EXECUTIVE FORUM
Successful Innovation Through Artful Process
Few deadlines in business are as unforgiving as a play’s opening night at the theater. The tickets have been sold; people will turn up; the curtain will rise. This launch can’t be postponed. Good theater companies routinely create value under pressure, plays that execute precisely again and again, incorporating innovations but finishing within 30 seconds of the same time every night after roughly two and a half hours—despite using none of the techniques a business would need to achieve such rigor. What business do you know that can innovate so reliably on such a deadline?
Can businesses learn something about innovation from artists? People not involved in the arts often think that artists “have it easier” than businesspeople. Some managers acknowledge that parts of business are more art than science, but they see this as a problem, not an opportunity. In business, labeling some task as an “art” implies that it is not quite trustworthy, that it resists reasonable description, that it can’t be taught or learned. Business leaders often look down their noses at leaders in the arts—especially leaders in nonprofit arts organizations—thinking them somehow less able because they don’t make “real” products.
Our research into the practices of successful collaborative artists suggests that these beliefs and attitudes are mistaken. Managers can learn from artists, and from the people who manage (that is, direct) them. This is especially true for the work we now call “knowledge work.” It’s a feature of such work that the people managed are often more talented, in at least some value-producing capacity, than those who manage them. In work such as high-tech strategy making and agile software development, practitioners have begun to embrace processes similar to the work of collaborating artists. Managers acknowledge this linkage with increasing frequency; for example, Eric Schmidt, CEO of Google, has said of the work done by creative teams (“Googlets”) in his company, “There’s little doubt in my mind that our Googlets are engaged in collaborative art.”
Making the Business-Arts Connection
Our unusual collaboration (a business school professor and former manager working with a theater professor and professional actor) followed a casual conversation. While we were explaining our interests, software development and play making, to each other, we stumbled upon a common concern: managing skilled people engaged in creative activities. We noted similar patterns and structures in our approaches to this work. Some recent methods and ways of talking about them in software development, especially in the so-called agile community, seemed almost identical to the methods and language of the theater.
During the four years of collaboration that followed, spent observing artists and business innovators in one venue or another, we spotted many things that managers of knowledge workers could learn from directors of artful workers. Here we’ll focus on four characteristics of artful work apparent from our observations:
We’ll address each of these in turn.
Reliable Emergence
What is emergence? Abigail Adams, artistic director for the People’s Light and Theatre Company in Malvern, Pennsylvania, describes emergent process in the theater: “My directing style is based on what the actors are bringing to rehearsal, and on making what you make out of those particular actors then and there. You discover the play throughout the process … sometimes, roughly, I know what the journey is. Sometimes I don’t … usually we try a scene or a moment so many different ways that the right choice makes itself known.… We work until we find that.”
During rehearsals of the People’s Light production of A Streetcar Named Desire, by Tennessee Williams, this sort of emergence led to a breakthrough in the staging of that play. In that event, the unruliness was physical horseplay, not in the script, between two characters, which created a tension and violence on the set that infected other actors and brought new energy to the play. The play’s director didn’t see at first how the disturbance would fit into her short-term plans. But she encouraged productive unruliness and trusted that the process would deliver emergent value. In the long term, the tension and violence had a definite place in what she was doing.
This degree of trust in an emergent process might seem risky to business managers, but we should remember that opening night is as relentless a deadline as any they face. And, as management researchers Henry Mintzberg and Alexandra McHugh have noted, this kind of emergence can be valuable in fast-changing businesses (which they call “adhocracies”):
“Strategies grow like weeds in a garden.… In some cases, many different actors converge around a theme, perhaps gradually, perhaps spontaneously; or sometimes senior managers fumble into strategies.… To manage this process is not to preconceive strategies, but to recognize their emergence and intervene when appropriate.…. To manage in this context is to create the climate within which a wide variety of strategies can grow … to watch what does in fact come up and not be too quick to cut off the unexpected.”
When you rely on emergence, you can’t typically aim directly at a target and be confident of arriving at the best outcome. Heading for a preconceived destination can keep you from arriving somewhere better—indeed, it can prevent you from noticing that better place. So, despite the management maxim, “If you don’t know where you are going, any road will do,” the very essence of an artful approach consists in managing successfully when you don’t know (exactly) where you’ll end up.
How is this relevant to business? Many people are surprised to learn that managers at Sun Microsystems, a pioneering Web technology firm, did not immediately recognize value in Web technology when they first encountered it. Web technology bubbled up from the bottom at Sun. The engineers noticed it first, played around with it, and tried to bring it to managers’ attention. Management actions such as instituting a financial penalty for using Web browsers (to reduce network traffic) beat them back again and again. The unruly engineers persisted, though, and eventually prevailed. On what managers and engineers now call “The Day the Universe Changed,” Sun realized that the Web, in combination with a new programming language called “Java,” created new strategic possibilities for the company. Sun redefined itself. A new firm emerged.
