A CDI strategy is especially relevant to mid-market CIOs who may not have the budget to buy proprietary CRM solutions or the time to invest in the typically arduous CRM implementation process which, according to Gartner's guideline for enterprise CRM roll-outs, can cost more than $US20 million over a three-year period. (Some CRM failures have run up to $US100 million in overall costs. See "AT&T Wireless Self-Destructs", CIO May 2004 for one disastrous example.)

The beauty of [the CDI hub approach] is that most organizations already have most of the pieces in place," Wang says. "They just need to find a way to pull it all together."

In the late 90s, CRM vendors promised that their software could give companies the ability to leverage customer data to boost sales. That software cost millions and took years to install, and yet at the end of those marathons many companies were left with tools and systems they couldn't or didn't want to use. Integration often was incomplete, data frequently dirty, and all too often companies had no guidelines for who would own the data or how it would be input and reconciled among systems. Eventually, business and technology executives became disillusioned with the enterprise approach. Many companies, large and small, turned to on-demand CRM, only to find out it also had problems with costly customizations and real-time integration challenges.

In a 2005 Forrester survey of 22 Fortune 1000 companies in North America, Europe and Asia, business and IT leaders voiced widespread disillusionment with their CRM implementations. Just 14 percent strongly agreed that their CRM applications had improved end-user productivity, and only 10 percent strongly agreed that they had achieved the business results they were expecting. CRM implementations "always seemed to over-promise and under-deliver", says Dolmovich. In fact, for many years UnumProvident's CIO forbade his IT staffers from using the CRM word to describe their customer data management plans because of the negative connotations attached to the acronym.

In the Forrester survey, executives acknowledged they were partly to blame for CRM's bad reputation. They confessed that they had not spent sufficient time on defining data requirements and managing data quality. In another survey, by Cutter Consortium, 64 percent of corporations admitted that they lacked a formal strategy for using the customer data they had spent millions to collect.

"When the company doesn't have rules and policies [for data], the data has been largely corrupt," says Anthony Lye, Oracle's group VP of CRM products.

The Importance of Business Ownership

The first step toward creating an integrated customer data system is to sit down with key business executives and ask them what they want. Do they want to focus on keeping the customers they have or on attracting new ones? Are they concerned more with decreasing lead generation costs or shortening the sales cycle? Once IT knows what the business side wants to achieve, IT can help the business identify which data sources are important and which are not.

Next, the business and IT need to agree on an information management policy: Who has access to what customer information and what can they do with it? How will they access that data? How will they make changes to it?

For CIOs, the key to success is making sure the business takes ownership of customer data. At AmerisourceBergen Specialty Group (ABSG), a $US7 billion pharmaceutical supplier, the mantra that "the business owns the customer data" has been critical to the company's CRM success, says CIO Dale Danilewitz. In 1999, when ABSG broke away from its parent company's systems, executives articulated what they wanted: more granular, reliable customer information accessible in one repository and accessible in real time. It was Danilewitz's job to make that happen. And although Danilewitz initially believed that an off-the-shelf CRM system might do the trick, he found that his business users' needs didn't align with what was on the shelf at the time. So IT cobbled together a mixture of applications and systems to form a home-grown CRM system, essentially a conglomerate of custom-built applications and vendor platforms and databases. In the centre, tying everything together, is a data warehouse that provides real-time and historic customer data, and is integrated with other data stored in ABSG's e-commerce applications, financial systems and customer data applications.

Today, Danilewitz says ABSG's system satisfies users from the sales, call centre and marketing sides. And because these business units understand the data's worth, Danilewitz says, they take pains to ensure that they don't add data that will "adulterate" their own systems. "The business users check the data, run reports on the data to make sure it's accurate, and run technical applications to check quality," Danilewitz says.