Episode #222 Dan Klyn - Determining What Good Means with Performance Continuums Live!
In considering your user's experience with your design, keep in mind that there's a difference between something looking good and being good. But how do you determine good? How can you measure it? If, for example, you’re a print company building a digital presence, do you focus on retention or acquisition based on the shifting experience? It’s easy to ask for something that already exists, but much harder to describe something that one might want or need.
In considering your user's experience with your design, keep in mind that there's a difference between something looking good and being good. But how do you determine good? How can you measure it? If, for example, you’re a print company building a digital presence, do you focus on retention or acquisition based on the shifting experience?
Dan Klyn of The Understanding Group believes that it all starts with language. It’s easy to ask for something that already exists, but much harder to describe something that one might want or need. Inspired by the work of Richard Saul Wurman, Dan introduced a diagnostic tool to his team based on the idea of performance continuums.
In using these continuums, you can remove “or” from the conversation. Instead of looking at design problems as a contradictory issue, you place your focus where it is appropriate on the continuum. In the example of the print company, Dan offers “We must continue to acquire new customers, yet we will service the ones that we have really well”. By seeing where you land on the continuum, you can get a sense if what you’re currently doing fits in with “good”.
Dan shares his thoughts in this podcast of his presentation from UX Thursday Detroit.
Dan Klyn: "Publishing Company" was accustomed to this. Paper and print and periodicals. What they were up against is digital.
They had adopted digital a long, long time ago to make printing paper things more efficient, easier, better and that sort of thing. They did it in a timely fashion. They had no shortage of digital technology to make the paper happen.
They didn't embrace the medium of digital as a publishing medium and they needed to catch up. They felt like they were being forced, this is 2011, they're being forced to transition to digital as a publishing medium.
They hired a UX consulting company, a famous one from one of the coasts who we won't name, to build them this bridge from print world, the print medium to the digital medium. Those people did a pretty OK job.
They had a launch of this new digital version of what had been a print product. This print product had started being sort of, became a thing in the 1840s. They had a very long tradition of doing their business in print.
They cut the ribbon on their new digital thing, and it was OK. It was OK. But what they've found out, there was a mix of people working on this thing and they had a blend of traditional publishing people and digital people.
The digital people knew how to measure the use of the new thing. What they found was disquieting, to say the least. What they found was that when people first came to this new digital thing, this digital version of something they were accustomed to in print, they liked it pretty OK the first time. But the more that they used it, the less they liked it. The less engaged they got.
That's a big problem when your whole business model, at least the print version of the business model, is based on continuous repeated use.
What they had on their hands was kind of like a turkey. At Thanksgiving, we all dream of eating that turkey. When we sit down after a year of being away from that turkey dinner, it's great. But then, five, six days later, when you're having your 16th turkey sandwich, you don't like the turkey anymore.
One of the people who worked at this company, we'll call her Doris, was concerned about this pattern. She knew it was at odds with what the business needed to do to succeed. She was lying in bed at night with her husband, we will call him Rock, and she was sharing her concerns. Like, our company, this isn't, we cut the cord, the people came, it was great, and now it's not. It's kind of a turkey. It's not doing good.
What Rock said to Doris was, "Have you guys tried information architecture as part of this making of this thing?" She said, "What is that?"
The people who built them their bridge from print to digital were an end-to-end UX provider who started with visual pictures of what the new digital thing would look like. When they figured out a picture of it that everybody liked, or thought they liked, then that same company built it and launched it.
So, laying in bed, worrying about this thing, her husband introduces this idea, the information architecture. Have you done that? Is that maybe what's missing in your problem? He said he knew a guy and that happened to be me.
We had this company, this new company called TUG, The Understanding Group. As I'm talking with my acquaintance's wife, Doris, about what they need, she was becoming increasingly excited about this idea of what we were offering, which wasn't a whole new big process, it was, let's figure out what you should do.
You have a thing, it performs OK, we know there are some problems. What is the plan for what you ought to do next? We're not going to talk about what it's going to look like and we're not going to talk about how we're going to build it. Let's just build a road map. Let's build a set of plans and that's a great idea.
