Technology: Predictive Analytics Lets Companies See into the FutureBy Larry Stevens | Posted 08-23-2006
Technology: Predictive Analytics Lets Companies See into the Future
How good has information technology gotten at predicting the future?
For years, companies have mined historical data in order to better craft future plans. By and large, those plans were based on the quality of managers' gut intuition. But as the complexity of markets grows, managers are seeking out new tools to help make critical decisions.
Consider the efforts of Center Parcs Europe to stimulate traffic to its 16 short-stay vacation villages in the Netherlands, Belgium, Germany and France. Center Parc's primary marketing effort was always a bit like Popeye's nemesis, Bluto: bulky and bruising, but lacking in the wiliness and cunning of his spinach-eating competitor. Twice each year, the company would blanket the continent with 5 million brochures. The strategy worked, but the collateral damage was becoming unacceptably high. "We were hurt badly by mailing and printing costs," says Richard Verhoeff, technology and e-business director at Center Parcs Europe N.V., a subsidiary of Pierre & Vacances Group.
Center Parcs Europe attracts more than 3 million visitors a year and generates revenues of more than $600 million; its 10,000 or so bungalows enjoy an average occupancy rate of about 90 percent. But Verhoeff believed that a more targeted campaign could raise occupancy rates even higher, while reducing marketing costs.
The plan was to shift away from the bulk mailings to many leaner campaigns. To be effective, however, the company had to segment customers so they received the right brochures at the right time. Verhoeff points out, for example, that customers are loyal to their home countries, and sensitive about driving long distances. Families prefer parks with lots of activities for kids, while older people seek out quieter surroundings. And mailings must be timed to coincide with the period when people typically make vacation plansnot too early and not too late.
These and a dozen or so other criteria were not difficult. "This isn't brain surgery. We knew what drives customers to register for one of our parks," Verhoeff points out. The problem was more fundamental: He couldn't find a way to take all the criteria into account and come up with a rating that would indicate which brochures to send to which customers, and when.
To solve his problem, Verhoeff turned to PredictiveMarketing, a predictive-analytics product from Chicago-based SPSS Inc. The application uses data from Center Parcs's customer database to rank prospects by the likelihood that they will respond to a particular campaign. Now, when the company wants to mail a brochure about a facility, the predictive-analytics application produces a list of prospects ranked by how likely they are to respond.
The results have been dramatic. In Germany alone, the number of Center Parcs mailings dropped from 2 million to 450,000, while occupancy increased by 10 percent.
Like other forms of business intelligence such as data warehousing, online analytical processing (OLAP), data mining and CRM, predictive analytics looks at historical data. But unlike traditional business intelligence, which involves the questioning of past trends and then using your gut to determine future actions, predictive analytics adds an oracular element. By applying statistical algorithms, predictive analytics finds patterns that can be used to predict future actions. These predictions, often expressed in the form of a ranking or a score, can foretell consumer behavior, market trends, the likelihood that a customer will make a repeat purchase, the effectiveness of products or services, and virtually any other factor that can be statistically analyzed.
Are you having problems making successful predictions?
What business intelligence tools are you currently using?
Not every prediction is important.
Kurt Schlegal, a research director at Gartner Inc., points out that to provide a good ROI, predictive-analytics applications have to be aligned with the corporate strategy. "Find out what drives the corporation's bottom line and implement a system that makes predictions that affect that," he says.
Take Aegon Direct Marketing Services Inc., the direct-to-consumer insurance subsidiary of Netherlands-based Aegon N.V., with $20 billion in 2004 sales. One of the largest direct marketers of insurance in the U.S., Aegon sells a number of brands, the most well known being Transamerica.
The firm's corporate strategy is clear and unambiguous. "We sink or swim on our ability to identify, with a great deal of specificity, people who will buy from us," says Mark Rude, senior vice president of marketing services. Last fall, when he was CIO, Rude implemented marketing optimization software, from SAS Institute Inc., to enhance that strategy.
The company pitches its life, health and disability products through direct mail, telemarketing, Web ads, inbound calls to its call center and other direct-to-consumer avenues. For years, it used a variety of business-intelligence and data-mining applications, some homegrown. But these applications generated prospect lists that had the dual deficit of being too long while also missing some of the most likely candidates. "Those problems were putting a lot of pressure on our bottom line," Rude says.
The new software accompanied a major shift in marketing strategy aimed at taking advantage of Aegon USA's list of current customers, which Rude felt was being overlooked by the previous process. Simply put, the company's lack of predictive analytics had precluded a robust cross-marketing strategy.
Aegon USA's previous marketing strategy involved a series of discrete efforts, each focused on a product or a campaign, not on customers. But Rude wanted to look at marketing from the customer's perspective. "For each customer, I wanted to know all the products he or she is likely to buy, in the order of likelihood," he says.
The software provides a predictive score for each potential sales opportunity the company has for each customer. The score changes monthly, as customers' situations change. For example, as a customer grows older, has a child or buys another Aegon insurance product, their score may change accordingly.
