and adapt it to make it work for your organization and your teams.

[00:20:33] And remember, becoming an expert at anything is really hard to do. And so we leveraged gamification techniques and play as a vehicle to learn and also to build that memory. It's been a lot of fun.

[00:20:52] Finally, treat personas as living and breathing assets. They're never really done. Launch a version of it even if it's not perfect, then encourage teams to start using it and experiment with it, so then you can make iterations and be ready to change and make the improvements so that they're more effective and actionable.

[00:21:12] And our last reflection. You remember this slide from earlier. We just want you to leave with this last piece, which is: traditional personas are really anchored in segments and industry, the ones that we're all familiar with, maybe coming more from the marketing space. And we think about product personas. The issue with that is, as maybe the industry shifts and you have environmental factors that you don't control, or as your company pivots and is changing directions or targeting a new segment audience, so changes the needs for your personas. What jobs to be done offered us really gets to the heart of why people use services and products, and that served as an awesome foundation to build our framework on top of, because now we can talk in a more holistic way about what is the core need of our users, and keep us focused on what matters the most, which is our customers.

[00:22:12] And it takes a village to do work like this, so I want to give some recognition to all of the many people who were involved in bringing these personas to life, some of whom are in the crowd today as well. So thank you to all of those fantastic people. And with that, thanks so much for attending our talk. We hope you found this informative and inspiring wherever you may be in your persona journey, and we'll take questions now.

Q&A

[00:22:44] Host: Thank you very much. That was really amazing, because I love those talks that are, how I perceive it is, you're treating your research as a product. So instead of just building it and hoping people will come, it's no, we have to treat it as a product. What's the onboarding? What's the usage like? Let's iterate and improve. So I just love that example, so thank you very much for sharing. We have some time for some questions, so is there anything online? I'm just going to jump to the online first. Yes, we do have questions online. Our first question is from Sylvia Kiter. What kinds of pushback did you encounter in consolidating personas, and how did you handle it?

[00:23:26] Speaker: So many. I think the biggest question, so we had to multiply the redundancy as much as we could. We had a multi-stage rollout plan, and the very first stage was just awareness. We just wanted to make people be aware of the new framework and know where they can go to learn more. And as we rolled out that first stage, the first question that came out from folks is, I'm working on an enterprise product, what about the persona for my enterprise people? Or, I work on this specific product line, what about personas for my product line?

[00:24:02] Speaker: And so hence the development of our modifiers concept, which really helped us to shift the narrative, keep it focused on the high level. Because what we learned is that if you look across these customers and what their needs are within a particular moment, they actually have a lot more in common than they have differences. Differences do matter, and in particular moments it's really important to specify what those are and build around that. But by and large a majority is the same. So that was one of the big concepts that we worked on and rolled out over time.

[00:24:39] Host: Great. So folks, some hands are up. We'll start here.

[00:24:51] Audience: I was just wondering how you named the assets. Is it by person or by the job to be done? How do you refer to each one?

[00:25:03] Speaker: I think it's more around trying to represent the job they're doing rather than the individual. Because one of the things we wanted to do is help reduce bias and not use names or demographic kind of aspects represented in there, just to keep the focus on the job that they're actually trying to perform, and how our tools might help with that. So then the customer is hiring our product for whatever they need to solve for.

[00:25:37] Host: Great, next question.

[00:25:38] Audience: Hi, my name is Kayla, and as an educator it was quite refreshing to hear the way that you employ teaching inside of your practice. So how did you lean into that role, and what benefits did you gain out of it?

[00:25:53] Speaker: Well, this started actually in the Persona Guild body of work. We wanted to create a rubric to help us think more critically about, well, what does it look like when they actually go from a newbie, a novice, to someone who's an expert? And so we leveraged some backwards learning theory to really design and craft rubrics that can help us be specific about the types of behavior we want to see on the ground. And how we got that information was by talking to our product team. So we actually did focus groups and we were like, tell us what it looks like when it's working well. So those were some of the things that we did to build out our education program. We really took a customer-centric approach at first.

[00:26:38] Host: Yes, we've got another question over here.

[00:26:40] Audience: Hi, I'm Kristoff. I have a question. I liked that you want to keep the personas updated. How do you qualify the feedback? Because I can imagine now every team is coming back and it's like, ah, but in our case it's like that, in our case it's like that. How do you qualify, and what's getting back into the persona and what's not getting back into the persona?

[00:27:04] Speaker: Yes. Well, we actually have a website for our personas internally, and we also have feedback forms, Slack channels. So we're constantly taking feedback. And not to mention, part of the Slack channel is, as folks start to use them and actually put them to practice, to share what's working well, what's not working well. So all of those things are getting considered. And as Marine mentioned, we're almost a year in. So some of the next things that we're doing is we're putting together a committee to help us evaluate which changes from a global level will emerge and make updates to the personas. But that's a really great question, and all I can say is it's a process, we're kind of in the middle of it.

