Business intelligence: Gleaning a 360-degree view of the customer

By Judith Lamont

Business intelligence solutions, once focused primarily on operational data, are extending their reach into analysis of customer relationship management (CRM) data and outside the enterprise to partners. New products are emerging to retrieve and organize critical business intelligence, much of which is being collected from e-business activities. Collectively, those new tools and strategies are providing a rich array of information through which to evaluate company performance and understand customers.

Companies are now seeking solutions to measure performance throughout the enterprise. “Demand for data to drive decisions at every level of the company is growing,” says Neil Patil, senior director of product marketing at the newly named Brio Software, formerly Brio Technology. “We are seeing this not just among a company’s managers, either, but also from partners and suppliers.” Patil says a broader definition of BI software to “business performance software” is emerging.

The Brio Performance suite of tools includes query, analysis and reporting software that can be integrated into Brio’s portal and will be augmented by Brio Performance Applications, a recently announced future product line intended to provide business process-specific metrics for various industries. Among the strengths of Brio’s tools are ease of implementation and use. Integrating all of a company’s information systems is important, says Patil, because it allows correlation between front-office information such as customer purchases and back-office information such as delivery time. By looking at those two data points together, it’s possible to find out if customers whose orders arrive quickly are more likely to order again, for example.

“We see BI starting to become part of the core fabric of enterprise applications,” asserts Andy Handford, product VP at Crystal Decisions. “Organizations need to get information related to all aspects of the business and then understand what that information means to them.” Crystal Decisions’ product suite offers analytical and information sharing capabilities, and has particular strengths in reliable and scalable delivery of reports. Crystal Enterprise integrates into e-business architecture and can access data sources that range from enterprise resource systems to customer relationship management (CRM).

Increasingly, companies are also providing business intelligence to their customers. Zenith Media Services , for example, uses Crystal Enterprise to deliver extranet solutions to its clients so they can evaluate the effectiveness of their advertising purchases. Zenith estimates that the savings in programming costs (compared to aggregating data in Excel) paid for the system in three months. The new system also benefited Zenith because its use allowed more individuals within the company to readily access the data. By integrating with Lotus Notes and Lotus K-Station, it also leveraged the company’s existing infrastructure.

With the dynamic growth of CRM systems, much more customer data is available for analysis. But CRM systems are designed to document customer transactions, not to aggregate and analyze the resulting data. To really make the most of that data, powerful analytic tools and statistical techniques must be called into play. Timo Elliott, senior director of strategic marketing at Business Objects, gives an example of what sophisticated analysis can offer.

“During the dot.com boom,” says Elliott, “a company came to us very pleased that their most profitable market segment had doubled in volume over the past year.” Analysis by Business Objects showed that in fact, volume had increased by a factor of six, but most of the customers had not been retained. “Segmentation shows not just how many customers are there at the beginning and end of the year,” says Elliott, “but whether they are the same ones.” As a result of the analysis, the company decided to spend more on retention, rather than marketing to new customers.

Customer loyalty is attracting a lot of attention because even small increases in the percent of retained customers translates into large increases in profits. Getting new customers is expensive, considering marketing costs and special offers. Serving them is also more expensive in comparison with long-term customers, who are more familiar with the company’s products and may have existing relationships that expedite transactions.

“It’s butterflies vs. barnacles,” says Elliott. “Customers who are lured by a lower price will also move on when a better deal is offered.” He maintains that offering top service to loyal customers is the best strategy. Waiving a fee for a banking customer who has multiple high-value accounts is a good example. But all too often, one side of a bank (the loan officer, for example) may not know what the other side (the investment counselor) is doing. Analytical tools help identify those “best customers.”

Ernex Marketing Technologies, which provides loyalty marketing and electronic marketing solutions for clients such as Nike and Levi Strauss, uses Cognos business intelligence products for analysis and reporting on customer behavior. Customers’ purchase patterns are tracked at the point of sale, and rewards in the form of coupons, points, or prizes are offered on the spot. Ernex delivers reports based on those analyses, and clients can drill down through the data for more detail or to see patterns that can guide promotional campaigns. The system uses Cognos’ PowerPlay, an OLAP tool, and Impromptu, an ad hoc query tool, to analyze the data.

ROI is of increasing importance, according to Jennifer Cullen, senior manager for CRM markets and applications at Cognos. Properly deployed, systems can pay for themselves quickly. A company may find out, for example, that it is using more people in the call center than are needed.

“The biggest mistake that companies make is not defining their questions soon enough,” says Cullen. “Without careful planning, they may end up collecting the wrong data, or overlooking data that could provide important insights.”

Some Web sites now specialize in collecting and analyzing customer data for other companies. PlanetFeedback was launched last year to assess consumer perceptions. Through BrandPulse, a set of analytical tools based on Business Objects WebIntelligence, companies can understand what the top issues are relative to their brand and how their products compare to those of competitors. The site offers consumers an opportunity to vent or praise the company, but also collects quantitative data about whether the customer would use the product or service again or would recommend it to others.

