You’re the chief designer of a product that is second in its field. It’s technically adequate, but essentially a me-too copy of the market leader. You want to take over the market. How do you do it? How do you discover the edge-the fundamental new perspective that will make the difference between the leader and the also-rans?

Or, you have the germ of an idea you think is fundamentally new, and will create a whole new market. How do you develop it into a real design? How do you work out the details so that you have a coherent, workable specification for something that will set the market on fire, given that no one has ever seen this kind of thing before?

In our work with organizations around the country, we have seen product teams struggle with both these issues. We specialize in coaching cross-functional teams in customer-centered design-the design and development of systems starting with understanding customers in their workplace and involving customers continuously throughout development. People who are not working from customer data often tell us that whatever merit our approach may have, it can help in neither of the above situations. Customers can’t give you totally new inventions; customers can’t imagine what their lives might be like if you gave them a tool they’ve never seen before. One company we know is proud of not allowing designers to talk to customers, because talking to customers would stifle innovation.

This attitude has always struck us as odd. How can you know where to innovate if you know nothing about your customers? How can you know whether anything will work in practice if you don’t develop it with your customers?

What makes creativity work?

The Invention of the Spreadsheet

Let’s look at an example of creative design in software as a test case. Arguably the single most creative act in software design was Dan Bricklin’s creation of the spreadsheet. This invention created an entire market out of nothing, spawned dozens of look-alikes, opened computer use to an entirely new set of people, and changed the way we approach whole classes of problems. How did he do it?

[I]

Bricklin’s background was in interpretive systems and in typesetting and word processor design-he worked on the software that became WPS-8. In the early 80’s he decided to go to business school, and in class after class he saw people struggle using spreadsheets to represent business problems. Apple was coming out with their first computer about this time, and as he sat in Aldrich 108 he began to daydream about the possibilities of the new technology.

A heads-up display, with a keyboard that moves-the display shows an accountant’s spreadsheet and as you move the keyboard, your focus shifts from cell to cell. Typing enters numbers and formulas and the formulas immediately recalculate the values in cells as you type.

The vision was there. Then came the reality of making it work. The Apple II couldn’t give you a heads-up display, or tie keyboard movement to screen position, so Bricklin made do with the Apple’s display and cursor keys (after trying and rejecting the joysticks as input devices). Throughout the development process Bricklin worked with his user community. He talked to the professors in each of his courses; he had people try the new tool on their own problems, testing as many different kind of problems has he could.

He never lost sight of his users and he never lost sight of his major competition: the back of an envelope. If people could do their thinking on the back of an envelope faster than in his tool, they’d never use his tool. Setting up and manipulating a spreadsheet had to be fast and natural.

Lessons learned

What can we take from this story about how to encourage creativity in software design? Lessons that jump out at us, and which are borne out in our experience, are:

  • Creativity comes from putting the technologist in the middle of the users’ problem. Bricklin was primed with knowledge of technology, especially interpretive systems (key to the idea of recalculating formulas) and visual representations of data (from the word processing work). And by going to business school he put himself in the world of his future users.
  • A creative idea jumps out as a whole, not as a string of incremental features. It may take longer to sort through and refine, but the initial flash of inspiration delivers the whole package. Inspiration is holistic, not analytic.
  • Creativity starts with a wacky idea. Heads-up displays, moving keyboards-these aren’t close to commercially available even now, ten years later. Maybe they’ll never be available. Maybe they aren’t even a good idea. And yet, intermingled, was an idea that could be implemented and would change the world. Bricklin didn’t just have the idea, he dared to go for it.
  • Despite this, the creative idea is obvious-after it’s explained to you. Of course you can put a spreadsheet on the computer-computers add and subtract, no surprise there. Bricklin says programmers couldn’t understand what the excitement was about. It was the users who saw that there was a new thing in the world.
  • Creativity is driven by stealing ideas from customers. Spreadsheets pre-existed VisiCalc by centuries-introducing a computer took the drudgery out, transformed what people could do, and radically changed how business people work today.
  • Creativity doesn’t stand alone. It requires development with users. Iteration with users and trying to solve different real problems with his tool was critical to working out Bricklin’s ideas. Even focusing on the competition-the back of the envelope-was a way of concentrating his user focus.

Roadblocks to Creativity

Put the technologist in the users’ world-this is simple. So why aren’t more organizations doing it? Why do organizations put process in place which stifle creativity? In our work, we’ve seen how organizations create mind-sets which prevent them from being creative:

People separate “requirements gathering” from system design. There’s a myth in the industry that these are separate activities. They’re not. Creative design is an immediate response to recognizing a need and the knowledge of a technology that addresses that need. Separating these activities across organizations or people stifles creativity every time.