As information architects, we'd like to invoke the metaphors and ideas of architecture. That picture of, wow, I could be working with the architect and we would have this great plan. We would figure out the plan before we do the things.
That sounded pretty good to her and to us, as well. We engaged for a project to build them a road map for what they ought to do.
What we learned was, there was a cast of thousands on the stakeholder team. We had been told that there were a lot of stakeholders and that it would be a pretty involved process. But there were 11 people who were deciders on this thing, and it wasn't 11 Dorises.
It was a weird mix of people from the old print part of the business, people who were new digital people who got brought in. Planning together the "we" of who is going to be planning was pretty complicated. It was this grand group with all these weird things that they were interested in.
Our plan was, because they had said to us...There's a HIPPO in here, there's a couple of HIPPOs. Do you guys know HIPPO? This might be a Jared thing, right? Did you make up HIPPO? Highest paid person in the organization? I learned it from him, at least.
In this group, there was the highest paid person in the organization, the second highest paid person in the organization, and then everybody else. They didn't want the bosses to just make a decision. What they wanted was a planning process that was inclusive and that channeled the collective intention and will of this group in order to make that plan for what we should do.
Our process for doing this was, well, we need to isolate them. We need to talk with them individually, figure out what each of them is about and what their priorities are and what they need.
That's Abby Covert and that's me and some of our other team members. Then we're sorting out the notes from all of these interviews. Our expectation is that we will be able to prioritize and figure out common threads and stack the paper and make the piles and we'll figure out what the plan is.
But that was a challenge. There were all these contradictions, seeming contradictions in what the people wanted. Some of the people wanted to focus on things like customer acquisition and if that was the problem.
This problem of losing engagement and interest as you use it more. Well, maybe we attracted, we acquired the wrong customers. If we acquire more digitally savvy customers, maybe that population will be better for our business. Other folks are like, no, we need to service the customers that we have.
As we work through these stakeholder interviews, all these seeming contradictions, we wanted to make a tidy little plan that everybody could agree on and have consensus around. But contradictions optimize what we've already built. They had built a ton of features.
Should the plan be optimize what we have and make the features we have work better? Would that help our engagement problem? Or should we keep innovating? Is there something missing from what we offer now, and ought we focus on that in our plan?
A huge one for them, based on the patterns that Jared talked with us about this morning, there's this, wow, they had a really poor mobile experience, all this pinching and twisting. They didn't have anything specifically for mobile.
Should our road map, should our plan be about what we're going to do for mobile now? Or should we just say screw it, we don't have a mobile anything right now? Rather than stopping the gap against the basic expectation for mobile, let's build something awesome for mobile that will be our future thing.
Then a really big one sort of, how are we going to measure success with this? Are we going to try to talk about building engagement? Because we started to look at the analytics, and while there was some problems with engagement, there was maybe some evidence that conversion was pretty good or could continue to be one of the ways that they could focus.
All these seeming contradictions. How do we sort this out? Because we have this perseverative interest in information architecture among our crew, we like to think in terms of architecture.
If we were building a physical space to accommodate all of these contradictory needs and goals, we might end up with something like this. In order to even make something like this, you would have to be able to change the laws of physics and gravity and stuff.
What are we going to do? What I usually do, so back to the storyline of this project, we had been engaged for a couple of weeks, they were expecting us to deliver them a roadmap for what, all the things that they should do, the plan. We had a meeting to show them the results of what we had learned that their intentions were in about a week or so.
We had been expecting to be further along in that process and to have an orderly way to talk about what the plan is going to look like, and we had nothing. We just had all these contradictions and complexities.
As I often do, rather than digging into the problem, I procrastinated and ran away from it.
In addition to all of this weird consulting work that I do, I do research into the life and work of a guy named Richard Saul Wurman. Anybody here familiar with the TED conference? The guy who invented the TED conference is an architect and I've been studying his work at the point in the story in 2011 for about 4 years.
