Michele Sliger is often asked if the agile approach can be used for things other than software development. She gave the question some consideration and found the following example of a non-IT case of agility in action, which she highlights in this week's column.
A few years ago, I had the pleasure of seeing Rudy Giuliani, the former mayor of New York City, speak at the Rocky Mountain Project Management Symposium in Denver, Colorado. As the keynote speaker, Giuliani spoke about the importance of leadership in his management of the city and how he attacked problems such as crime and poverty. I was impressed by his approach and tactics and the agile way in which he accomplished his goals.
The first thing he did was identify the most important issue, because he realized that he may not have time to address all the problems he faced during his term in office. As an elected official, he had a short window within which to accomplish positive change, so he created a prioritized backlog of items he needed to address, and crime reduction was number one.
Working with the police department and university fellow George Kelling, he decided on a strategy to reduce crime: the "broken windows theory." This strategy originated from a paper written by James Q. Wilson and George L. Kelling, who said that by focusing on fixing small things like vandalism, further neighborhood damage is discouraged and major crime is prevented from developing. Giuliani implemented this approach, focusing on things like subway-fare evaders, public drunkenness, and vandals.
At this point, Giuliani had followed several agile practices: He created a prioritized backlog of work; he collaborated with the team (police and other advisors); and he defined a vision and an approach to be implemented in an iterative fashion.
In order to determine how well the broken windows project was progressing, they measured crime in every borough every day. At first, they were just looking to see if their attempts were resulting in decreases in crime, but then they began to notice patterns. By checking every day, they were able to shift the police force around to areas and times where they were most needed. They were constantly inspecting and adapting based on the metrics they were gathering. The end result was a reduction in crime by seventy percent by the end of Giuliani's tenure--and Giuliani stated that crime continues to drop every year as the police department continues to follow this successful method.
So, two more agile practices helped the team to be successful: checking in on the progress every day and inspecting and adapting each iteration. While they weren't viewing burndown charts, they were reviewing metrics that showed what was working and what they still had left to resolve. These metrics allowed the team to track improvements and make appropriate changes. You'll see that the metrics in the next example needed a bit of work before they were able to foster change.
The second item in Giuliani's backlog was welfare. He called this a problem of hopelessness and was compelled to help these individuals on welfare find good jobs that would restore feelings of productivity, accomplishment, and self-reliance.
The first thing he discovered was that the metrics that the welfare offices were tracking were inducing behaviors in staff that did not end up helping the welfare recipients. The primary measurement of success of a welfare office was based on how many people were approved for welfare that month. The more people the office had signed up for welfare meant the more productive that welfare office was, and the more attentive it was to its people's needs. The metric drove the offices to treat the symptom, but not the problem.
Giuliani changed this metric to begin measuring how many people the office had found jobs for and how good those jobs were. He put programs in place to assist with job training and placement. He found that what is measured changes the thinking of those people performing the work. The mission of the welfare offices changed, and, with constant inspection of the metrics and adaptation of the programs, the number of welfare recipients was reduced.
Giuliani went on to discuss effective leadership--a critical component of these successes. He stated that leaders must do the following things: have a vision and execute and accomplish. With the vision alone, you're just a philosopher, not a leader. Giuliani also stated that leaders need to be optimistic in their problem solving, but not foolish. Treat the problem as a challenge, but don't kid yourself and others with falsely optimistic statements or "spin." Instead, be prepared to face and deal with the brutal truth of the issues.
Resolving these challenges called for input from everyone and resulted in what he termed "blue sky sessions," where anyone could suggest any solution they wanted--no matter how crazy it sounded. Giuliani affirmed that there has to be a level of trust for this to work, and that it is the leader's responsibility to establish this trusting and collaborative environment.
What struck me all throughout his speech was the clear use of agile practices used to tackle these issues: working in a timebox, using a prioritized backlog, defining and measuring success, inspecting and adapting, effective leadership, and collaboration in an environment of trust. Rudy Giuliani used an agile approach to managing a city and was incredibly successful!
For more information about Giuliani's use of the broken windows theory (which is also mentioned in the book The Pragmatic Programmer by Andrew Hunt and David Thomas).