In 2005, Sarah B. Nelson contacted Greg Allen, the founding director the Neo-Futurists, a theater collective known for performing an evening’s worth of two-minute plays. Nelson, who has played violin since age three, was not looking for a job, but for ideas for Web design teams―as part of her research for a master’s degree from Chicago’s Institute of Design. This academic work was a break for Nelson, who had already been in design business for ten years. But she called Allen to follow a hunch: that when it comes to creative collaboration, Web designers have much to learn from the theater.

The similarity between Web design and the performing arts is especially pronounced in the early phases of the process, where two-minute plays written by troupe members get critiqued. First, the group makes suggestions on how a work can be improved. But then, the group turns a corner. It considers which of those plays are good enough to perform, and which should be (to put it bluntly) tossed out. Then, having made those decisions, the troupe starts rehearsals. These two phases, which Nelson calls “⁠divergence⁠”⁠ and “⁠convergence,⁠”⁠ require a different mindset. Divergence is about exploring ideas; convergence is about executing them. It’s important, she says, to understand the difference and to know which phase you are in.

Nelson has also experimented with a technique many professional Web designers would find appalling: bringing in clients to participate in the early stages of the design process. While she doesn’t pretend the process is easy, or that it works for every client, Nelson argues that getting early involvement of the group can save hours of revisions later on.

Nelson’s portfolio of clients includes artistic organizations like the The Metropolitan Opera and Seattle Symphony, Flickr, Nintendo and AOL Mobile, the Nordstrom department stores and several financial companies. She writes a blog, Cartographies of Imagination, and recently departed on friendly terms from the San Francisco design firm Adaptive Path to found her own firm, Tapir Studios, in the city’s Mission District. When we spoke by phone, Tapir was so new that Nelson had yet to put up a Web page.

A lot of your research work has focused on the relationship between Web design and the performing arts. What’s the connection?

When I first started working in interactive media, it occurred to me that operas were the first multi-media production. They require music, theater, costumes, set design―all done by specialists who are collaborating. That can be true with Web design, as well. If the project is large enough, no one person can do it themselves. You really must work with other people. You can choose to be like a bulldozer―plowing through and insisting on your own ideas. Or you can use the power of the group. I know that my work is a lot stronger when I am working well with another group, or at least one other person.

How did you become interested in creative collaboration?

I got interested, ironically, because I was very uncomfortable with it. My training was in visual art, which emphasizes each individual’s point of view. That idea works during the later production phase of Web design. When the group executes an idea, there are clear role definitions: the art director, the developer, the interaction designer. But that’s not so true earlier in the process during the design strategy phase. Here, you are trying to generate ideas, and anyone can have a good one, regardless of their job title. So in the first stages of a creative collaboration, the roll definition gets blurry.

You came from a musical background, playing the violin from age three. Did you learn about collaboration from that?

I played in a lot of orchestras, so you would have thought I would have grown comfortable with collaboration from an early age. And so I wondered why I had so much trouble with collaboration during the strategy phase. But classical music has very clear roles. You are given your part, and your job is to learn that part. Then you learn to bring your playing together with others. But in design strategy sessions, this orchestra analogy doesn’t work. There is no equivalent of “⁠sheet music,⁠”⁠ so people often don’t know where to go next. Or they encounter a problem they can’t quite identify. So as opposed to playing in an orchestra, the design strategy sessions are extremely ambiguous and highly subjective, with the “⁠parts⁠”⁠ overlapping. It isn’t about one person’s ideas; it’s about the group of ideas and how all of those work together within a complex set of criteria and boundaries.

In graduate school, you studied a theater group, the Neo-Futurists. Aside from their being down the street from your home, why did you choose them?

One reason is that they create two to 12 new plays every week. That requires a very rapid creative process. I was curious how they did that. The other was that the actors were both writers and performers ― I thought that had some parallels to design work.

What parallels did you see between them and design work?

We had at least this in common: the group had to hit deadlines, no matter what. Their curtain time was fixed, and they had to be ready. In design, too, we are often under very hard deadlines. The Neo-Futurists do that every week, so they get a lot of practice. Also, because they work together as a team over many performances, they tend to learn from each other. That can happen with design teams, as well. Over time, they get to know each others⁠’⁠ strengths and weaknesses, and know what each other needs in order to support them.

Later I began to see another parallel that was less obvious. The troupe members each write and perform their own plays. And as they worked together, each would take ownership over part of the process. This didn’t happen formally, but in time, each person took charge of something, and that something seemed to come out of their own heart. Moreover, I saw that people were very willing to change the nature of their role. When they worked on their own play, they were able to own the process. But when they needed to be the support person for someone else, they were able to do that too. They would switch back and forth in these roles. “⁠This is my vision, and so I own it. This is someone else’s vision: so I will help support that. Ultimately, of course, it is important for the leader to specifically assign these role so that it is clear who is responsible for what.

