Halfway into that decade of data, MGI has looped back to produce an extensive 136-page overview, The age of analytics: Competing in a data-driven world. While the report focuses largely on uses of data in the commercial world, the authors say they learned much about the state of data adoption in the public sector.
The story of government data remains largely one of untapped potential, they say.
“While lots of good work is being done with data, we are estimating only 10 to 20 percent of the potential has been realized. We are out of the starting gates but this is a long, long race,” said MGI Partner Michael Chui.
In talking about unrealized potential, Chui has in mind the vast volumes of data that already exist in government hands and are not being put to good use.
“Many times the data is sitting there and either not being used or being used for purposes other than to optimize operations,” he said. “Nowadays you cannot execute a public assistance transaction, or a procurement transaction, or almost any other transaction without data being generated. It just happens. Much of it is not captured and much of it is never transmitted."
Often information is underutilized. Government may use data to schedule the maintenance of trucks, but it could be using that same information to improve routing and scheduling of those vehicles. Data that tracks tolls on a bridge could be used to inform bridge upkeep schedules as well. “Sometimes we do need more data, but a lot of the time we just need to make better use of the data we already have,” Chui said.
Other studies bear out the general impression that government, while interested in the promise of big data, has yet to become fully engaged.
A survey by ViON Corporation, for example, found that while 92 percent of federal chief data officers say their agency uses big data, 58 percent grade their agency’s data strategy at a C or below.
At the same time, data trends are evolving in the private sector that could have an impact on government. Machine learning in particular could drive resource allocation choices for the government agencies always struggling to make their numbers work.
“If you are trying to optimize how you spend tax dollars across different programs, you can pull data from history and use that to determine what sorts of programs create the most value,” Chui said.
In the private sector, such data techniques are used to drive personalization — a trend that has some government IT leaders nervous. But the trend is one that Chui argued should be embraced.
“In the private sector, marketing organizations want to look at different consumers differently, to customize what they offer," he said. "That is something that is outside the mindset of public servant who, for fairness reasons, wants to treat every citizen the same."
Rather than frame it as an issue of fairness, government could approach big data as a means to enhance service levels. “If you knew exactly what was most valuable and what would help that citizen the most, then you could provide the best service to that specific citizen. We can use machine learning to do that,” he said.
While some public sector entities have made headway on data, others have been slower to come around. In general MGI found that beyond the simple mechanics — the databases and analytic software — government entities struggle to adapt their internal processes to the challenges of data-driven management.
“Even after you realize that you are getting great insights, you still need to change the organization to make use of those insights. That is hard work and it isn’t necessarily about the technology,” Chui said. “You do need the technology and you need the data scientists, but ultimately you need to change the way things are done.”
This struggle “is not unique to public sector,” he said. “Change management in the private sector is also hard, but it can be harder in government. The practices may be written into legislation or there may be regulations, and all that can slow the pace of change.”
In addition to improving its own processes, and opening data for general use, government has yet another role to play in driving the data revolution. As the prime movers in education, public-sector entities could be doing more to build a data-savvy citizenry.
“Government can help to make sure we have people who have these skills: Not just to produce data scientists, but to build data skills for all of us as citizens,” he said. There needs to be a greater focus on metrics and analytics, a better understanding of averages and probabilities, beginning at an early age.
“Even first-graders can flip coins and roll dice," Chui said. "Pulling the education lever will help all of us to be better citizens."
