Maps are back. More than a decade ago, geographic information systems took population and demographic data from the Census Bureau and other sources and gave businesses new ways to see opportunities, literally. Consumers could be segmented by zip code, for example, or regional sales could be analyzed with maximum granularity, the results writ large on color-coded maps. The immediacy of the visual presentation was not lost, particularly for those falling below quota.
The advent of the global positioning system (GPS) has added a new technological wrinkle: the dynamic mapping it makes possible now creates an entirely new universe of applications. Today, for example, if your dishwasher goes on the fritz and you call a Sears service center, odds are good that the technician will find your home by using mapping software and a laptop computer. True, you can buy or rent a car with the same feature. But in this case, Sears isn’t simply replacing a paper map or a cell phone with an incrementally better solution, it’s combining that mapping capability with wireless data applications that not only get the driver there faster but also usher in a raft of bottom-line improvements.
“We strongly believe the future belongs to those companies that successfully innovate and use technology to drive down costs and increase service levels,” says David Sankey, director of process and technology development at Sears Product Repair Services in Hoffman Estates, Illinois.
Sears isn’t alone in looking for new ways of “seeing” with visualization tools. Many companies are looking for ways to cope with information overload. While dashboards and other techniques continue to gain in popularity, maps are proving surprisingly versatile as well, even for deskbound workers.
“Academic research shows that we can’t keep more than seven data points in our head at one time,” says Dan Vesset, research director for business analytics at research firm IDC. “If you have a table with more than seven rows and multiple columns, you very quickly get lost.”
The versatility of spreadsheets is beyond question, but when a company needs to locate a new store or identify the site of a power outage, numbers don’t tell the story nearly as well. “With some applications, you can look at a spreadsheet all day long, but rows and columns will never show you the pattern that emerges on a map instantly,” says Robert Hazelton, technical director at Information Builders Inc., a business-intelligence software company in New York.
According to IDC, the market for interactive data visualization tools will grow by 7.5 percent a year through 2007, reaching almost $7 billion. The category includes software ranging from visually-intensive engineering applications and image editing software to geospatial information management tools and animation software. Also included is a segment of business intelligence that brings data visualization to executives, line-of-business managers, and analysts. It provides a number of ways to present data visually, such as maps, diagrams, stoplights, flashing alerts, decision trees, and dashboards. Several kinds of business-intelligence software vendors are adding visualization, frequently through licensing agreements with niche providers that do business with the governmental, scientific/engineering, and academic communities.
Now You See It
David Lewis learned fast when he launched both the marketing and information technology departments at ING Direct in 2000. Now one of the fastest-growing U.S. banks, with more than 2 million customers and more than $30 billion in assets, ING Direct initially let departments choose their own software for reporting and other needs. As a result, producing monthly finance reports required two or three full-time employees who could look across disparate systems and pull numbers together.
That made it difficult to understand performance metrics, so the company switched to an enterprise database from Oracle and business-intelligence software from Cognos. The latter includes a dashboard, which provides a flexible way to display those metrics visually and customize them for every department. “This allows the people who best understand a given type of data to analyze it easily,” says Lewis.
Dashboards can link cause-and-effect and other relationships between numbers in a way that traditional reports can’t. “A supply-chain dashboard might show that inventory is low in a certain area, yet sales are good,” says John Kreisa, a group marketing manager at Business Objects, another business-intelligence/dashboard supplier. “So you know right away that you need to do something about it.”
Dashboards have evolved from executive information and decision support systems, and now fold in more data in more customizable ways. What began as a set of distilled numbers provided to executives too busy to crunch data can now be displayed via flashing lights, maps, and other means, all in an effort to give as fast a read on the data as possible. Drill-down capabilities allow a user to click on summary data and trace it back to its various sources.
The goals are still to improve decision quality and to fight the so-called fact gap. “Many decisions are being made within an organization that aren’t based on factual information; they’re more based on gut feel,” says Kreisa. “The bigger that fact gap — decisions made without facts — the more at risk a company is of decisions going the wrong way.”