It sounds like you learned something about roles, in general. In a creative collaboration, sometimes a person’s job title matters. Sometimes it doesn’t.

On a successful design project, you start from a very ambiguous place and wind up building something very concrete. It’s that first, blurry phase, what I call “⁠divergence,⁠”⁠ that I’m talking about here, and that’s where the roles blur, too. A design team might consist of a creative lead, a designer, a developer, and a copy writer. Each is a specialist, but in the beginning, where you are exploring the context, it’s important for all these people have the same understanding. The whole team should know the background of the organization. They should all have some direct experience with the company’s customers and first-hand experience with the products. But the formal roles for each member don’t much matter. It doesn’t make a lot of sense to say that―because I am a research person, only I can have valid insights into what we all saw in a customer visit or usability test.

Eventually, the team moves beyond the idea generation stage, and now, everyone does bring in their unique professional skill sets, along with their own human experience. As a project lead, I see my role as bringing all that out in people.

You say there is a transitional point where the brainstorming ends and the critical work begins. In the first stage, no idea is bad. But then suddenly, you must kill all the bad ideas off.

Yes―this is the trickiest part, requiring two different mindsets. A lot of strife comes from people being at the wrong mindset at the wrong time. It’s also a difference between developers and designers. Developers are often tacked on to the end of the design process and they are under the gun to make the thing happen. Execution is their biggest concern. Whereas designers are always exploring possibilities, so it’s hard for them to close down that process―which can lead to chaos. Turning that corner can be difficult, and it’s up to the leader to set the appropriate tone. But both divergence and convergence are important. Design is not just about generating ideas, but evaluating them so that you get what you need.

You’ve said that during production, the Neo-Futurists were good at quashing ideas while still giving people their dignity.

Well, the Neos wouldn’t put it that way: they are much more gentle than that. But they do have a well-defined technique for removing plays from the previous show’s menu. They read a play, and if nobody says “⁠keep⁠”⁠ or at least comments on it, the play get junked. The group doesn’t discuss or critique it: they just silently kick it out, which preserves the dignity of the author while recognizing the piece’s time is over. That’s important in design, too, even though an idea may be very near to the heart of its creator.

Web designers sometimes fall in love with their design, but forget their audience.

It’s important to judge ideas based on strict criteria. That’s why in Web design you research the competitive landscape, and get to know the organization and its customers. You lay this foundation so that when people become emotionally attached to ideas, you can shine the light of rationality on the process. You can ask: does this design idea actually help this company compete? Does it differentiate them? Does it address its customers⁠’⁠ needs? Web design isn’t a science, but it has science-like moments where you go back to the project goals, to the criteria we laid out for success, to what users want and the business needs to make it successful.

Designers work is in service to a business and its customers. That means we have to constantly compare our own agendas with what is actually best for the client. That’s a constant battle for all of us. I see this when young designers first start: they have strong opinions. Many of them are valuable. Their next step in professional growth is to learn to back away from those opinions when they don’t meet the larger goal.

You’ve actually brought clients into the brainstorming process. How do you make that work?

One of the biggest challenges is that you are often working with people who are not creative professionals―so they don’t have the tools, language or experience that the members of the design team have. On the other hand, they know things about the project that the designer doesn’t and can make a big contribution. So, when this technique succeeds, you are designing with an organization, rather than for it.

What kinds of clients make the best collaborators?

The single most important quality is the ability to let go of their own idea in favor of another, no matter where that idea came from. Part of this is group dynamics. Some groups are very hierarchical, with some people deferring to others. For them, collaboration can be challenging.

We in the West tend to think of Asian business cultures as being highly hierarchical.

There is definitely a cultural component to the collaboration. I would love to hear about how it could work in a Japanese environment. I learned a lot about collaboration from the Korean students when I was in graduate school. As Americans, we learn from an early age to talk a lot and to jump into a conversation. Whereas, the Koreans were often very reserved. They only contribute ideas after they had carefully thought about them, whereas we “⁠think⁠”⁠ with our mouths―we throw out ideas without worrying if they are perfect.

So I would structure a workshop in Korea very differently from one I would do one in the states. I would want to understand the culture better. And I would try to figure out how to get ideas from people, without making it uncomfortable if those ideas aren’t always the best.

So if the best idea came from an administrative aid out of college, his boss should be able to accept that.

