look at it another way

pouring wine into the "glass" that appears between two other bottles

This article originally appeared in issue 267 of A List Apart in September 2008. Reprinted below.

“Before you can solve a user’s problems, you must see them as that user sees them. Once you understand what drives people’s behavior, not only do new ideas flow freely, but the ideas that flow are appropriate and useful.”

Look at It Another Way

Remember as a kid the first time you saw the vase/faces image? Remember staring until your perspective shifted and you saw the “other thing” in the picture? You probably felt a thrill accomplishing it. Did you show your sibling how easy it was, vaunting your ability to see both images? It felt powerful—a revelation.

The field of user experience (UX) helps us change that. UX practitioners examine the everyday lives of the people we endeavor to help. Bill Buxton, in his presentation at Interactions08, offered the best illustration of UX I have seen. He showed a catalog picture of a mountain bike next to a picture of a rider careening through a stream with water splashing everywhere. You couldn’t even see much of the bike in the second picture. Bill said, “You don’t buy a bike. You buy the right to scare yourself to death.” This is exactly it!

Image of mountain bike from Specialized catalog, shot against a white background with no foreground, showing only the bike.

Catalog bike image from www.specialized.com. Used with permission.

Photo of bike being ridden through a stream in the Quantocks, UK, where all you see is half a wheel, two legs, and the brilliance and motion of splashing water.

Rider challenging his skills by riding through a mountain stream, Quantocks, UK.    photo credit: Gary Foulger (Creative Commons) Flickr.

The people who designed the bike talk about what the bike can do, but the rider wants to find out what she can do. In the former vocabulary: “There are 20 gears.” In the latter vocabulary: “I’ve decided to bike to work twice a week, but I fear the pain of getting up that steep hill on the way there.” If the bike designers were smart, they’d be talking about making it easier to get up hills while commuting to work, or suggesting alternate routes or techniques so that you’ll arrive at the office without needing a shower and a nap.

Pretend your organization doesn’t exist

Thinking from the potential customer’s perspective is a Zen-like exercise. It requires you to drop your problem-solving role completely, and spend time—at least two hours, or maybe two weeks—engrossed in the world of this person. Stop thinking of them as a “user” of the thing you provide. Think about how and why they accomplish what they want to get done, not how or why they might use your stuff. Pretend you and your organization do not exist, and study what this person does with all the resources available in her life.


For example, what does a citizen need from her town government? She needs a way to get from her house to the grocery store, the library, the post office, her workplace, etc. These could be roads, bike paths, public transit, and sidewalks. She needs utilities like water and electricity to be delivered to her property. She needs assurance that her property will be defended from fire, protected from floods, and accessible during a disaster. She wants to feel safe from assault, whether by a human, an animal, pollution, noise, or disease. This list goes on. When you look at how this person would approach the town government, she is faced with a list of departments:

  • Parking Services
  • Public Works
  • Redevelopment Agency
  • Emergency Services
  • Community Development

To whom would she turn if she wanted to report a strong pesticide smell? If the list contained titles such as the following, she might find the right person a little faster:

  • Getting Around Town
  • Water, Sewer & Utility Infrastructure
  • Fire/Flood/Disaster Preparation
  • Concerns About Feeling Exposed/Invaded
  • Changing Your Property
Look for differences in audience beliefs and behaviors

This step can be tricky. Almost always, there are different types of people trying to do slightly different things. Most organizations divide these people into segments separated by their relationship to what the organization produces. For example, I see much attention paid to whether a segment is “web savvy,” has a certain social or economic power, or is a new or existing customer. Defining groups by their relationship to your product blinds you to the relationship they might have with products you haven’t thought of yet.

That may sound pretty axiomatic, but consider that the leap from existing products to something groundbreaking is still thought of as genius. To dispel this belief, many folks have argued that innovation is the result of constant, earnest effort. This constant effort becomes frustrating when aimed at the customary perception of audience segments. You can wear a groove in your brain thinking through the same old problems defined in the same old way. Rather than looking for differences in how segments use your product, look for differences between the beliefs and behaviors of these segments in real life.

