The view from inside MetLife is just as admiring, but managers are realistic about the challenge. "It's a dramatic change," says Bob Marzulli, MetLife vice president of information technologies, who heads the new technology plan. "The business units involved were dedicated to different lines of business and different products. They had a vertical perspective. But we had to tell them to take a much broader outlook to provide the customer base a wide range of services, not just a few related ones.
The CIF network began to take shape in early 2000, when a team headed by Bob McKinney, then MetLife's Individual Business CIO, was given the task of instituting a central client file system to replace multiple files within his unit. IBwhich includes life and property insurance, annuities, mutual funds and financial planningis responsible for $12.4 billion in revenue in 2003, about a third of the company's sales. McKinney's group discovered that each product line had its own administrative system and none of them was related. If a customer had two life insurance policies, an auto policy and a mutual funds account, he had four different client files of information, one for each contact.
The solution seemed obvious to McKinney's group: A single client file should be the operational system of record for each customer. Separate business lines could access this file each time they needed to manage or update the account, but the client's basic pertinent informationincluding any combination of name, address, demographics, assets, account activity and the likewould be centralized and transparent throughout the organization.
But a problem immediately arose. There was no software available that seemed capable of handling such a vast application, especially because the customer files MetLife proposed to centralize were located on many different programs and written in any number of database formats. Eventually, when the new technology is implemented throughout the company, it will have to manage all customer records, including more than 10 million life insurance plans, 37 million employee benefit programs the company administers for U.S. corporations and more than 2 million automobile and homeowner policies. MetLife had a few CRM programs scattered about the company, but they simply weren't up to the job.
The answer was found when Bob Shafto, at the time the CEO of New England Financial, told Tony Candito, the CIO of IB and Marzulli's boss (he's since left the company), about a recent customer data integration application developed by DWL Inc., an Atlanta-based company of which Shafto is a director. The software is a client information hub that overlays existing CRM, ERP and other databases and acts as a so-called single authoritative source for client filesa "golden copy"for front and back-end corporate systems. In essence, CDI software is an added layer of intelligence that not only supplies a single, readily accessible location for customer records, but also turns these files into potential transactions. For instance, when an address is updated on a policy, CDI can alert salespeople of the opportunity to sell a homeowner's policy.
Says Justin LaFayette, DWL's CEO: "MetLife was skeptical about whether our program, which was only being used by small companies with a limited number of files, could handle the millions of records MetLife had." This doubt was alleviated when DWL agreed to help MetLife customize its CDI program to match the insurer's needs, on the condition that DWL would be allowed to sell the newly developed application to other financial services firms.
Meanwhile, Benmosche threw another curve ball: While the plan was to make CIF an enterprise-level initiative, most of its funding would not come from corporate coffers. Instead, each business group that would be using the application would have to pay for its implementation. It would come out of the IT budgets of the middle managers and business unit heads in the Individual Business group. "If you fund it from the top, no one has a stake in it," says Marzulli. "But if your skin is in the game, if your dollars are paying for it, you're going to be more involved."
Another implication of the funding policy: It forced cultural issuesany resistance to the project because it clashed with the traditional way of doing businessto be addressed before the effort began. If there was no financial pain attached to installing the system, then those who viewed it as an intrusion on their work activities would treat it with the corporate equivalent of passive-aggressivenessletting it be built with no intention of ever cooperating with its development and implementation. With dollars on the line, though, sales agents and their managers, who tend to dominate the culture at MetLife, would be more public and vociferous about their disagreements.
The biggest source of friction had to do with the implications of data sharing, especially when it involved cross-selling. Like many insurance companies, MetLife's agents are used to a strong degree of independence; historically, it was their sales performance that dictated their compensation and value to the corporation, not how well they shared data and promoted additional customer sales by other business units. Some sales managers argued that to close deals quickly, customer records maintained locally were more useful than centralized ones, primarily because local files served as an extended Rolodex. Salespeople put their personal insights about clients as well as tidbits of information about prospective customers in these records. Why share this data with everyone else at the company? Why let their customers be inundated with offers for other MetLife products?
