By contrast, when there is an ineffective balance between formal and informal structures, it often shows up as an inability to manage bad ideas effectively. After a formal decision has been made to advance some ideas but not to pursue others, the company expends considerable effort to plan the next steps for the winners. But no one thinks actively of planning next steps for the losing ideas, to put them to rest, free up their supporting resources, and (ideally) identify and share any lessons or insights gleaned from the experience. (See “How to Shut Down a Project Gracefully,” by Rita Gunther McGrath, s+b, Summer 2009.)
When the balance is ineffective, the informal parts of the organization quietly kick into high gear to “defend the abandoned children.” Advocates of the failed project continue to spend time and resources on it, out of either inertia or false hope. The pride of seeing a personal project succeed overwhelms the pride that people might otherwise feel in supporting the group’s best assessment of what will succeed in the market. Efforts to move that investment to other projects don’t quite work. This tends to occur most often in “hero cultures,” where recognition often accrues to the “creative genius” who pushed forward despite resistance.
The formal parts of the organization tacitly reinforce this pattern. Metrics, and thus career advancement and recognition, are tied to expected results. People know that an innovator’s role in a successful product launch could be part of the justification for a promotion, raise, or merit bonus. And they know that no one is rewarded for failure. Thus, in very formal organizations, no one may explicitly shun a failed product, but employees have little or no incentive to celebrate or learn from failures. People try as long as possible to turn them into successes, and dispose of them only when it is clear no hope remains.
In the worst cases, the process leads to a vicious circle: Winning ideas aren’t resourced appropriately, and thus don’t achieve their predicted success; this undermines the credibility of the development system and thereby encourages more people to pursue their pet projects as mavericks. Meanwhile, since no one captures the lessons from the products and services that failed, companies then make similar offerings, with similar poor results in the marketplace, again and again.
Setting the Process in Motion
How, then, can companies integrate the formal and informal structures to support the best ideas, effectively abandon the weaker ones, and drive superior innovation over time? Successful organizations call on a few useful design principles when developing their innovation systems.
1. Weigh the balance of formal and informal mechanisms to meet your company’s current needs. Companies tend to vary by industry and market, and especially by their life-cycle stage. For example, startups tend to manage innovation very informally; they have less of an established business to run, and everything must shift quickly as they learn from their first efforts to bring an idea to market.
It is typically only after several years that leaders recognize that they need formal protocols, decision rules, practices, and metrics, applied across the board, to manage complexity more efficiently and ensure that people are treated fairly. After that, they may overcompensate with formal rules and processes borrowed from other companies. These tend to conflict with the informal rituals — practices that, because of their responsiveness and adaptability, were once core to their success — thereby diminishing the effectiveness of both.
Similarly, in larger companies, leaders sometimes become aware that they need to reintroduce some informality to become less rigid and more responsive. They, too, can overcompensate, creating overly flexible skunkworks that aren’t plugged into regular operations and that put forth maverick ideas that aren’t taken seriously by the rest of the organization. Some companies create a pendulum effect as they alternately emphasize formal and informal mechanisms, introducing inefficient change efforts with each oscillation.