Like the vase and the two faces, suddenly new, less complicated segments will come into focus. For example, rather than categorizing customers of a statistical graphing package by industry and expertise, they might fall into two behavioral segments like “people who think visually and like to sketch what-if situations” and “people who need to communicate a result.”

Play psychotherapist to key audiences: explore root causes

Select the groups most important to you and get to know them deeply. Ask them to explain their approach to achieving something. They may not have formally defined the approach in their head yet, so keep asking and help them articulate why they do certain things. Is their behavior motivated by a belief or a personality trait? What are their reactions to things that come up along the way?


Let’s say you’re the administrator for a public recreation center. You zero in on the pool as a conduit to increase summer revenue, and you know that swimming lessons are an important contributor. You tried offering more levels of swimming lessons last summer, but it didn’t attract the crowds you were hoping for. This winter, you decide to study people who might be interested in swimming lessons. Instead of looking at “kids” and “adults,” you divide people into new groups:

  1. People interested in fitness.
  2. People inspired by the current Olympic swimming hero.
  3. People afraid of drowning and want to do something about their fear.
  4. People compelled by others to take lessons.
  5. People interested in improving their technique to get faster.

You skip the fourth group because you feel like you understand that group pretty well. (Think of kids enrolled by parents to take swimming lessons, or asked by swim team coaches to learn a new stroke.)

You study the remaining four groups. During listening sessions with members of group three, you are prepared to hear a lot of emotion. What surprises you is that these people don’t ever see themselves in a pool at all. They see themselves surviving a capsized boat on their upcoming nature cruise in Baja. They want to appear relaxed and natural at the family reunion at Lake Shasta—able to play with the nephews and nieces in knee-deep water. You get the idea that they don’t want swimming lessons so much as “drowning avoidance”/confidence lessons. They want their fear respected, recognized, and taken care of.

Based on your research, you advertise a new kind of class. You run it in the evening when the pool is not full of observers. Participants don’t wear bathing suits or even get in the water—they’re encouraged to describe the specific situation they’re afraid of, and watch as an instructor demonstrates different methods to survive in the water. You get six people at the first class—one who is 13 years old and others aged 37 through 63. You get 22 people at the next class. You really met the needs of this group, and word gets passed throughout the community. It’s a big success throughout the summer.


What if you’re the VP in charge of internet customer support at a really large telecommunications corporation? Customer satisfaction surveys indicate that people really aren’t getting what they need from your division, yet figures show that 88% of the questions are “completed.” The kicker is, your boss tells you to expect budget cuts.

Your predecessor established all sorts of workflows to allow customer service reps to route customers to the right information. Last year, the voice-activated system was upgraded to try to answer more questions before involving a real live rep.

Instead of erecting more processes, you decide to explore the root of the problem. You review 40 recorded calls from the past month and a picture forms in your mind. Callers have a really negative reaction to interacting with a system, even if that “system” is actually a rep. The negative reaction stems from being forced to state things discretely, one idea at a time, in a branching manner that the “system” can deal with.

System: Is this line you’re calling from the line you are having trouble with?

Customer: Well, yes, but I’m not specifically having trouble; I just want to ask about compatible modems for my DSL service.

System: Is this line you’re calling from the line you are having trouble with?

Customer: Yes.

System: Are you having trouble with your password or with connectivity?

Customer: No, neither!

System: I’m sorry. I am having trouble understanding your response.

Customer: Just give me the OPERATOR! [yelling]

Rep: Hello, I’m John. Welcome to XYZ where we appreciate your business. Let me look up some data about this call. Are you Suzie Queue?

Customer: Yes.