Efforts to present data more effectively have been galvanized by “spatial information technology,” a combination of software, satellites, sensors, and technologies such as radio frequency identification tags that enable users to geosynchronize production inputs and thereby streamline operations, says Susan M. Wachter, professor of real estate and finance at the Wharton School of Business at the University of Pennsylvania. “We now have the capacity to affordably geosynchronize the supply chain,” she says. Companies can now track resources and people in real time, operating “just-in-time” and “just-in-place.”
“This is not just a step up, it’s a leap,” Wachter says. Retailers can go beyond zip-code analysis by converting the addresses of known customers to dots on a map and adding such demographic information as average incomes and average spending on specific items, such as furniture or sporting goods. Armed with that view, optimizing the location of new stores becomes far more accurate.
The Sound of All Hands Mapping
The insurance industry belatedly discovered the power of visual representation following the September 11, 2001, attack on the World Trade Center. “There was a tremendous concentration of risk in lower Manhattan,” says James Bisker, director of insurance research at TowerGroup. “Not just the buildings themselves, but how much was insured by one company in each building.” That data lends itself well to visual analysis, he says, but “because the information was buried in tabular reports, it wasn’t as obvious.”
Gen Re Intermediaries (GRI) has used mapping software for years to model its exposure. Now it’s sharing this capability as part of its competitive strategy, allowing customers to check their own catastrophe exposure via the Internet, looking for geographic hot spots — too little (or too much) earthquake coverage in California, for example.
Mapping capabilities can now be integrated with other enterprise software applications such as ERP, CRM, SCM, and business-intelligence systems. These systems hold vast amounts of data, and just getting to it can be a challenge. “If you’ve ever tried to retrieve information from an ERP system on a particular physical asset, you typically go through four or five screens and may have to rely on a little cheat sheet to remember different asset codes,” says Steve Benner, director of strategic accounts at GIS developer ESRI Inc. in Redlands, California. “Mapping software lets you see the asset visually and provides direct access to data in the ERP system by simply clicking the asset on the map.”
Mapping software can also show at a glance where resources such as field workers are in relation to one another and lets you use spatial functions to query data. You could ask, for example, “What percent of our sales comes from customers within a 15-minute drive time of our stores?”
GIS software varies considerably in price, from $995 for a single seat at a small business to $500,000 for large companies, which typically spend another $500,000 for implementation and staff training. Even at those prices, analysts say ROI is often quick, particularly if companies haven’t made use of mapping previously.
So what’s holding companies back? “It’s a complete mind shift for most businesspeople to even begin considering ways that map-based representations can be used in their work,” says Information Builders’s Hazelton. But as more software vendors build mapping and other forms of visualization into their products, people’s eyes may open.
Connie Winkler writes about management technology from Seattle.
The Big Picture, in Words
Despite today’s zeal for presenting data visually, sometimes a word is still worth a thousand pictures — or at least a handful of rows and columns. So says Sageworks Inc., a Raleigh, North Carolina-based software company whose ProfitCents product turns numbers into text.
“Graphs, pictures, and other displays may be easier to understand,” says Sageworks CEO Brian Hamilton, “but they depict numbers rather than address the basic question: ‘What is the big financial picture for the company?’ ” ProfitCents analyzes dozens of metrics, ratios, trends, and key comparisons, and produces a narrative summary. One evaluation of a firm’s assets, for example, begins: “The net profit margin fell as fixed assets were added. This generally means that the company is now less efficient with the new assets than it was previously without them.”
Plain English has proved a selling point for Sageworks’s customers, which range from large enterprise companies to midsize banks and accounting firms. “We give our customers the graphs and all that kind of data,” says Hamilton, a former CFO, “but we put it in language that they can easily understand.” CEOs and board members like it, he adds, because it answers the age-old question: “What does all this data actually mean?” —C.W.