The University of California’s Office of Risk Services worked with IBM to design and build an Enterprise Risk Management System based on IBM’s analytics, portal, and collaboration software to manage risks and improve security. The move helped the school mine its databases to spot trends, such as a rise in pushing and pulling injuries at medical centers. Once that particular trend was spotted, officials were able to take action to prevent such injuries, like purchasing better equipment and launching training programs designed to limit them.

Information about the real and potential risks from departments and locations across the university system’s 10 campuses, five medical centers, laboratories, and field sites is aggregated for better insights and management, so that UC administrators can isolate recurring incidents and break the cycle of injuries and costs that ensue.

As a result, injuries have been cut by 39 percent—and the cost of the university system’s insurance has dropped by $167 million since the system’s initial deployment in 2006.

“We are now able to determine where we are the most vulnerable by creating dashboards so managers can access their data in real time,” says Grace Crickette, chief risk officer for the University of California. “They can target the key variables that influence outcomes and make changes to increase productive trends or intercede in operations that are having a negative impact.”

Traditionally, risk management programs in many sectors have relied heavily on quantitative analysis, but they have lacked qualitative measures and analysis. Using the advanced analytics and business intelligence capabilities of IBM’s Cognos software, UC’s Enterprise Risk Management System uses both quantitative and qualitative data to highlight emerging risks and assist with avoidance measures. The software also has helped university officials know the best way to deploy resources—money, people, and time—which is vital during a time of severe budget cuts.

But analytics alone are not capable of making change happen, Crickette points out.

“One of the benefits of analytics is it can be a great motivator of people,” she says. “Our tools can show us how [we] are performing in different areas, which helps to motivate people to implement good policies and procedures. Analytics for analytics’ sake is, without action, not very useful. But analytics that help drive action can be very powerful.”

—J.N.

In the last few years, IBM has acquired two companies to help it improve its offerings to the education industry in the predictive analytics space.

In 2007, IBM acquired Cognos to accelerate its information-on-demand business initiative, followed by the purchase in 2009 of SPSS.

IBM Cognos helps schools improve student performance, deliver on performance mandates, and improve financial performance, IBM says, by aggregating critical data and identifying trends.

For example, Cognos can help schools:

• Calculate curriculum costs, or identify good fundraising programs.
• Monitor student headcount and performance, program outcomes, school reputation, national agendas, and other key performance indicators.
• Share secure web-based information with all stakeholders.
• Manage endowments and recruitment through driver-based planning.
• Spot high- and low-performance schools or programs.
• Map enrollment to attendance and attendance to performance.
• Speed compliance reporting.

IBM Cognos is currently in use by more than 1,000 institutions of higher education and more than 530 K-12 school districts (representing more than 20,000 schools), IBM says. Additionally, 13 state departments of education and the federal Education Department use IBM Cognos.

IBM’s SPSS Modeler is another tool that schools have found of great benefit in predictive analytics. With Modeler, users can access a broad range of data, including data stored in operational databases and files, as well as unstructured data such as call center notes, eMail messages, Web 2.0 sources, and survey responses, which can be mined, modeled, and deployed via a simple desktop tool or via advanced client server architecture. This allows organizations to integrate predictive analytics into their everyday business processes, IBM says.

The data can be used to create predictive models in a way that doesn’t require programming, which means users can access information without waiting weeks for their IT department to respond to data requests.

SPSS Modeler provides deeper insights and more accurate predictions than simple analytics, because it uses all data assets to provide a complete view of a school district’s data, regardless of where these are stored.

For example, a community college in California uses Modeler to predict which students are less likely to return to school, helping faculty and administrators improve retention by providing appropriate counseling, financial aid packages, and curriculum offerings.

Partly as a result of these programs, the college ranks third among the state’s community colleges for the percentage of students successfully transferring to the University of California system.

—J.N.

The Houston Independent School District faced the same management challenges encountered by many large urban school districts: a wide variety of incompatible information systems, reliance on multiple manual processes for transferring data between those systems, and no way of looking across systems to find redundancies or business process problems.

The school district decided to improve operations in key organizations, consolidating business applications onto a common platform with a centralized, enterprise-wide administration database. The district worked with SAP, using its enterprise platform for education. HISD broke even on the project in about five years, with an anticipated 10-year ROI of 151 percent, officials say.

Broken down, this equals:

• More than $3 million in labor savings and warehouse cost reductions in food services;

• More than $5.5 million in savings on labor, paperwork order reductions, fewer vehicles kept in inventory, and improved maintenance in Fleet Operations;

• $5.7 million in reduced headcount and other savings in Administrative Operations; and

• $52.3 million saved in labor, bulk purchases and negotiated cost savings, inventory reduction, and eliminated waste in Materials Management.

To make such an implementation work, educators and administrators should take certain steps, HISD managers say. Like Kirk Kelly of the Hamilton County, Tenn., Department of Education, they suggest that stakeholders be engaged early in the process to help identify goals, including annual reporting requirements. They also suggest appointing a “project champion” at the highest levels of administration. Such a person can be valuable in making sure the project has adequate resources and that individual departments are on board and remain committed.

Team training is also an ongoing need. Work up a communication plan and stick to it to keep all stakeholders informed of progress or issues, HISD suggests.

—J.N.