In many endeavors, let alone product development, there's a gap between knowing what's best and actually doing it. 

Gaining behavioral insights about your potential customers usually falls into one of these gaps. The reason? It can be awkward (and exceedingly time-consuming) observing and recording the behavior of people you've never met as they use (or struggle to use) your product.

"A conversation with a stranger can be a little, well, strange," writes Jon Kolko, Vice President of Consumer Design at Blackboard and the Founder and Director of Austin Center for Design, in his new book.

But founders, even shy ones, should take heart. For one thing, there are ways to overcome the social challenge of ethnographic research. For another, the insights you'll gain about your customers' behaviors will drastically supercede anything you'd learn from focus groups or online surveys.

Due out this week, Kolko's book, Well-Designed: How to Use Empathy to Create Products People Love, provides a smart, replicable set of steps for observing customer behavior in authentic settings--and acting on the findings. Kolko's research on students helped him build MyEdu (a startup that Blackboard acquired) into a platform more than 1 million students use in their job hunts. He has seen firsthand that a product's emotional acuity can influence customer decisions far more profoundly than its functions and features.

Here, in 10 steps, are Kolko's methods for gaining behavioral insights about your customers: 

1. Establish and articulate a focus. Specifically, you should write a one-sentence description of the scope of your customer research. Here are three of Kolko's examples:

  • To learn about the way people use and think about banking services.
  • To watch the way orders are placed at a given business.
  • To see what it's like for families to go to the movies. 

The idea here is twofold: First, it can help you choose which people you want to observe--and in which settings. Second, if you're worried about the awkwardness of talking to strangers, your articulated focus can help you steer conversations in a useful direction.

2. Prepare a set of questions. The questions should focus on customers' behaviors and emotions. Examples include: 

  • Can you show me how you use the software to process that order?
  • What's your least favorite part of shopping at this store? Can you show me why you don't like it?
  • Can you remember a time when your vehicle broke? Can you show me what part broke?

If you noticed "show me" is a big part of these questions, you've got the gist. You want customers to show you their actual behavior in decision-making situations. By contrast, you'll learn little--and observe little--if you ask yes-or-no questions such as "Do you like this product?" or list-based questions such as "Which three features do you use the most?"

Those questions are ideal for online surveys in which you're seeking a high, rapid response rate. But in real-world settings, you want to probe for the emotional details informing customer decisions. 

And under ideal observational circumstances, you might not even need to ask your scripted questions. You'll already have gained the relevant insights by simply watching how your customers behave, at the bank or at the movies or in whatever setting pertains to your would-be product. In other words, customer actions, in and of themselves, might provide most of the insights you're seeking. There'll be little need to follow up with questions probing for verbal explanations of their actions. 

For example, Chris Michaud and Kevin Young, two of the designers at Felix, a $1.5 million startup which makes cases and stands for smartphones and tablets, didn't ask would-be customers what they wanted. It was obvious to them--simply from watching people struggle to prop up their iPads on the seatback trays of planes and trains--that consumers were frustrated by difficulties in setting the tablet at a stable reading angle. In designing Felix's FlipBook, which serves as both a case and a stand, the design brief was simple: "You should be able to set it with no thought, and it has to hold," Young recalls.

3. Get in context, and record everything. Context is crucial. You don't want participants recalling how they felt when they were making decisions. You want to watch them as they act, in real time, in the relevant setting.

And in the same way your participants shouldn't rely on memory, neither should you. Ask permission to record them. Even if it's just audio. You'll return to it repeatedly. You may even need the footage at a later date for marketing, or to persuade colleagues about one thing or another. In addition, you should pay attention for other methods to document their behaviors. Including these next three steps:

4. Ask to see examples. In Kolko's ethnographic research for MyEdu, he and his team spent a "ton of time with college students in dorm rooms, looking through their book bags and apartments [with their permission]," he told Inc in a recent phone interview.

What came out of that was a sense of what you could call student artifacts: relics and objects of a cultural interest. You can even draw parallels between the ethnographic methods teams like Kolko's have used and the artifact-seeking methods legendary explorers Meriwether Lewis and William Clark used to learn about native populations in the western part of the United States in 1803. 

For example: Rather than relying merely on interviews, Lewis and Clark made sure to sketch and collect every artifact they could. "Everything from Arikara corn and tobacco seeds to a Mandan buffalo skin painting," writes historian James P. Ronda, Barnard Professor of West American History at the University of Tulsa, in Lewis and Clark Among the Indians. 

5. Ask to try it. "For example, if you are watching a butcher prepare meat, ask if you can make a few cuts. If you are observing a college teacher grade papers, ask if you can grade some," writes Kolko. "You haven't lost anything if the person says no." But, he adds, if the person says yes, you'll be that much closer to gaining the empathy that superb product design requires. 

Remember, trying something can take many forms. For example, on many occasions, Kolko's MyEdu team watched TV with undergrads. It provided them with a window to the world of student emotions. "And what came out of that was insight about their decision-making," adds Kolko. "It was not utilitarian, it was emotional. College students are filled with anxiety. Around what school to go to, what major to pick, what classes to take, and what job am I going to get."

