Business Analytics Turns Data Into Intelligence
By Tony Kontzer
Judging from recent headlines, it would be easy to conclude that the real value of business analytics software is in helping organizations contend with their so-called big data problems. However, many organizations don’t yet have a big data problem.
In fact, aside from the biggest players in the most data-intensive industries, most organizations are simply trying to put whatever data they have to work. And thanks to the fact that powerful analytics tools—formerly the domain of huge enterprises—have become available to businesses of all sizes, any organization can now turn its data into actionable intelligence.
Take the city of South Bend, Ind. Like many American cities, South Bend has an aging infrastructure that has sometimes created significant issues for civic decision-makers. For instance, up until 2005, the city's 1950s-era combination water delivery and sewer system was plagued by as many as 30 overflow incidents a year, during which debris dislodged during storms would collect at the "on-ramps" to the main sewer arteries, says Gary Gilot, president of South Bend's Board of Public Works.
Sewage would overflow damns, back up into homes and cause other damage, resulting in frequent threats of multimillion-dollar fines from state and federal regulators. The only way for the city to minimize those incidents was to have field workers physically observe each of the system's 36 on-ramp junctions and, if possible, remove the debris. It was an inefficient approach that didn't provide sufficient public health and environmental safeguards.
But then Gilot came across two technologies he thought could help: a real-time sewer-monitoring system being developed by entrepreneurs at nearby University of Notre Dame and IBM's Intelligent Operations Center, an analytics-based software package that's designed to help cities monitor and manage services. The city installed credit-card-sized computer sensors at 116 locations in its sewer system and started feeding data from those sensors into the analytics software.
Once the technology was in place, it was as if the city had a whole new sewer system, without replacing a single piece of pipe. "We now have the smartest sewer system not only in America, but perhaps in the world," Gilot claims.
Gilot and the city's sewer technicians and operators suddenly had Web-based access to a tool that works like a GPS traffic report, using color codes on a map of the sewer system to indicate where potential problems are brewing and allowing sewer staff to prevent backups and overflows before they happen.
"What I found from the data analytics is that I could have room in my downstream pipes, but have overflow in my upstream pipes due to weather," says Gilot. With a system of real-time sensors in place, the city can now ensure that when rainfall is heavier in a certain part of town, the resulting sewer flows can be routed to areas where the pipes aren't getting as much use—much the way networks route Internet traffic.
As a result, the city has eliminated 95 percent of the overflows it once endured. Better yet, the entire platform cost the city $6 million, a far cry from the $120 million in sewer improvements it faced if it had opted to make traditional fixes.
Meanwhile, the analytics setup is delivering far more additional value than originally anticipated. Water distribution system operators are using the system to monitor when they flush out iron and manganese buildup, a process that can result in overflows if it's not closely controlled. And the mayor of South Bend has a Web-based dashboard view that lets him keep tabs on the state of the water and sewer system.
In addition, Gilot says that over the next year, the city will start using data the system contains about water service shutoffs to inform police about the locations of abandoned and foreclosed homes. This will enable the police to efficiently provide additional patrols in those areas and avoid unnecessary blight.
What's more, the city has identified an additional $23 million in potential operational savings over the next five years by adding sensors, purchasing more licenses for the analytics software and using the system's data to help make a variety of services more efficient. One such service involves closely monitoring sidewalk temperatures in order to more accurately plan for purchases of de-icing materials. To that end, Gilot says he's working with the entrepreneurs at Notre Dame to be able to tag more types of data in the Intelligent Operations Center to support such efforts.
"It's one platform that's beginning to look across the silos and run a smarter city," says Gilot.
An Analytics Renaissance
Such efforts to get smarter are a big reason for the renaissance currently under way in the global business analytics market, which grew 14.1 percent during 2011 and is expected to reach $50.7 billion by 2016, according to recent research from IDC.
While the campaign to get a handle on big data is at least partly responsible for this anticipated growth, John Myers, senior business intelligence analyst for research firm Enterprise Management Associates, says most companies haven't actually arrived at a big data issue yet. Rather, the market is being driven by more traditional business goals, such as the desire to fuel top-line revenue growth, reduce costs and replace gut feelings with more informed decisions.
But what Myers believes is most responsible for the growth of the analytics market is the fact that analytics technology, "is bubbling down into smaller organizations so they can take advantage of it," he says. "You have a whole new avenue that's opened up and is available to these users."
Delta Dental of Virginia is a perfect example of this. Having chugged along for nearly 40 years as a small provider of dental health benefits, the company has experienced meteoric growth over the past decade, with annual revenues increasing from less than $20 million to more than $500 million. It now provides benefits to nearly 900,000 subscribers though 4,000 groups.
That growth brought a need for big changes; no longer could Delta Dental run its business by the seat of its pants, says John Sheffield, director of software development. It needed to get much more sophisticated about how it priced and packaged its services, analyzed the competition and monitored claims.