Sun’s managers embrace this emergent management style. Despite the management actions that discouraged the engineers, Sun leaders have intentionally fostered a culture in which the engineers persist when they have a good idea, even if it doesn’t fit into short-term plans. Says one senior manager of this sort of employee unruliness: “We don’t necessarily stop that—he might be working on something that leads to the next big product.”
The Shape of Artful Processes
The industrial processes most used by managers are quite different from the artful processes used by creative artists. But the emergent processes employed by successful collaborative artists have a number of advantages over less flexible industrial processes. In the future, mastering them will be an essential task for any leader who wants to succeed.
It’s useful to observe that industrial processes tend to be linear, or sequential, while artful processes tend to be helical, or iterative. Figure 1 compares a classic sequential industrial process for making cars with an artful iteration for making software.
FIGURE 1. COMPARING THE SHAPE OF INDUSTRIAL AND ARTFUL PROCESSES.
The artful process is modeled on the way Texas software company Trilogy manages its “fast cycle” development and consulting work. Rather than step sequentially through phases of planning, design, and operation (as in an industrial process), Trilogy loops rapidly through each of these activities again and again, each time generating a prototype—a product, however rough, that can be experienced directly and to which makers and customers can react. Such methods, described collectively as “agile software development,” are growing in popularity.
This process shape can also be found increasingly elsewhere; in fast-changing industries, strategy making takes on this shape as workers refine and often completely reconceive products and plans. This is also the shape of many collaborative arts processes such as theater rehearsal: actions are tried, examined, and tried again, each time reconceiving old outcomes into something new.
There’s a good reason why these diverse processes can share an iterative shape, and it has to do with cost. The iterative approach works only when iteration can be cheap and rapid. For an iterative process to make business sense, you need to work quickly enough to generate a large number of prototypes so you can be confident you’ll build atop old ideas and achieve sufficient value. Iterations must also be inexpensive; if each iteration costs a lot, artful process becomes economically unsound.
We need to worry about two kinds of cost. First, reconfiguration cost: rearranging the process to produce a different outcome. Second, exploration cost: the cost that must be endured when a prototype doesn’t work. A simple example helps distinguish between these types of cost: If an airplane maker decides to make a new kind of airplane, the costs of rearranging the factory to do that are reconfiguration costs. If the prototype airplane takes off and then crashes, the costs associated with the crash are exploration costs. Both these costs must be sufficiently low for an emergent process to make economic sense.
What factors result in cheap iteration? First, materials: if you make things with costly metals (for example), they may be difficult to bend in different ways (reconfigure) and expensive to replace if you ruin them (while exploring). Ideas and symbols, such as computer code, are inherently easier to unbend and replace when ruined. The materials of theater rehearsals—or budget meetings—are people; they are very flexible as they work out new ideas.
But the inherent malleability of knowledge work materials is not usually sufficient in itself to allow cheap and rapid iteration. Supporting technologies must play a role in reducing the cost. In software development, for example, automated application development and testing technologies allow for rapid-fire generation of prototypes, reducing reconfiguration cost. Configuration and version control systems keep bad prototypes from reaching customers and make it inexpensive to roll back to a previous version when necessary, reducing exploration cost. Only recently are these technologies getting good enough to allow truly artful process and reveal the convergence in process shape between artful processes such as theater rehearsal and certain business processes.
Openness to Uncertainty
The traditional industrial response to uncertainty is to protect processes from it. A classic example from industrial (service) management: McDonald’s requires that the frozen potato sticks delivered to its restaurants be exactly 9/16 inch wide. Suppliers comply, or else. But not all processes can be so well protected from uncertain inputs. Raw potato processors have a different problem: potatoes cannot be grown in identical shapes (not yet, anyway), so raw potato processing must adjust to inherent variation in the shape of its inputs.
Many industrial processes have difficulty protecting themselves from uncertainty. One famous attempt to automate meat packing foundered because cows won’t grow to specified shapes and sizes. A similar problem may have contributed to the demise of Digital Equipment, a computer company that was on top of the world in the late 1980s. The uncertainty came not from materials input but from large, unpredictable customer orders that played havoc with efficiencies. Digital’s efforts to protect operations struck some observers as odd; customer orders are surely the most compelling reason to be responsive. The time Digital spent trying to protect its process from uncertainty might have been better spent creating ways to remain open to uncertainty and still succeed.
An artful manager keeps processes open to many sources of uncertainty, by design. Uncertainty—the unexpected action of another player, say, or the reaction of a customer or audience member—fuels artful process. An unexpected event triggers an unpredictable reaction. Others react in turn. Those reactions trigger further reactions. The ripple effect of interdependency produces novel outcomes. According to a (true) story often retold at the People’s Light, Abigail Adams once raised a performance to a new level by asking an actor to “walk the other way around the table.” The suggestion had no apparent connection with the play’s action. For the actors, the move was a change, something to be absorbed into their work. But when they did the move, it led to other moves and countermoves, each prompting collaborative adjustments and reconceptions that, in combination, transformed the emerging play. As at Sun, the important leap forward arose from moves and countermoves triggered by uncertain events, not from intensive planning and execution of the plan.