Likewise, Lewis and Clark's expedition participated in hunts, games, and ceremonies, whenever it was possible. John Ordway, the sergeant major, once tried to play a game involving hoops and poles. He couldn't understand the rules. But the point is, he tried to play. The analogy here is that you want to ride alongside customers in the Mustang or watch them as they install their software. You don't only want to interview them later about it, useful though that may be.

6. Watch the outliers. Think of this as a form of beta testing. You haven't really tested your product if your only participants are mainstream users with predictable behaviors. You need to gauge--and gain empathy for--the responses of unformulaic minds.

"Try to find anomalies or outliers, people who have extraordinary views or attitudes," writes Kolko. "These anomalies present new, provocative frames, and they can be useful in helping you see the world in a new way."  

7. Transcribe, externalize, and build the product synthesis wall. This is not a figurative wall. It's an actual wall, in an office, like the one depicted below. The wall should include quotes on Post-its from research participants; photographs of those participants in action, in context; even artifacts from your research that can be pinned or taped to the wall. 

 inline image

The idea is to "see the data in a new way and to question your preconceived understanding of hierarchy, relationships, and causality," writes Kolko. In so doing, you'll also create a visual "blend" of your research. This is why the designers at Allsteel, a maker of office furniture based in Muscatine, Iowa, refer to their product-synthesis wall as "the blender."

Moreover, the wall allows everyone in the organization to glimpse--and comment on--your findings. It can jumpstart dialog on a project, and inspire colleagues as a visual manifestation of the sweat equity that real ethnography requires. 

And make no mistake: Creating the wall is tedious. "You'll be tempted to make your intern do it," writes Kolko. But it's vital for you to do the work yourself--including the labor-intensive transcription of quotes and observations from your research participants. "As you type, word for word, what you heard and saw, you'll find yourself reliving the experience and achieving a sort of meta-analysis of what each person said," he writes. 

In his interview with Inc, Kolko stressed that there is no short-cutting this process with digitized sharing. "It doesn't work at all," he says. "It even serves a negative role because shareholders think they're following your train of thought. But they don't get the nuances and extremities and insights that lead to the conclusions you need to make."

8. Identify patterns and anomalies. You're trying to figure out why customers behave the way they do. Over time, as you scan the wall, you should highlight behavioral patterns. Feel free to move Post-its and pictures around the wall, pushing the similar ones together. 

This process will take at least one week, likely longer. How many hours should it take? Kolko suggests a multiplier of roughly three hours per research participant. As in: For eight to 10 research participants, spend 20-30 hours with the data. 

9. Visualize behavior across time. Drawing diagrams on whiteboards is essential to this step. Your aim is to illustrate customer emotions and decisions across stages of a process. 

These diagrams need not be comprehensive. You simply want to map out a customer's emotional journey. For example, at MyEdu, Kolko's team recognized that college students felt a great deal of anxiety around their resumes. They felt (not inaccurately) that recruiters were making snap judgments of their credentials based on rapid resume scans.

Worse, they felt as if the traditional and online formats of resumes didn't accurately capture the skills and accomplishments of their experiences as students--or their ethical character as people. "And so we capitalized on that in creating MyEdu profiles with movable tiles, a highly visual format," says Kolko. 

As opposed to conventional resume formats (both on-paper and online), which array one's credentials in a rigid vertical stack and tend to emphasize job histories, the movable tiles allowed students to feel as if they had plenty to offer--even as newcomers to the workforce. In this way, MyEdu's solution addressed an emotional pain point that one set of its users (students) were feeling at the early stages of their journey in the job-seeking process. 

Likewise, the tiles also allowed another set of users (recruiters) to do what their jobs necessitate: Make fast decisions based on visual snapshots. "The tiles met the demands of the recruiters, who are in the snap-judgment business," adds Kolko. 

10. Communicate your results. It's one thing for you and your product-development team to grasp the emotional realities behind customer decisions. It's another to share those insights with other relevant shareholders. The product synthesis wall is a useful first step. It can break the ice to colleagues who are outsiders to the behavioral-research process. 

Beyond using the wall, Kolko has found it helpful to sketch your key takeaways and conclusions. "Two really low-hanging tactics work every time," he says. "Tell a problem-and-solution story and visualize it." It doesn't have to be anything high-tech. "A crass or crude comic strip really helps ground what you are talking about and why it's appropriate," he adds. In your sketches, you'll want to include: 

  • Any quantifiable data connected to key behavioral insights.
  • Quotes, audio, or video from participants illustrating those selfsame insights and bringing the numerical data to life. 

If you've done the advanced legwork of letting outsiders view your product synthesis wall, then very little you communicate will surprise them. You'll already have buy-in. Those outsiders will already "understand the invisible or intuitive storyline," says Kolko, "and start selling it for you." 

After reading about this process, you might conclude it's too complicated and draining for your temperament. You may be right. But remember: Human behavior, too, is complex. If you want to design a product befitting such nuanced behavior, there's no sidestepping the heavy lifting of this process. "It will feel hard, tedious and time-consuming," writes Kolko. "But the results that pop out the other side of synthesis are the elegant truths of innovation, grounded in their humanity and beautiful in their simplicity."