Rather than dealing with this project problem, I turtled into my happy place of this kind of stuff, the work that Richard Saul Wurman is doing in architecture in the 1970s. La la la. [laughs]
This is great. But what I found was, I had this gift of synchronicity or coincidence or whatever. In this book from 1972, which is the year of my birth, so 41 years ago, Richard Saul Wurman writes a book. It's called, "The Nature of Recreation," and it's about Frederick Law Olmsted, the landscape architect. It's about how do you design and plan parks and recreation?
In the middle of this book is this concept that he introduced that he calls, "performance." What is so peculiar to me at this time trying to avoid my work problem by perseverating on what this guy was doing with architecture is his description of how you can make the planning of parks and recreation in the built environment be good isn't about bricks and mortar and sod, or...It's not about any of those things. It's about language.
His contention in this book is that the reason just what's possible for us to build starts with the language that we use to describe our intention of what we think is going to be good to build. What he observes is people know how to ask for a product of something that they've already seen, but they don't have a language to describe what they actually want or need. They can say, "It's kind of like Pinterest for dogs."
Another thing that Wurman is figuring out in the seventies is that there is a difference between things that look good and things that are good. He's saying, "Let's not talk about what any of this architecture is going to look like. Let's develop a language of performance to describe what would be good to make."
One of the things that he says in this book on page 19 is, "This language of performance that would not be talking about products, but talking about what would be good for us, that the way to pull this off is often with continuums." Rather than saying we have an inherent contradiction -- this park is either going to be for young people or old people -- he was saying, "No, let's put these on a continuum, because we have to do both, because that's how people are.
People are complex and contradictory, and just saying, "We're going to make something for the young or just for the old, that's not going to be good enough. If we want to make things to be really good, we have to find a way to describe these interesting contradictions.
And so, he proposed a language of performance using continuums. Some of the continuums he described for the nature of recreation, young and old use of that park. There's a continuum, and where on this continuum is good for what we're trying to accomplish? Or the land that we're building on, some of it is flat. Some of it is sloped.
Where are we going to focus? And it can be both. It doesn't have to be either/or. And so, thinking about what these people...These people's problem of all these seeming contractions, and are we going to have to say some of these are winners and some of these are losers? Maybe what we could do...so here's an example of one of those contractions. Are we going to have this plan be about acquisition or servicing?
What if we got rid of, "or," and what if we said, "yet?" We must continue to acquire new customers, yet we will service the ones that we have really well. If we put this on a continuum, we could say, "With all of these interesting challenges, what would be good for us to do? Where should we be at on these continuums?"
Rather than saying we need a this. It's kind of like a that with a little of a that thrown in, if we could describe our desire and our need and what would be good as far as we're concerned with this rich language of continuums, we could make some pretty great things.
We could also use it as a diagnostic tool so once we figure out what are these seeming contradictions that the client is wrestling with, we could say, "Today, what you present, the balance that you are making between these competing concerns is about a two towards the this end," and then either through an expert recommendation or through a consensus building effort, we could say, "Our target is to move a couple of clicks over toward that.
We're not making a pivot from, "either," to, "or." We're saying, "We will continue to address this, but we will tend more toward as we go forward toward catering a little bit more toward that," which is great.
For Publishing Co., we use this tool. I've been hiding in my research to not have to do my job, and here on a silver platter from 1972 is a way to negotiate these problems. Because the client didn't want the HiPPOs to be the ones who are just calling the shots, because they said they were interested in having it be a group decision, we anonymized their intentions.
We gave them worksheet to fill out, and we knew who they were. We put their names on the worksheets, but we anonymized the results. We knew that this was the boss, this was the second boss, this was most of the people on the team, and then, these were the developers, but they didn't know that.
We had thought that maybe what we would do is we would say what's green, what's the plan, is the statistical average of everybody's intention as mapped to this continuum. We did that, and what we found out was that doesn't work.
We ended up with a problem that Brian Eno has described really well. Brian Eno uses continuums. He calls the points on these continuums. He's interested in fashion and art, and so if he's talking about punk and mod, two fashion movements in England, that can be on a continuum, and that's how he would think about them is on continuums.
