Analyze your customers

Alan Radding

Updated · Jun 12, 2001

When AXA Financial Inc., a New York-based diversified financial services company, focused on its products, it didn’t bother with customer relationship management or CRM analytics. Customers? They weren’t even on the radar screen. AXA is a company with $565 billion in assets under management, and about two years ago things started to change dramatically. New management declared customers, not products, to be the primary company focus.

“This shift kicked off a number of changes in IT. Basically, we were going to transform the agent from an insurance person to a financial person,” explains Sharon Sibigtroth, managing director, strategic data at AXA Financial. Now, the agent–and everyone else–needs to understand the customer, his or her motivations, retention issues, and more. In short, they need information about the customers. So, the IT group set about building an enterprise data warehouse as its first step toward providing this information.

Very quickly, the enterprise data warehouse was bulging with customer data. “There was a lot of data. Humans can’t really handle it alone. We needed analytical and datamining tools,” continues Sibigtroth. By 1999, the company added SAS Enterprise Miner from SAS Institute Inc. as its global choice for CRM analytics. “We set up a new group to help us understand customer needs, motivations, and retention issues, and SAS’ Enterprise Miner will help us do this,” says Sibigtroth. Insights gained here will be used to personalize communication and interaction with customers.

AXA Financial isn’t the only company adopting CRM analytics. From startup dot-coms that desperately need to convert browsers into buyers to established companies seeking to get closer to their customers, businesses are turning to CRM analytics.

Mining insightful nuggets
CRM analytics is the latest label for an old concept, datamining. For a decade or more, large packaged goods manufacturers and direct marketers turned to highly sophisticated datamining tools for insights into how best to promote and merchandise to their customer base.

Proctor & Gamble Inc. and American Express Co. Inc., among a host of leading marketers, employed many statisticians and Ph.D.s armed with complex datamining tools to cull insightful nuggets from volumes of point-of-sale (POS) data and other information they collected. Just a few nuggets could help these companies boost the response rate to a campaign or promotion by a few percentage points, which could easily translate into millions of dollars in additional sales.

“CRM analytics builds on data warehousing, datamining, and business intelligence,” observes Phillip Russom, director, data warehousing and business intelligence knowledge center at Hurwitz Group, a research firm based in Framingham, Mass. CRM analytics, or CRA (customer relationship analytics) as Russom prefers, deals with the analysis of information while data warehousing focuses on the management of information. Businesses also use online analytical processing (OLAP) and even ad hoc querying for the purposes of CRM analytics.

CRM is attracting big money. International Data Corp. (IDC) of Framingham, Mass., which does not break out CRM analytics separately but considers it part of the CRM market segment, expects worldwide revenues generated from CRM data warehousing software and services to soar from under 4.2 billion in 1999 to over $20 billion by 2004 (see chart, “CRM growth skyrockets”).

While any data analysis tool can be applied to CRM and most CRM tools provide some analytics, tools intended for CRM analytics differ in several ways. Often they can capture data directly off an Internet site and perform some analytics in real time rather than in batch mode, thus enabling the organization to respond while the customer is still on the telephone or online. They also will include prebuilt models that address the types of CRM problems many managers typically wrestle with, such as converting browsers to buyers or boosting customer retention.

Driving the interest in CRM analytics, both directly and indirectly, is the Internet, suggests Dan Vesset, IDC senior analyst. The Internet, for example, gives companies access to massive amounts of data that can be analyzed and massaged to reveal valuable insights into customer behavior. “Every click is data to be stored and analyzed,” Vesset explains. And this Internet data can be brought together with other data in one place–typically the enterprise data warehouse or a marketing datamart–to create a complete picture of the customer.

AT A GLANCE: AXA Financial Inc.

The company: Part of the global AXA Group, which operates in 60 countries,
New York-based AXA Financial Inc. manages nearly $565 billion in assets. Its
major U.S. brands include The Equitable Life Assurance Society; Donaldson,
Lufkin & Jenrette (DLJ); and DLJdirect.

The problem: The company needed IT support for a major business strategy
shift from a product-centric to a customer-centric focus.

The solution: Implement a CRM IT infrastructure beginning with an enterprise
data warehouse built on a UNIX/Oracle Corp. database platform and customer
information architecture. It is installing the Siebel CRM product and has
implemented SAS Enterprise Miner for datamining and CRM analytics.

Drilling the data for driving forces

The Internet also magnifies a trend that began before the Internet–the shift in the balance of power from the seller to the buyer, the customer. Customers who had growing numbers of alternatives before the Internet now find a mind-boggling array of options available to them on the Web. If you don’t like the product selection, price, or terms being offered, a simple click of the mouse brings up a completely new set of options. In this kind of competitive environment, businesses must become highly attuned to the customer’s every whim, identifying and responding to even slight shifts in customer preferences. This is where CRM analytics pays off.

eBags Inc., a Greenwood Village, Colo.-based pure-play online retailer of a wide range of carrying products from handbags to luggage, turned to CRM analytics to better understand its customers and their behavior online. Using Broadbase Software Inc.’s CRM application, Foundation, eBags already has segmented the visitors to its site and is beginning to mine the data. “This will put us light years ahead of what our merchandising and marketing people are used to,” notes Mike Frazzini, vice president, technology.