Rep: Thank you, Suzie Queue. You have reached the technical support department. We are here to answer your questions. I see that you have Super-Ultra-Internet service and you’ve been a customer for eight years. How may I assist you today?

Customer: Well, I want a list of compatible modems for this service I have, and I want to find out how to log into the free wireless that comes with my account when I’m at Starbucks.

Rep: Let me see if I understand your question. You want to purchase a modem?

Customer: Um… Not right now. I just want to see which devices are compatible so I can do a little shopping around on the internet. I did have two questions, but we’ll get to the second one later, I guess.

Rep: You don’t want to purchase a modem?

Customer: I do, eventually, yes.

Rep: I apologize, but I cannot help you here in tech support. I’ll need you to call the Store for that information. Here is the toll-free number…

You realize that even your reps have been dumbed-down to replicate the digital system, reading possible topics from a branching list without putting any effort into a human conversation.

The root of the problem is that customers are looking for answers they couldn’t find online or through the digital voice recognition system. When they reach a rep, they expect an intelligent being who is able to provide more complex answers.

You decide to add “none of the above” options at every level of the digital voice recognition system. This allows the customer to navigate through the system or not, acknowledging that they know whether their question is simple or complex.

You decide to put more complex information online where customers can find it, based on customer needs mentioned in the recorded calls. You describe it to other VP’s as “maintainer information,” while pointing out how it differs from the pre-purchase marketing information currently out there.

Finally, you yank all the branching topics from the reps and ask them to talk to the customer informally, the same way they would if they were a detective helping someone solve a mystery. By getting rid of the self-limiting branching topics that assume there are no mysteries, you train reps to extract information through conversation, and how to use the internal company knowledgebase. You allow them to instant message each other and other key people in the actual production departments.

Over the first month, the reps identify among themselves the people who have the most knowledge on a certain topic. They are free to say, “I don’t know the answer, but let me ask somebody in another department who might know.” There is now a culture of respect for those with the knowledge among the reps, and a bit of competition to become the “go to” person in a certain area.

Those that aren’t as enthusiastic about the new culture leave. Those that remain become much better equipped to move up into other positions within the company. In the next customer satisfaction survey, your scores increase. You have fewer people answering phones, and more questions being answered by information online.

Build a model of behaviors and how you support them

When you understand what drives people’s behavior, you can imagine new ideas pretty easily. In each of these examples, the protagonist formed a model of a specific segment of people. The model was based on listening sessions and outlined the mental processes of the people as they went about their business, including the underpinning philosophies and emotions behind certain actions.

The protagonists took these “mental model diagrams” and mapped how they were already supporting certain things but missing other areas. In the areas where there were gaps, the protagonists came up with several ideas that might support that part of the segment’s mental model.

Sketch your mental model like this: put behaviors, beliefs, and reactions across the top, and the ways you support these things as a foundation underneath.

A thumbnail of a horizontal mental model diagram with a callout showing detail of one section. One of the towers shown contains behaviors such as

An example mental model diagram of dating from Engage.com, with behaviors, beliefs, and reactions across the top, and the ways Engage plans to support them underneath.

Harness the visualization for a wider perspective

Mental model diagrams can fall prey to our own assumptions and understanding of a particular field. When creating one, turn off your internal problem-solver and just listen to people. Allow patterns of behavior and motivation to reveal themselves to you. Work from the bottom up, rather than designating several behavior areas and trying to fit people’s actions into them. Make sure you’ve asked questions that dig into a person’s soul. Find out what makes them tick. For example, you’ll find that many health patients go to a doctor not to “remove pain” but to “get back to my old life.” Removing pain is just one underlying root motivator.

When your organization redefines the way you see and support customers, you take a step toward a more mature design approach. Instead of simply making existing solutions work better or applying stylish cosmetics, you open up new perspectives that allow you to see something that wasn’t apparent before.

Focus first on what it’s like to be these people, and then focus on what you have to give them. With this vision, creativity tumbles forth.