What he said somewhere -- I don't have the reference for it. Jorge Arong can get you the reference. He gave it to me -- is that the points on those continuums are cultural addresses. The cultural of the company is expressed at what point are they at on that continuum, and if you are proposing a move that that can be terrifying to people.
It was, and when we showed them without the color code sort of the uneven distribution of their intent, and we asked them to do a next step with us before road mapping to clarify that intent, they couldn't do it. We lost the rest of the engagement. They didn't want to move forward, because in order to reconcile themselves to a change in the cultural address of where they would be focused, that was beyond what they were capable of doing.
It reminds me a little bit of in, "The Tempest." Caliban is a creature that lives on this island with these shipwrecked English people. The first time that he sees a mirror, he freaks out, because it's the first time that he realizes he's different from everybody else. It's the first time he's seen the truth is looking in a mirror.
These continuums are a mirror of what's true about the organization, and that truth for this client was too vivid and they couldn't handle the truth, and so we lost the rest of the engagement. Goodbye Tug. Thank you for showing us how disordered and uneven our intention is, and we can't do anything with this.
That's the sad part of the story. The happy part of the story is that we have this tool that we now use with every engagement that we do in order to model the intention of what the people say that they want, in order to determine what good means. Not some sort of universal flat design is better than skeuomorphic design. But in the context of this company and their culture--what does good mean--we now have a tool to do this, which we love.
What does "good" mean? What do we need out of our work relationships with our clients? We need approval. They need to sign off. I don't know about this morally righteous stuff but what good means is a hard conversation to have.
What Mr. Werman does, rather than talking about some sort of universal good, he talks about choosing the right way to structure information to help people navigate through it with an aim of making something be good and that being good is not the same thing as looking good.
This tool of the performance continuum happens before any sort of visual description of what you might do is going to happen. They had been accustomed to being shown pictures and then iteratively changing the pictures to be more and more what they think they want. This tool can happen independently of any sort of modality. This is describing what good means.
I'm over time, and I have a few tips for you. The first tip would be to use these performance continuums. Back to so what does good mean? How many of you have had a project that you deliver, and then you get feedback from the client that uses the word "too"? "It's too flat." "It's too clean." "It's too much like Instagram." "It's too Twittery." You can use that future objection to what you might do to figure out these performance continuums.
They could say, "It's too this" or "It's too that." Rather than picking one of those, what if before we make the thing we could present that on a continuum and say, "In this range of what we might do, how would we describe what would be good?" Another tip for using these that we found effective is using grammatical symmetry on either side of the continuums. It just helps people understand better than a cluster of words on one side and a single word on the other side.
You want it to appear balanced, and then let's find the right balance. If it's right out of the box, it looks unbalanced. It looks like there's more on one end than the other. That might influence how people interact with these things.
Using true opposites. If we're saying it's either going to be awesome or it sucks and put that on the continuum, that doesn't work. Or it's going to be tall or short, well, that doesn't really work very good either.
Rather than saying, "What are these true opposites?" let's find things that want to held intention. We try to sort these continuums into similarly granular clusters. There are certain performance continuums that we use at a feature level, at an interaction design level almost, but then there are other ones that are really big global ones that have to do with the business.
Clustering them and using similarly granular continuums is a good way to use them. Another thing would be don't start sorting and narrowing these potential performance continuums too soon. Many of our information architects build a Google doc with a stack of 50 of things. As the project discovery process progresses, they winnow it down. So making things be good, not just look good.
Objections to what I have just described? I want objections first.
Woman 1: Do you have...
Jared: Wait, hold on.
Woman 1: ...stories and then a successful one that you did?
Dan: Yes. The American Society for Aesthetic and Plastic Surgeons engaged us to consolidate a couple of micro sites and their main site into one unified presence. The reason that they had spun off these micro sites was because they thought it was an either/or kind of world.
We can either have a site that's for people who are OK with surgical procedures. But then there are people for whom Botox and more lightweight types of procedures are OK, but if they see pictures of operations, they're going to freak out.