The initial eBags CRM analysis is based on OLAP. Managers download CRM data into multidimensional cubes provided by Broadbase. The cubes allow managers to view the data from different angles and, at the same time, drill down or up to get more detailed or higher level views. “We offer a huge selection, which makes it hard for people to find what they want,” Frazzini continues. The company has already redesigned its tab system based on CRM analysis to make it easier for customers to navigate the site. It is now analyzing the CRM data to understand what drives increased sales at the product level.

Sunrise Medical’s Respiratory Product Division, based in Somerset, Pa., a conventional provider of medical equipment, has turned to CRM analytics to help it manage its customer base of 12,000 distributors in the United States. It has collected every transaction from the past three years–more than 10 million records in all–in a data warehouse. Using Cognos Inc.’s PowerPlay, it now loads that data into multidimensional cubes for CRM analysis.

The goal is to boost customer and product profitability. “We have thousands of products, some more profitable than others. So we try to understand and control the mix of products customers buy,” explains Richard Kocinski, president. CRM analytics help guide the company’s salespeople in putting together programs and offers that meet the customers’ needs while ensuring account profitability.

Companies learn to manage relationships

Data warehousing software and services for customer relationship management will increase dramatically over the next five years.

Copyright: International Data Corp.

As interest in CRM analytics spreads, both new and established vendors are rushing to offer solutions. All the data warehousing vendors either directly or through partnerships offer CRM analytics. Similarly, established datamining and OLAP firms such as Cognos and SAS are providing tools suitable for CRM analytics. Finally, a crop of young companies, including Broadbase, Digital Archeology Corp., E.piphany Inc., thinkAnalytics Corp., and Unica Corp., offer CRM analytical applications. The big CRM application providers–Clarify Inc., Clarus Corp., Peoplesoft Inc., and Siebel Systems Inc.–are starting to address analytics, but their offerings are rudimentary at this point. Peoplesoft, for instance, promises a full set of CRM analytics, but it won’t appear until the fourth quarter.

Vendors rush in

“A couple of the tools are pretty good, but for most, the best you can hope for is that they take over a lot of the dirty work,” says Bob Moran, managing vice president, decision support research at Aberdeen Group, in Boston. The dirty work revolves around capturing the data and getting it into a form that can be analyzed. Most of the CRM analytic offerings require technical skill, and users must be prepared to do some serious work to get meaningful results. At best, they provide some basic cubes or analytical models. Beyond that, you are on your own. A few, such as Unica Corp., promise actual applications–something managers can run almost out of the box. Almost, but not quite.

Star Tribune Direct, the direct mail operation of the Minneapolis Star Tribune newspapers, turned to Unica’s Affinium application, which offers a push-button model for CRM analytics, to analyze newspaper subscriber data, zip code address data, and demographic data to create effective direct mail campaigns for its clients. “With Affinium, we just have to set up the criteria for the campaign,” explains Don Poepping, director, direct marketing operations. The application handles scoring and modeling and allows managers to run what-if scenarios. As a result, the direct mail group is able to put together complex campaigns using newspaper, direct mail, and telephone, hitting each person in the target market the way he or she prefers to be contacted.

Arc International, in Melville, N.J., sells glassware through retailers. It collects data captured by POS scanning from its retail customers and combines it with syndicated research provided by companies like ACNielsen Inc. But it ran into difficulties because of differences between the various data sources. “We had groups look at different data sources and come to different conclusions,” recalls Narsi Bodapati, director of business development strategy.

Often the mistakes occurred because of the time lag between when an event would show up in one or another data source, such as order and shipping systems, and customer POS data. This could easily result in faulty demand forecasting and inventory imbalances. To overcome this would have required an enormous data warehouse integration effort, but “our IT group was already overtaxed,” Bodapati explains.

Instead, the company turned to thinkAnalytics, which enables Arc to analyze patterns in data coming from multiple data sources. “This is very complex, but thinkAnalytics was able to do it,” Bodapati reports. Setting up the linkages between the data and mapping the different data sources took considerable planning. Bodapati’s team had to be particularly careful, for instance, to match product numbers to ensure the system was tracking the right item end to end. The effort took just a few months, and Arc expects to collect its payback within six months.

Customer analytics as company strategy

CRM analytics is taking on increased importance as CEOs everywhere shift focus to the customer. In a recent Compass America Inc. survey of 400 CEOs, the most important IT area for impacting competitive advantage turned out to be building customer relationships, with 38% ranking it among the top three. Exploiting new technology and supply chain management rounded out the top three at 33% each.

But as AXA Financial discovered, a customer focus involves more than implementing CRM analytics. “CRM is really a strategy consisting of a number of components,” says Sibigtroth.

CRM analytics becomes just one piece of a larger customer-centric IT effort. IT must provide an entire infrastructure to support a customer-focused strategy, including data warehousing, datamarts, datamining, OLAP, modeling, data integration, customer interaction, and CRM itself. Only then can the organization capitalize on CRM analytics.

Alan Radding is a freelance writer specializing in business and technology. You can contact him at [email protected]

Reprinted from Datamation.

More Posts By Alan Radding