[00:27:51] Host: Do we have any other questions? We'll take this question here and then we'll try to get it over to another side of the room.

[00:27:59] Audience: Hi. So I'm at a company where we talk a lot about jobs to be done, and the thing I think no one can ever agree upon is how big is a job. Is a job a tiny task, or is it like a huge overarching goal? I know you showed that pyramid, that you sort of rank these things, and I'm curious how you differentiated between those levels and what actually counted as a job.

[00:28:26] Speaker: I'll take that. We get that question a lot. I think at the higher level, where you had the pyramid, it's thinking more about big chunks of actions that are happening. So if you're baking something, for example, you have to decide what you want to bake, then you have to go grocery shopping, and then you actually bake. But once you get to the more granular level, it's like, okay, I've got the ingredients, I've got to put in one teaspoon of sugar, and that gets more to the micro job and user task level. So think of the more high level overview, those are the journey phases, which we added, or the core jobs, and then the more granular is those steps.

[00:29:10] Speaker: And again, if you look at it out of the box, it's very prescriptive. So figure out what works for your company and what resonates, and feel free to adapt it, so you don't have to stick with going by the book.

[00:29:25] Host: Okay, so this is an interesting question from Lindsay Defore. It seems like there's quite an investment into creating these personas, maintaining them and rolling them out. I gathered that leadership and cross-functional teams were supportive of this effort from the start. Is that true, and can you speak a bit more about that?

[00:29:45] Speaker: Yes, that is a true assessment of the situation. One thing that worked to our advantage was this new wave of leadership that came through that are very, very customer-centric, and care about our customer experiences, and striving to deliver delightful and 11 star experiences. So they were really supportive too, and they knew this new work needed to be done, especially once we took inventory and we're like, okay, we have 60 personas, what are we actually going to do with this thing? So yeah, a lot of support from leadership for sure.

[00:30:20] Host: Good. I think we have time for another one or two questions. There's people in the center, just in the right here. I can see the mic is running around towards you. Oh, okay, over there, we'll come to you next.

[00:30:34] Audience: I am Jason.

[00:30:36] Host: Hi Jason.

[00:30:37] Audience: So one question I had was, how did you differentiate between a job map and a user journey map, and was there one that was prioritized more over the other?

[00:30:50] Speaker: Yeah, I think with the journey map, again, it gets to more of a process map almost, and it's very product focused. It's capturing how the customer interacts with your product. Versus the job map is what's happening in the context they're in. So again, in our case it's agreements, like what are the core jobs they're trying to accomplish in an agreement context, and focusing on those things. So I think the big difference here is one is more product focused and the other one is more job focused and customer focused.

[00:31:22] Host: And last question.

[00:31:26] Audience: Thank you so much. So I'm Lindsay, I do product research at Priceline. Awesome talk, thank you so much. One of the questions I think that's occurring to me, and I get this a lot, is people on the product team really try to understand how to use these personas in whatever shape they are and with the modifiers. Coming from 60 plus personas and modifying it this way, I imagine there was a certain way they were using the 60 plus, or not using, as the case may be. As you sort of made this journey into a new format, I'm just curious if you could speak to the education around, here's how you can now use this and make them actionable.

[00:32:02] Speaker: Yeah, we had a lot of strategies. We actually had three tiers as part of our education. On the bottom, I mentioned before, was just awareness and knowing where to go, and having that fluency starting from leadership all the way down around, okay, here are the new things and what they mean, and just to be able to talk about them.

[00:32:20] Speaker: But we also realized that everyone jumps really quickly into wanting to apply them, and we were a little cautious about that, because they hadn't actually learned what the actual personas were and how to contrast and compare and be able to identify which one should their teams be using. So we actually slowed down the roll on the application piece. We didn't give them the tools to apply right out of the gate. We sort of created a forcing function by rolling out these educational games. We did games to help them name the persona and compare aspects of the persona to each other, so they had that next tier of understanding and could actually be able to say individually, as a designer or researcher or product manager or an engineer, oh, it's this persona, and I know why it's this persona. And they could, almost like you remember a friend, recall some of the details. Once they got to that level, then we were ready to level them up, and we had all kinds of tools and additional games to get them to the application level as well. So we slowed the roll down to speed the roll up, so to speak.

[00:33:24] Speaker: And feel free to find us. If you have additional questions we'll be around.

[00:33:29] Host: Thank you. So thank you once again, Megan and Marine.