Some Individual Business unit heads, who believed that cross-selling could be a benefit both to MetLife and to their own departments, took an opposing point of view. But they were unconvinced that a massive, multi-year project estimated to cost between $50 million and $100 million could achieve better results than a smaller, more targeted updating of customer records.
|MetLife's Primary Business Lines|
Institutional Business | Group insurance, retirement and savings products to corporations
Revenue | $14.2 billion*
Individual Business | Life insurance, annuities, mutual funds and financial planning to individuals
Revenue | $12.4 billion
Reinsurance | Corporate reinsurance programs
Revenue | $3.2 billion
Auto and Home | Personal property and casualty insurance
Revenue | $3 billion
International | Individual and group life insurance, retirement and savings products, sold through ten subsidiaries worldwide
Revenue | $2.5 billion
*All revenues for year ended December 31, 2003
Source: Metlife Inc.
Another cultural obstacle involved the cozy relationships that had developed between the IT groups attached to specific business lines and the managers of those sections. Traditionally, when the life insurance group wanted to upgrade ways to sort client files, for instance, unit chiefs negotiated with the group's IT department to complete the project. And with this another moat was created in MetLife's database structure. With the CIF system, all these insular networks would be wired together and department-level IT groups would lose their control of the information and systems. "There was a lot of pushback," says Yehuda Gourvitch, vice president of Individual Business operations and services. "All of a sudden, central IT was more important than they were. So they had to think about how they could integrate with what we were proposing. It was changing their world.
Gourvitch, whose career includes stints as an IT manager as well as in strategic consulting, was given the task of breaking down the intransigence. Like most cultural issues at companies, Gourvitch believed there was very little room for discussion. In this case, Benmosche and the other top MetLife executives were committed to the data initiative and convinced that it was essential to improve the company's troubled financial performanceand no amount of grumbling from the business units would change their minds. But if Gourvitch presented it this way, he knew the result would be lack of participation and poor morale, and that, ultimately, the project would fail. It was his job to make the organization feel like everyone with a stake would have their say before final decisions were made.
So Gourvitch decided to attempt a campaign of old-fashioned educating and politicking. First, he created a training program and presented it in a series of road shows for middle and upper management in the IB group. The central theme: they had to trust the benefits of the new systemchief among them, the potential to help clients purchase more products; the ability to mine data in real-time to target new customers; and increased productivity, since the system would eliminate the need to enter new data every time a customer moved, changed account information, or signed up for another offering. And Gourvitch made the argument that there could be significant cost savings from this project. By centralizing customer data in a form that would make it possible to manage and sort without too much difficulty and with a high degree of certainty that the information is accurate and up-to-date, business units wouldn't have to reinvent the wheel by conducting costly one-off marketing and sales programs every time they launched a campaign.
Return on investment was another critical pitch. Gourvitch told the IB group that MetLife's top brass expected an ROI of about 15 percent from the project, and he asked each business unit to come up with scenarios for how that return could be realized. As for the one-on-one IT relationships, Gourvitch told the technology groups, "This isn't the time in any company to be in the way of creating economies of scale."
Gourvitch left the most compelling point for last, particularly because he knew that words alone wouldn't diminish the skepticism. He told IB chiefs that since it was up to them to fund much of the customer data initiative, management fully realized that a new round of budget discussions would need to be held annually, based on available funds. The hidden message: by funding the initial part of the project, Individual Business units were not committing to the same level of expenditures each year. If they weren't satisfied with the progress, there would be opportunities to reevaluate.
"That's the key to cultural changemaking sure people feel like they have a lot of room to maneuver, that they're not being forced into something," says Gourvitch. "And although I knew this project was critical to the company's successand I knew one way or another it was going to be implementedI felt initially it was important to plant the seed, get the system off the ground, and then let the chairman's customer-centric philosophy work its way through the organization and drive the project to completion."