Rude points out that predictive-analytics applications also allow organizations to add business rules, which consider strategic factors that may affect customers the organization wants to solicit. The customer most likely to purchase a product may not always be the best one to sell to. In fact, Aegon includes a number of such constraints when generating its prospect lists. For example, the company may decide not to pitch a low-return product to a customer, despite his or her high likelihood of buying it, because the same customer may be likely to purchase a more profitable product. "There are a lot of balls to juggle. It would be virtually impossible without predictive analytics," says Rude.
Can we identify our most profitable products?
Is our CRM system up to the task of providing accurate information on customers?
Top predictive analytics hurdles: creating the proper statistical algorithms and finding, sanitizing and accessing the needed data.
Implementing predictive analytics requires either a statistical analyst who understands your business and your data, or a team of people who can work together to combine data, business rules and business strategy.
When Yorkshire Water, the U.K. city's water and sewer utility, decided to use predictive analytics to determine the likelihood of failure in any of its pipes, "we had to create a pretty robust proof of concept first," says Melvin Parkinson, IT specialist at the utility. "And to do that, we needed to get our operations people together with technical people like myself and people with statistical knowledge."
Yorkshire Water, which provides 2.2 million households and 140,000 businesses with water and sewer services, decided it didn't have sufficient in-house expertise. So it turned to SPSS's consulting services to help implement the firm's data-mining and predictive-analytics software, called Clementime.
Predicting the likelihood of a pipe failure would allow the utility to fix the problem before a pipe burst. The utility's operations people already knew what factors increase the likelihood of flooding: the number of cellared properties, the length of sewers, rainfall, minor and major excavation work in the area, age of pipes, and dozens more. But when deciding which parts of the system needed work, most operations people could only juggle about four factors in making their recommendations.
The work with SPSS's statistical consultant did not always flow smoothly. "It was very time-consuming to explain our business and maintenance methods, and have the consultants program that into statistical form, test it, and then fix things that didn't work well with the program," says Parkinson.
But as a result, Yorkshire Water now has an application that can analyze more than 120 factors, and ranks sections of the utility's system by risk of failure on a weekly basis. "At this point, the system is pretty much self-working, although we are always refining it a bit here and there," says Parkinson.
Unfortunately, not every organization has the large, robust database Yorkshire Water had. Sometimes the need to create such systems becomes a major problem in developing predictions.
Time/Warner Retail Sales and Marketing, a subsidiary of Time Inc., is a magazine distribution company that regularly places about 400 Time Inc. and client magazine titles in 120,000 stores, generating more than $1 billion annually. For years, the division's marketing people made decisions about inventory in each store by poring over dozens of wholesaler reports. That laborious task led the company to consider predictive analytics to help in inventory, pricing, cover design, and other factors. But executives had to contend with the fact that much of the required data was owned and controlled by the wholesalers. Says Dilip Patel, Time/Warner director of BI systems and information management: "Building the database was not an easy task, because we had to deal with five major systems and dozens of smaller systems from mom-and-pop operations."
Time/Warner turned to Insightful Corp.'s S-PLUS Enterprise Server to deliver analytics to marketers, as well as to client publishers. But before it could be implemented, "we had to find common ground among wholesalers' systems," says Patel. The problems ranged from major formatting inconsistencies to relatively trivial but irksome idiosyncrasies. Some wholesalers didn't list stores by their actual retail addresses, for example, but by a description of their locationshelpful to the driver but useless in a database.
Time/Warner appointed Pittsburgh-based Management Science Associates Inc. to collect, standardize and consolidate the wholesaler data. MSA collected data on every Time/Warner publication, as well as data from its competitors (most wholesalers, who owned the data, agreed to provide it to Time/Warner). The competitive information helped Time/Warner to determine inventory for new publications.
MSA is responsible for cleaning, matching and verifying the store data, and for providing a monthly updated store-level sales data reporting system. According to Time/Warner's Patel, the ROI on the application is 282 percent since it was implemented in 2001, thanks primarily to a reduction in the amount of waste and publications returned from stores.
Do we have sufficient statistical analysis resources in-house?
What data do we have in-house, and how is it stored?
How powerful and widespread can predictive analytics become? Most analysts are upbeat about predictive- analytics applications. Says Dan Vesset, research director for analytics and data warehousing at IDC: "I believe we'll see increased sales, deeper penetration into industries that never used predictive analytics before, and widespread use in companies for which customer segmentation is strategic."
Some form of predictive analytics will increasingly be incorporated into other applications such as CRM, call-center software and industry-specific applications, Vessel believes. "More and more packaged applications will include predictive-analytics modules," he says. And while those modules will not be as full- featured as dedicated applications from SAS, Oracle Corp., SPSS, Insightful and others, they will be easier to implement.
Overall, the good news for companies is that predictive analytics is becoming more robust and more widely used. They should allow companies to make better predictions about outcomes of strategic maneuvers. And while these applications will never replace the judgment and intuition of experienced professionals, they will provide a good tool for testing hypotheses and for creating automated systems.
What kinds of analysis are you hoping to do in the future?
How can we integrate predictive analytics more tightly into our business processes?