About five years ago, after struggling with a few commercial products that proved too complex, the company came across Pentaho, an open-source analytics solution that Sheffield says allowed Delta Dental to "do the crawl-walk-run thing." The company immediately started pumping data from its transaction system into an analytics model and provided that data to its customers so they could more effectively manage their contracts.
It wasn't long, however, before the company's use of its analytics engine turned inward, and it started using it to study its claims history and determine where provider fraud might be occurring. The resulting tool, developed over nearly a year by Sheffield and the rest of Delta Dental's IT staff, is now used by more than 50 people who manage the company's provider network, and is the subject of a patent application. More importantly, once the company began letting providers know how it was using its analytics capabilities, it began to see what Sheffield dubbed a "sentinel effect."
"We've seen provider submission behavior change radically because of telling them what we're doing," he says. In other words, providers no longer feel they can sneak excessive charges into their invoices.
Meanwhile, the company's underwriters also started using it to refine the pricing of its coverage options, studying transactional data and competitors' pricing in an effort to become more competitive without allowing prices to go too low. They also began using it to analyze brokers, causing a subtle shift toward the brokers who achieved the best outcomes for the business. The result has been a 10 to 15 percent improvement in top-line premiums.
Additionally, after spending about $400,000 on support and consulting services in each of the first few years developing its Pentaho environment, the company has seen its need for customization drop off as the tool's feature set has expanded, enabling it to cut the related annual costs in half.
Now, says Sheffield, "We couldn't run our business without something like this."
Bringing Discipline to Data
After more than a decade of enduring analytics capabilities that were limited to whatever an Excel spreadsheet could spit out, executives at Allconnect finally cried "uncle" in 2010. That's when this Atlanta-based company, which resells home phone, television and Internet services to consumers who are referred by utilities providers, implemented SAP's Business Objects analytics package, bringing a new level of discipline to the company's use of data.
While the technology initially was intended as a sales-focused tool, it didn't take long for Allconnect to start crunching financial, customer satisfaction and quality assurance data, too. "We have put every bit of data that drives our business into our Business Objects platform," says Bobby Nix, IT director.
Employees now can use the platform to link various pieces of customer data—such as the customer's dwelling type, the referring utility and the services they purchased—and see the connections between these factors and the resulting customer satisfaction levels.
"We've begun to bring in a much richer context around the data," instead of relying on analysts to provide what they hope is the right data, Nix says. That's led to a measureable improvement in customer satisfaction and an increase of more than 10 percent in revenue per call.
Going forward, Nix says Allconnect would like to roll out more dashboard views for exposing this data, as well as plugging in some predictive analytics capabilities that will enable it to make more dynamic offers to customers.
Predicting Business Outcomes
Many organizations are using analytics software to predict likely business outcomes and become more proactive, reports Scott Schlesinger, vice president and head of the business information management practice at consultancy Capgemini North America.
Schlesinger says predictive analysis is critical to getting ahead of the competition in an increasingly data-intensive era. "Any companies that aren't looking at their data to gain competitive advantage could become irrelevant," he says.
At the Thule Group, best known for its rooftop car storage products, the need to better predict demand was the main reason for adopting business analytics. As recently as 2006, the Malmö, Sweden-based company relied on a spreadsheet-driven process for financial analysis and forecasting that created a lot of confusion about what numbers were right and often resulted in lost spreadsheets, says CFO Mark Cohen.
That was when Cohen and the company's vice presidents of sales and operations started looking for an analytics tool that could complement the company's use of PivotLink, a Web-based app it was using to publish raw sales data. They eventually settled on Host Analytics' financial analytics application, using it to crunch data they pull from their Oracle finance system into PivotLink.
Initially, Host was rolled out to more than two dozen functional managers and a handful of sales managers, all of whom began using the resulting insight to better predict demand for the company's products. Just a few years removed from having to rely on the finance department for all of their forecasting data, those managers now are able to ensure that there's always the right amount of inventory on hand, and that staffing levels meet expected demand.
"We're doing a better job of laying inventory into the building," says Cohen. "We're not sitting on excess inventory to buffer a lack of visibility into what we're going to sell."
That's not all: Thule's Host environment is enabling it to integrate new acquisitions into its planning and forecasting processes without turning to IT for help, as it did last year when it acquired Calgary, Alberta-based Chariot Carriers. And the company's sales and marketing staff uses Host to analyze data collected from the 15 to 20 trade shows Thule attends each year, using the resulting knowledge to prioritize the use of trade show resources.
Meanwhile, individual users are bypassing Oracle and using Host to build their own customized reports. "That was one of the things that told me this was working much better than I thought," says Cohen.
Organizations like the City of South Bend, Delta Dental, Allconnect and Thule might not be at the big data stage yet, but they're clearly getting big results with analytics solutions.