We built out a set of performance continuums to describe all of these tensions in audience types, business models, performance criteria. How should we measure the success of this? That tool allowed stakeholders to agree on what good means for this project to take fragmentation and make a consolidation out of it.
I would say they are successful in all of the projects that we've done since Publishing KO. The interesting silver lining to Publishing KO's story is we're working with them again. They shed a HIPPO.
Dan: They were able to agree more on what "good" means. So we're working with them again.
Woman 2: Hi. Over here.
Woman 2: This seems like a great tool for getting your stakeholders on board. How do you reconcile it with what the clients actually want? Because that's something that I run into a lot. [laughs]
Dan: In our work, and this is maybe just an anomaly of the way that we work, our clients and stakeholders are synonymous. We have the luxury of working with deciders. It's their budget. It's their project. It's their business performance that is on the line, so we use these tools with them because again, approval, paying the last invoice, [laughs] all of these things that we like to see happen depend on a shared understanding of what "good" means.
If we have this robust way to describe with language before we're talking about any pictures, of what good means, executives can deal with a couple of words and some lines and where should we go? Some of them love it. They want to do it in real time. They're like, "All of you shut up. Boom, boom, boom, boom, boom, boom. Go." That's what good means.
If they come back to us during the design process and they ask us to develop features and functionality that are not aligned with what they said good means, we then bring back the performance continuums and say, "Hey. I would love to make that feature for you, but it flies in the face of what you said good means here, here, and here."
"Would you like us to refactor these performance continuums? Can we resave this document and have it be version 1.7 because we've changed what good means? Then our team will refactor all the stuff that we're doing toward that new target of what good means. But if you're saying what we said good means a couple weeks ago stands, then the thing that you asked for is actually against what you said you wanted."
Woman 3: You find that syncs up with what the users ultimately want? At all?
Dan: The user validation side of this can get really interesting because stakeholder intent doesn't necessarily come from an understanding of what people want. In certain contexts when we have the budget and the time to apply this process after the stakeholder intent, then we've also used it internally.
We haven't sold it yet as a deliverable in the user research phase, although we try to [laughs] say, "OK, based on all these conversations with users" we sneak it into our persona work. That's the place where we express it. We use the continuums to describe personas and where they're at.
But yeah, what the bosses and what the stakeholders and the people who are paying the invoices say is good, once you start to figure out relative to what they said they wanted and what would be good, OK, now what do people think? Yeah, sometimes it's like...
Man 2: Do you try to find yourself thinking about the user on your side when you're having those discussions around what is good? Do you inject yourself into that sort of thinking? "OK, there are people that they might not be thinking about." Do you become that advocate?
Dan: At the early upfront discovery and strategy point of the process, which is mostly where we've used these today, we do not yet know anything about users. We actually use this so that we don't have to boil the ocean with user research.
We trust that the people who are the bosses who are paying our invoice who incepted the project know enough about what people want to tee this up. Then we figure out later on how close to what they thought was good is what people think is good.
Jared: We have time for one more question. Somebody else?
Dan: I will also still entertain an objection. I would prefer one, in fact. One of the questions that I've had with these is why is there a zero? Does there have to be a zero in the middle? For some people, the zero is a destination. If we were at two toward this, having zero in the middle is helpful because zero is a destination. Zero is a weasely place to land, to start from.
If they're saying, "I want a perfect exact balance between these two things" and maybe that's another do and don't that I should add is try to encourage people to not start from...if where you land with a stack of 15 of these is in the middle of all of them, you have to do it over because that isn't something you can design against. Yes?
Man 3: What types of skill sets is on the teams of yours that are determining this and the that and the what's better or what's good?
Dan: We have a number of graduates of the University of Michigan's School of Information who I will cherry-pick before anybody else can have a crack at them.
Dan: Some of you are next. So master's degree in Human Computer Interaction or Library and Information Science is some of the background that some of our people have. Two of our team members have undergraduate degrees in philosophy. Many of us have interests and training in linguistics, semiotics, and nerdery around language and meaning in language.
Jared: Thank you, Dan. Thank you very much.