Failing the Right Way
Government innovators need to encourage a culture of innovation in which failure is not a four-letter word.
"Ever tried. Ever failed. No matter. Try again. Fail again. Fail better." -- Samuel Beckett
Innovation is about experimentation; experiments often fail. Encouraging innovation requires a culture where it is safe to take smart risks -- even when they don't work out.
Traditionally in government, however, a "can't afford to fail" environment prevails. Governments often engage in large-scale planning processes with rigid program mandates that remain stubbornly static in the face of unexpected obstacles. Projects become too big to fail.
Fear of failure and wariness of funding untested approaches pose significant challenges for innovation. Political realities drive this risk aversion -- failures make headlines while successes are spread like butter across too much toast, yielding often invisible improvements for millions of constituents. The fact that many innovations don't bear fruit within the political election cycle doesn't make things any easier.
The $64 million question for government innovators is how to encourage a culture of innovation in which failure is not a four-letter word while also ensuring accountability and protecting taxpayers.
Fail small, fail fast. Many governments suffer from big-ticket failures with projects that run too long and cost too much money. Successful innovators, in contrast, adhere to the motto "fail small, fail fast." The idea is to fail quickly if you have to, learn from the experience, and move on to the next big idea. This approach entails a willingness to build small prototypes and pilots, conduct staged rollouts, and build smaller failures into the process in an attempt to avert larger failures at a later stage. By allowing their people to fail small and fail early, innovators can frequently detect and correct errors before they spiral out of control.
Governments can take similar managed risks by testing potential programs. New York City Mayor Michael Bloomberg mobilized private funds to launch innovative pilot programs to transform the city's underperforming public school system before spending public money on a citywide rollout. One example is the Empowerment Schools program, in which schools sign performance agreements committing them to high levels of student achievement. In return for this commitment, schools receive greater local autonomy over their operations.
By targeting a subset of early adopters, Bloomberg and Joel Klein, his school superintendent, receive feedback that allows them to improve their programs in advance of wider adoption.
Use partners and networks to prototype new ideas. The Bloomberg model points to an effective tool governments can use to reduce innovation risk: partnering. The Science & Technology Directorate at the Department of Homeland Security, for example, uses grant money and develops formal relationships with academia, state and local governments, other public agencies, and private organizations to test and prototype emerging technologies. The CIA created a nonprofit, In-Q-Tel, to find and deliver technological solutions to the agency for a wide variety of needs, including data mining, strong encryption, and the ability to comb the Web for valuable information. In-Q-Tel provides seed capital to start-up companies to develop promising new technologies that could generate IT solutions for the CIA.
Borrow proven ideas. The public policy process offers any number of opportunities for crushing innovation. New ideas can fail to become law, or they can be botched during implementation, and of course, seemingly good ideas sometimes prove faulty when put into practice.
For all of these reasons, one of the best ways to reduce risk when launching an innovation is to borrow a proven idea from elsewhere. An idea can be argued ad nauseam, but results are difficult to dispute. "We borrowed and adapted several ideas from other states and cities during my tenure as governor," explains former Pennsylvania Governor Tom Ridge. "It worked well because we could see what worked elsewhere and why, and then tailor and deploy those best practices in a way that fit our state's unique needs and circumstances."
Manage risk through a portfolio approach. Vivek Kundra -- formerly the dynamic chief technology officer for Washington, DC, and now the new federal CIO -- created the District's OCTO (Office of the CTO) Labs. The OCTO functions like a portfolio-management company, funding high-performing projects and killing others that are unlikely to deliver.
Each project is treated as if it were a stock and evaluated on the basis of its management team, customer satisfaction, and how likely the project is to finish on time and within budget. Kundra's hope for this approach is that the aggregated wisdom of employees working in the trenches will increase the likelihood of funding ideas that actually work.
This is one way of changing government's risk-averse culture. "I think controlled risk is very healthy," explained Kundra. "That's why I created OCTO Labs. The idea was that we would throw hundreds of ideas on the wall and even if five of them survive, they will be transformative."
"Progress always involves risks. You can't steal second base and keep your foot on first." Embedded in this good advice from writer Frederick B. Wilcox is something deeper. The word "risk" usually brings big risks to mind -- the equivalent of trying to hit a home run every time you go to bat. Risk diminishes, however, when runs are produced one base at a time. Governments, too, need to think of ways to make one-base-at-a-time progress by providing incentives for taking chances, and at the same time, creating mechanisms for risk management that prevent the fear of failure from trumping the desire to innovate.
Shalabh Singh is a manager at Deloitte Research. This article is adapted from Eggers' and Singh's new book, The Public Innovator's Playbook, co-published by Deloitte Research and the Ash Institute for Democratic Governance and Innovation.
Join the Discussion
After you comment, click Post. You can enter an anonymous Display Name or connect to a social profile.