Companies seek out consumers who access such Web sites because those individuals tend to be opinion leaders. PlanetFeedback participants are 54% more likely to discuss their experiences with others as compared with non-participants. Giving proactive consumers an environment in which to express opinions that they are already sharing with others helps the company tap into important trends. Some organizations also host their own customer feedback sites. Whether it is provided by a service such as PlanetFeedback or is an in-house effort, customer feedback via the Web will become valuable only if carefully analyzed, integrated with other sources, and interpreted.

With roots in statistical analysis dating back to the late 1980s, Insightful offers a business intelligence tool called StatServer that excels at data mining. The term “data mining” is often used to refer to analysis of information contained in data warehouses, but a more accurate meaning is the ability to detect, in a semi-automated manner, patterns in the data that may not have been previously recognized. While OLAP tools help reveal correlation between such variables as sales figures and promotional programs, data mining can help predict what products might be successful in the future.

“The ability to forecast demand for new products is very valuable,” says Peter Hallett, VP of marketing and business development at Insightful, “but it cannot be achieved by looking at historical sales of existing products.” Predictive modeling helped Pillsbury succeed with the introduction of a new chocolate breakfast pastry when the company used StatServer to analyze information in the 600,000 calls it receives each year from consumers. A model may draw upon sales information in a data warehouse, but it also incorporates other data sources to predict customer behavior.

Integrating disparate sources of information is an essential element of business intelligence, and portals serve an important role there. Enfish offers a family of integrated portal products that automatically index and cross-reference information and organize it in context.

“A distinguishing feature of Enfish Enterprise,” says Louise Wannier, CEO of Enfish, “is that it integrates corporate knowledge with information from the user’s hard drive, as well as from disparate sources on the server.”

Jeff McNaught, VP of market strategy for Wyse Technology, is typical of how executive-level staff can benefit from Enfish. “It saves me significant amounts of time every day,” volunteers McNaught. “I don’t have to remember file names or locations--Enfish presents information as I need it.” When a user’s schedule shows a company name, for example, all the information related to that company is automatically presented.

Xerox recently introduced askOnce Version 2—a source integration tool that is effective in gathering business intelligence—into the U.S. market. AskOnce was developed by the Xerox Multilingual Knowledge Management Solutions (MKMS) group based on research from the Xerox Research Center Europe in Grenoble, France. It conducts a federated search across multiple data sources. One of its strengths is that it can connect to the search front end of “deep” Web content (as opposed to information on a public Web site) to access information.

“If a company has a subscription to reports from the Delphi Group , for example,” says Mary Bennett, director of North American operations for MKMS, “askOnce can search those reports using ‘wrapper technology’ that lets it connect to the data source.” AskOnce can be run on a regular basis to retrieve information about a company and its competitors. Search results come back in the form of links to the data sources. The software also enhances the query and retrieval process; for example, if a data source does not support an “OR” operator, askOnce compensates by combining the results of two separate searches. In addition, it automatically removes duplicate results from search results.

With increasingly sophisticated and easy to use analytical tools along with Web data gathering strategies, companies have an impressive array of options for obtaining customer information that can support data-driven business decisions.

Tapping consumer buzz

One source of customer information has gone largely untapped until recently. Consumer discussions on bulletin boards, message boards and news groups provide windows into their opinions, but the information has been difficult to capture.

“A strong argument can be made,” says Karthik Iyer, senior VP of business development at Intelliseek, “that this unsolicited, implicit consumer feedback is of greater value than traditional methods such as focus groups and interviews.” Intelliseek’s Corporate Intelligence Service (CIS) monitors such sources to provide insights into consumer perceptions and sentiments on brands and issues.

CIS evolved from BullsEye, a search and retrieval from Intelliseek that conducts a federated search of information sources on the Internet and provides filtering and analysis. BullsEye customers had wanted several new features, including enterprise search capability for their own intranets, a browser-based rather than client-server model, and XML capability. In response, Intelliseek developed an Enterprise Search Server. CIS is built on top of that server, and adds the ability to access sources of unstructured consumer data on the Internet. The process includes discovery of the sources and collection of data, both of which are managed by the server. For analysis, Intelliseek has developed a variety of patented analytics and metrics. For reporting and presentation, Intelliseek chose the MicroStrategy 7 Business Intelligence Platform from Microstrategy.

According to Iyer, the greatest use of the new technology has been by industries that devote significant resources to building consumer-facing brands, including auto manufacturers, retail companies, telecommunications and financial services.

“They are trying to find out where the buzz is taking place, what else are they talking about, how loyal they are,” says Iyer. Companies also are attending to what is called “reputation management,” tracking rumors and responding to them if necessary. He says that the problem with Firestone and Ford appeared on the Internet well before it received attention from regulators.

Judith Lamont is a research analyst with Zentek Corp., e-mail jlamont@sprintmail.com.