That's exactly what Tim Blevins did as soon as he became CIO of the Kansas Department of Revenue (KDOR) in 1997.
"This was a department that had never really dealt with ways to see data on an organizational level," said Blevins, who was brought in to lead the Topeka-based agency with 1,100 employees and 124 IT staffers.
Typical of many public-sector financial organizations, the KDOR had many stand-alone applications managing different types of taxes with no common data architecture, data reporting or analysis tools. "We had to gather data from all these places and then write COBOL reports to merge it all together," Blevins recalled. It took IT staff time to create and write these reports, and it pitted the end-users against IT, because the users could never get enough access to data in a timely fashion. "Something had to be done," he said.
Because the KDOR was already in the process of replacing legacy systems, Blevins made a strong pitch to former Secretary of Revenue John LaFaver to employ business intelligence (BI) tools from Cary, N.C.-based SAS Inc. He had been involved in BI implementations in other state departments and saw the transformation data warehousing and analytical tools could bring.
Blevins found a partner in General Counsel Shirley Sicilian of the Multistate Tax Commission, whose analysts had been starved for data in their efforts to quickly field requests from the Legislature for information involving tax bills.
Since that time, the KDOR has gradually integrated data from all its systems, and developed an enterprise reporting and data warehouse environment across all four of its divisions -- taxation, motor vehicles, property valuation and alcoholic beverage control.
"I am convinced we could not do our business without this toolset," Blevins said.
A New Generation
Blevins is part of a new generation of public-sector CIOs who are aggressive about eliminating data silos and employing business intelligence software to enhance their organization's ability to act on information requests. BI applications include data mining, online analytical processing and the creation of executive dashboards. Their common goal is to help employees access enterprisewide data and to provide tools that enable data analysis.
With many governments legislating performance standards, BI tools are often used to track programs' progress accurately. They also let agencies see their data in new ways. For instance, Census Bureau data can be sliced and sorted to create more intelligent views of household and geographic data. State employment departments can offer a more comprehensive view of data about employment sectors and unemployment.
Human resource executives use BI tools during negotiations involving union contracts. They can speed up the negotiation process by running what-if scenarios based on different union requests.
Many of the current applications are financial. For instance, the Department of Defense (DoD) procurement officials who use query and analysis tools can discern that Army, Navy and Air Force bases in the same state are all purchasing truck tires from different vendors. By consolidating those purchases, they can leverage the department's purchasing power to negotiate better deals, said Gino Magnifico, deputy program manager of the DoD's Standard Procurement System in Fairfax, Va. The next challenge, Magnifico said, is to use BI tools in ways that feed meaningful reports on the data to senior executives.
A Window on Government Data
BI tools empower government employees, but some agencies and municipalities also use them to open a window on government data for citizens.
"Part of the move toward improved e-services is giving citizens access to data that can affect their decisions," said David McClure, research director of Gartner's Global Public Sector research group. Albuquerque, N.M., is a good example, he said.
The city, which had been using Cognos' BI tools to extract data from its legacy financial systems for years, decided in 2002 to use Cognos to offer citizens reports on statistics previously unavailable to the public.
City IT staff built a Web-based analysis and reporting environment that offers both neighborhood-specific and citywide crime statistics by individual crime and trends over time. Using Cognos PowerPlay Web Explorer, users can look at everything from bribery to kidnapping by neighborhood, time of day, day of the week, month or police beat.
Real-estate agents use the crime statistics quite often, said Chris Framel, an applications group manager for the city. Moving beyond crime stats, Albuquerque now offers citizens air quality data in a Cognos report, as well as analyses of campaign financing reports.
The city also uses Cognos NoticeCast to notify citizens when their water bills are due. "Using Analysis Studio, they can see reports on trends in their own water usage, which is an important issue in New Mexico," Framel said, since there's scarcity of water due to arid conditions.
"Albuquerque is a great example of being more proactive with its citizens, sending out messages when taxes are due, rather than them having to log on to a Web site to find data," McClure said. "They're being very innovative in terms of pushing data out."