If it’s the right idea and supports the business and its customers―then yes. It does help if that spirit is part of the corporate culture. In companies where I’ve seen the technique work best, people are already talking to each other a lot. But even in the best situations, this fluidity doesn’t happen immediately. Collaboration takes effort.

Can you give me an example of how the process actually works?

I worked with very large financial institution whose design team alone was 80 people. We ended up working with representatives of their internal design and about 30 different business stakeholders within the rest of the organization. These people were definitely not from design. They were experts in retirement funds. They were used to working together in meetings. They were used to communicating through presentations.

The first session was a structured brainstorming session where I used scenarios to frame the idea generation. For example, imagine a working couple age 35: what would be the best way to help them save for retirement? Each group would do this for a different kind of customer.

So we all sat around a U-shaped table, and I explained how the session would work and what we hoped the outcome would be. But the group didn’t want to move on. They wanted to talk more about the process. It was as if they had reached the end of a diving board and were afraid to jump into the pool. Why? Because they weren’t creative professionals. They weren’t use to thinking like this. They were used to making concrete business decisions, while this process was abstract and hypothetical.

So I told them to try it―just once. They got into their groups. I put a facilitator with each group to move things along. It started slow, but then the ideas started to come out. We wrote them down, put them on the walls, the walls got full, and put more ideas on the windows. Eventually, the ideas covered the entire room.

What happened after that?

These guys went through three more work sessions to evolve the ideas further. At the end, we combined all the ideas into 12 idea sets―for an online application or a data visualization of a retirement account. Then we had them do a business case evaluation for each. Does the idea set really serve the customer? How could you make it better? How viable is in the idea in the marketplace? Can the company really deliver on it? Each person then put a dot on the ideas they thought were strongest, and we narrowed the field to a couple of top candidates. The designers started there, took some strong concepts from others, and that became the foundation for the prototype.

It sounds like every project starts of as a democracy, but it still needs a dictator.

Every project needs a decision maker. On the financial project, there were two things that the group did not do very well. One was synthesis―pulling a bunch of ideas together into something cohesive. That’s what designers are good at. And the other was coming to a decision. In the end, we had two ideas and for an hour we debated the merits of both. No decision was made, and I don⁠’⁠ think a decision could have been made in that environment. What was needed was the strong recommendation of the designers and a strong decision maker internally.

The client put in a lot of effort at the beginning. Did that pay off in the end?

Right: the process is top-loaded. You put in effort at the beginning and avoid those endless rounds of revision that happen at the end. Sometimes at the end of a conventional design project, someone who didn’t feel listened to will throw a monkey wrench and bring the whole process to a halt. That didn’t happen on this project. Our final recommendation was a little bit different from the initial idea sets, because we as designers used our expertise. Even so, everyone said “⁠okay.⁠”

But I had another client, an academic institution, where people even voted on the design of the labels for every navigation section. That’s not something I recommend. It was laborious, but we finally got to the point where we could start the design work. And then a key stakeholder came in who had expressed zero interest in being involved. He said he trusted us. But then, after 18 rounds of revisions, he wanted to step in. We spent two years on that project.

You’ve also worked with creative institutions like symphonies and operas. Are they any better at collaboration and idea generation?

I don’t think it matters. Major art institutions are still organizations. They have the same constraints that businesses do and have to figure out a way to run efficiently. They have to attract people and think about marketing just like every other organization. No matter what the nature of the project, company, or business situation―the biggest challenge is in dealing successfully with people.

Sidebar: Tips for Managing a Creative Environment

At last year’s South by Southwest (SXSW) festival in Austin Texas, Sarah B. Nelson and her Adaptive Path colleague Bryan Mason gave a presentation on managing a creative environment. Their collective wisdom came largely from conversations with people in the performing arts, including the Neo-Futurists and Avenue Q.

1. Cross-train the entire team: everybody should have some familiarity with everyone else’s job.
2. Rotate creative leadership.
3. Actively turn the corner: phase 1 (divergence) is exploring new ideas; phase 2 (“⁠convergence⁠”) is narrowing those down. Don’t confuse the two.
4. Know your roles: once production begins, work mainly within your own expertise.
5. Practice, practice, practice: not just your own skills, but those of the group.
6. Make your mission explicit to the whole team.
7. “Kill your darlings⁠”: the expression means that even if you are in love with an idea―if it’s not working, kill it.
8. Leadership is the ultimate support service: be a facilitator, not a dictator.
9. Generate projects around the group’s creative interests.
10. Remember your audience.
11. Celebrate failure: you learn more from what went wrong than what went right.