Failure as a Step on the Way to Valuable Innovation
Early in this research, thinking about cheap and rapid iteration as a way of working, we found ourselves talking about the concept of failure. One of us, the business guy, observed that an iterative work cycle must include many failures on the way to success. The theater guy agreed—but resisted the term failure. Failure isn’t the right idea. In rehearsal the iterations all interact with each other. This run-through is the main material for the next run-through. Each trial is a necessary step on the way to what’s good, essential to the final success. To call an essential step toward success a failure merely tortures language. What’s more, the word failure applied to routine work could poison the quality of group work essential to rehearsal.
Then we got to thinking about IDEO, a leading product design firm that employs an iterative approach, and failure came up again. We agreed that IDEO’s work process was an artful one, but they talk about failure all the time, saying things like: “Fail often to succeed sooner.” When professors use IDEO as an example at the Harvard Business School (HBS), they note the difference between a “failure” and a “mistake.” In the HBS view, IDEO cherishes failure because it generates new information. But a failure that doesn’t generate useful new information is called a mistake. Touch a hot stove and burn your hand—that’s a failure. Touch it again and burn your hand again—that’s a mistake.
This resonates with many MBA students and even executives, but makes no sense to artists. The distinction between failure and mistake imposes an unreasonable limit on exploration. Though artful making is reliable and even efficient, it has little use for the efficiency of rules like “Avoid touching a hot stove twice.” Touching the stove twice (or ten times) may be what’s needed to break up a creative logjam. Just as an athlete may need to execute the same painful movements (lift the weight, run the interval) over and over on the way to new levels of performance, so you may need to make the same mistake many times on the way to an innovative leap. Burning your hand is a small price to pay for a good idea. The apparent imperfections in an artful approach are a primary source of innovation, a vital part of making something truly new. Artful creators respect and even cultivate “mistakes.”
Reliable Innovation and the Knowledge Economy
The power of industrial management derives from standardized process, from interchangeable parts (and people), and from the resulting economies of scale. Businesses that could reliably and cheaply do similar things again and again could achieve competitive advantage. The not-yet-fully-tapped power of management in a knowledge economy will have a different source: it will use past ideas as material for new ideas. In the future, businesses that can reliably and cheaply do things differently from what others did before will achieve competitive advantage. To be more valuable, practical disciplines must now be focused not on enforcing compliance with preconceived process but on creating emergent processes of iteration to make something brand new.
It shouldn’t surprise us that we are only beginning to understand how to do this. It took a long time to master physical work. Figure 2 shows U.S. employment statistics at three points between 1850 and 2010. In these numbers we can see the shifting basis of the American economy. Between 1850 and 1950, we learned a great deal about farming, and the dramatic shrinkage in agrarian employment numbers show that management’s industrial expertise yielded higher productivity. As many have observed, a similar transition is now under way, described by a comparable shift in employment statistics. By 2010, based on current estimates, most workers in the U.S. economy will be found in neither agrarian nor manufacturing jobs. Many of them will be doing knowledge work.
FIGURE 2. AGRARIAN TO INDUSTRIAL TO INFORMATION ECONOMY
Source: Ronald E. Kutscher, “Historical Trends, 1950–1992, and Current Uncertainties,” Monthly Labor Review, 1993, 116(11); and various reports from U.S. Department of Labor, Bureau of Labor Statistics <www.bls.gov>.
As we began to manage manufacturing, our first instincts were to apply agrarian management frameworks to factories. It took us decades to sort out which ideas were transferable from agrarian to manufacturing settings and which were not. As we begin to manage knowledge work, we begin, not surprisingly, by applying frameworks that we use successfully in factories. It may take us decades to sort this out as well.
Managing Artfully
Several ideas stand out when we view successful business innovation through an artful lens: the sense of an emerging product as something better than anything a single person could have preconceived, the interdependency of iterative collaboration, the unpredictable results that in retrospect seem inevitable, an understanding of apparently wrong choices as steps along the way rather than setbacks.
When you create something entirely new under time pressure and right on cue, your artful accomplishment is not so much the product created, but the team that can make that product. As we move forward into the knowledge economy, the path we must take will become clearer. To see a preview of what the future holds for successful managers, find a way to look in on collaborative artists at work. We must develop new reflexes and skills if we are to see the future through this artful lens. Some who have devoted effort to developing their (often successful) industrial management skills will not at first welcome this message. Some, though, will learn to translate their management success into a new kind of work for the new century. 
Rob Austin is a professor of business administration at Harvard Business School. Lee Devin is professor emeritus of theater at Swarthmore College, and a dramaturge at the People’s Light and Theatre Company. They are the authors of “Artful Making: What Managers Need to Know About How Artists Work,” recently published by Financial Times Prentice Hall.