Realigning the Data
Despite these promising examples, governmental entities have generally lagged behind their corporate counterparts in taking advantage of BI tools, according to vendors and analysts. Because they are frequently dealing with multiple legacy systems, a lot of their data is not structured in an appropriate manner, so it takes a reorganization and realignment of data to answer the questions they want to answer, said Robert Latham, director of U.S. government systems engineering for SAS.
Despite these obstacles, Latham said BI has become an increasingly hot topic in the public sector over the past five years.
According to The Challenges and Promises of Business Intelligence, a Gartner survey of 218 government CIOs published in March 2006, BI applications ranked fifth among their top 10 technology issues for 2005, and respondents predicted it would be their No. 2 technology issue in 2008, ranking behind mobile work force enablement.
Although Gartner doesn't break out its estimates by industry sector, the company predicts the overall market for BI software will grow at a compound annual rate of 6.5 percent over the next five years, making it a $3.2 billion market by 2010.
Gartner's McClure noted that in the private sector, BI tools are used in increasingly sophisticated ways to spark business transformation. Using predictive modeling, decision-makers are moving beyond examining historical trends to look at future scenarios and possibilities.
"We're not quite there yet on the public-sector side," he said, adding that there's much pressure on state and federal governments around accountability and metrics.
In the public sector, BI is most often used to aggregate data from disparate systems -- pulling together information for internal managers and oversight agencies. McClure said the challenge for public-sector CIOs is to help their organizations take advantage of the analytic power of BI tools to get beyond things like measuring city employee cell phone usage to actually improving processes. For instance, are traffic signage changes actually reducing accidents? Are new programs geared to reduce repeat offenders actually having the intended impact? "They should move toward spotting trends so that improvement is proactive," McClure said, "not reactive."
Kansas Takes Action
Some government organizations are taking full advantage of BI tools for trend analysis and projections. Blevins said predictive modeling, previously a weakness of his department, is now one of its strengths. For instance, if the Legislature is considering a change in withholding tax, its research analysts can predict within 24 hours how the change will affect individuals in different income brackets, as well as the state's revenue over a five-year period.
Another strategic aspect of the deployment was to change the relationship between IT and other KDOR staff.
"The strategy was to push the data analysis and reporting functions down to the desktop level," Blevins said. This allowed the IT department to evolve from a programming staff that spent time writing reports to one that manages the systems and their enhancements.
Because the analytical tools take time to learn, Blevins explained, BI implementation at the KDOR had to be gradual. "It definitely is a stair-step evolutionary process. They have to start at the bottom, and they can't climb it all at one time," he said. "There's a need for training at the desktop."
But he said the environment they've created turns data management, analysis and reporting over to the business analysts so they can support themselves.
The number of self-service reports generated by staff analysts has climbed gradually from 500 in 1999 to more than 6,000 in 2005, he said.
Avoiding the Pitfalls
CIOs have the potential to make a huge difference in their organizations by introducing BI tools, but they should not take on too much responsibility or ownership for BI implementation, warned Gartner's McClure. "It's a tough balancing act for CIOs," he said.
The CIO may drive that discussion of the goals, he added, "but it's a nuanced position they have to take -- when to step into the light or back into the shadows."
If the CIO is seen as owning the project and it doesn't deliver as expected, he or she is left holding the bag. "That's accepting too much of the risk," McClure added.
Another issue for CIOs is that their own IT staff may not have a clear understanding of the issues on the business side. The IT staff has to walk in the business side's shoes about accessing data and understanding the pain points. A lack of business skills in the CIO's office can undermine a project as well.
To increase the likelihood of project success, some vendors promote a "BI competency center" concept. Sometimes large organizations have several BI projects going in different departments with no standards or centralized understanding about tools and training, said SAS's Latham.
"Sometimes each department builds its own BI silo and the tools aren't shared. Instead of getting rid of silos, you just get taller silos," he said. The competency center is a way to enhance knowledge sharing, Latham added. "We usually suggest it's in the CIO's office. It could be a physical or virtual group with people from other departments. The idea is to allow the organization to spend more time focused on managing business needs and analytics, not just the technology."