People think about product design as a list of features. They think about a new version in terms of what features to add. They think about their competition in terms of what features the competition has that they don’t. Teams we work with generally start with ideas for specific features, each responding to a single customer problem. These can only lead to incremental change and “feature creep.” Working with customer data to represent the customer’s whole work situation allows a team to respond with a synthesis that represents a wholly new approach. The management of one team was astounded at the change in the product design after working closely with customer data: “we thought you’d never get out of the box.”

People are scared to let their engineers talk to customers. Some are afraid of letting an unwashed engineer talk to an actual customer. Some think their engineers’ time is too valuable, or that talking to customers is someone else’s job, or that customers shouldn’t see ideas which are still tentative. (In one organization, product marketing blew a gasket because we were going to take paper mockups out to customers. Paper!) Eight years of experience taking engineers to the field shows that they are perfectly capable of working with customers. Customers love to talk about heir work and try out new designs-even when they are in paper (in the above case, they wrote an article in their internal newspaper about how their supplier came to share their newest ideas).

People are scared to work together. Bricklin combined multiple skills to see the opportunity. Products today are build by teams which could combine the diverse skills of their members to address a problem, but organizations structure themselves in ways that keep people apart. They break projects into separate units, assign “ownership” of units to individuals, define “architects” who can make decisions by fiat, define “hand-off” from one group to another, all so people won’t argue. This also means that each part of a system can benefit from the creativity of only one person. People can contribute to a common effort-if they are willing to tear down the walls and work with each other as people. One person we worked with asked the team to help him manage his tendency to dominate the meeting. This helped him to monitor himself, and gave the team permission to modulate his behavior.

People are scared of new ideas. It’s ironic in an industry with as much turnover as ours, but that’s what we see. Individuals won’t put new ideas on the table: “It’s different, and no one else is thinking of it, and maybe it’s silly after all.” Teams won’t allow themselves to pursue new ideas: “Management has all these expectations, and it’s not in our charter, and if they find out they’ll ax us.” Organizations won’t adopt new ideas: “We are in a competitive market and can’t possibly spare the resources to pursue something that is untried.” No one can take a risk if it’s not safe to fail, and new ideas are always risky. One team we worked with was so bound up in worrying about management, they arranged a meeting with their division head. He was incredulous: “How could you think I wouldn’t support this? I depend on you for new ideas and direction.” They went back to their team room energized and started planning how to drive their organization.

People don’t want to find out their ideas don’t work. This gets especially bad when deadlines loom: “If we go find out about the customer, either we will discover new things or we won’t. If we don’t, we’ve wasted our time and we have no time. If we do, then our design will be broken and we have not time to fix it. So let’s not find out anything.” Going to customers early helps engineers let go of existing ideas, and make room for new ones. This can be fast-we sketch out a part of a system, mock it up, test it with users, and interpret the results within 48 hours. We’ve watched any number of engineers return to their teams ready for new ideas after having watched users recoil from their pet idea in horror, or dismiss it as irrelevant.

People don’t respect their customers. Some groups are invested in being experts. Taking orders from users is hard for them. You hear it in their language: “they don’t know how to use the system right,” “they don’t even get trained,” “bad user on device,” “people can’t tell you how to use a new technology.” People make reasonable trade-offs between learning tools and doing their real work. The short cuts people put in place let them get their job done. People have figured out how to live their lives and do their jobs. One team started out sure that certain cards their customers used were clumsy and unnecessary. After studying the work, they discovered the cards supported eleven different forms of communication. This was incredibly efficient, and the team switched to designing ways to support the communication directly.

Learning to Be a Rainmaker

So, if current organizational practices don’t work for creativity, what will? It may not be possible to call down the lightning, but is there a thunderstorm and a bare hill to stand on?

First, put your designers in the data-put them in front of real users while they work. Second, get the fear out of the organization-make it safe to go beyond the accepted ways of thinking. Define creativity as part of your charter-which means that ignoring turf and challenging the business that is currently making money is also part of your charter. Third, define a clear process that starts with real customer data and takes you to a shipping product. Customer data is your touchstone throughout the process.[II]

Customer-centered design is old. It’s been tried and found to work for decades-for as long as people have designed products for others rather than themselves. What’s new is that customer-centered design is now being defined, formalized, and moved into standard practice. Teams which have taken on a customer-centered approach are opening new avenues to creativity.

Footnotes

[I] Our thanks to Dan Bricklin for telling us this story.[II] Our approach to customer-centered design is described in “Making Customer-Centered Design Work for Teams,” K. Holtzblatt and H. Beyer, Communications of the ACM, October 1993.

Published in IEEE Software 11.5 (September 1994): 106.