Piercing the Veil

By Lauri Giesen

As banks make more use of outsourcers, they must improve their access to customer data controlled by these third-party entities.

Financial institutions increasingly are embracing outsourcing as a way to improve efficiency, capture external expertise and free themselves to focus on core competencies. Outsource providers not only manage back-room technology and processing operations, but they also develop the credit cards, mortgages and investment products that banks then market to their customers.

But as this trend has proliferated, so too have problems in coordinating the flow of customer information between institutions and their external partners. First Tennessee National Corp., for example, works with seven different annuity providers. To access all its data on annuity customers, the Memphis-based bank had to develop standard information requests for all seven providers.

"Getting a full view of our customers is becoming increasingly difficult when that information is spread out among so many different outsourcers," says Dave Miller, senior vice president of the strategy consulting group, which is responsible for managing First Tennessee's customer databases and getting that information to the marketing and customer service departments.

Such episodes can throw a monkey wrench into service, sales and marketing. Retail banks, for example, want to know as much as possible about their customers — such as relationship profitability and demographic profiles — so they can market additional offerings to them in the most effective manner possible. It's difficult enough to integrate the customer data contained in their own files distributed across multiple units; the problems are magnified when that information has to be filtered through an outsourcer.

The problem is partly political, in that some outsourcers are reluctant to share the portions of customer data that they control. The bigger issue, however, is that outsourcers often don't format their data the same way as their clients, nor do they update it with the same frequency. Reformatting the information requires additional work from them — and for the bank as well.

One implication is a crying need to improve the groundwork for outsource deals. Data-sharing procedures should be written into service contracts, and in great detail. Institutions "need to get as specific as possible about what data they want and when they want it," says Kathleen Khirallah, senior research analyst at TowerGroup. "They can't always go back later and change the rules."

In the multitudes of cases where players rushed in without doing all of their homework, revisions of operating standards and even some contract renegotiation may be necessary. Since such work can be costly, managers will need to carefully examine each situation to weigh whether the customer information desired is worth the additional expense and trouble of acquiring it.

Proprietary Data

While banks have always possessed great reams of customer information in their files, they weren't able to pull all this data together for marketing purposes until the advent of high-speed computers and information storage technology. Data gathering and management has now become an integral part of banking, at least for the larger institutions.

"A few years ago, it might have taken three to four weeks to gather and cleanse the data, and by that time, a lot of the information would be obsolete. Now the same project can be completed overnight," says Jay Desai, founder and senior principal with Chicago-based Knightsbridge Solutions LLC, a consulting and systems integration firm.

Banks still face a serious challenge in assimilating data from disparate sources, however. This is a particular problem for large, multi-regional banks that grew through acquisition and now have to deal with multiple legacy systems. In that sort of situation, the mortgage unit typically formats its customer data differently than the credit card unit, which may be on a different system entirely from the branch system, and so on.

These problems are compounded when customer data is provided by outsourcers. A particular outsourcer may not only format the information in a different way than the bank client, but also may consider some of it proprietary. "It can be a political issue," says TowerGroup's Khirallah. "Some processors are saying, 'This is our data and we don't want to hand it over.'"

In most cases, however, the data that the bank wants is either more detailed than what the processor is accustomed to providing or not updated as frequently as the bank would like. This latter problem is particularly prevalent with insurance and investment products, where data typically is gathered on a monthly or quarterly basis.

The discovery of such problems in the course of an engagement creates pressure to revise complex arrangements that are already in place, oftentimes provoking resistance from outsource providers.

"Service bureaus become accustomed to providing data in certain formats, so what they are trying to fight now is changing how they present that data," says Charles Bruney, senior vice president with Speer & Associates Inc., an Atlanta-based consulting firm. "They may be reluctant to meet the bank's request if they have to go back and make an extra effort."

This is not a comment on attitudes, but rather on economics. "Most processors know that getting this data is critical to us and they try to help," says Miller at First Tennessee. "The only time we get much resistance is when our requests create a lot of extra work for them."

Credit cards present a good example of the type of information banks would like to receive from outsourcers, but can't always access. When banks outsource their card processing operations, they typically receive account summaries from the processors. These summaries may not contain information about customers' patterns of making late payments, however, or which customers are slowly increasing their revolving balances. Both of those behavior patterns might indicate that those customers are ripe for other credit products, such as debt-consolidation and home-equity loans, which is useful marketing information for the bank.

Banks would also like to know which products a customer is purchasing with the card. A number of big purchases from home improvement stores, for example, could indicate the customer is receptive to a home equity loan. A slew of auto repair bills might suggest a car purchase down the road.

"There is a lot of cool stuff in the credit card files," Miller says. "If we can match some of the purchase behavior against the demographic information we already have on our customers, we can get a real good picture of a customer's lifestyle."

Fortunately for First Tennessee, its card outsourcer, Columbus, Ga.-based Total Systems Services Inc., does provide the needed information. But other banks struggle to get it, experts say.

Data Integration

Beyond prevailing upon outsourcers, banks can do some things on their own to cure data disconnects.

One technique is to use a centralized clearinghouse to funnel information to the bank. First Tennessee, for example, offers its customers investment products from dozens of providers. These outsourcers formerly submitted their data separately. Then, First Tennessee hired a clearinghouse that gathers the information from all the brokerage houses and mutual fund firms before passing on one file to the bank. The clearinghouse does not handle annuities and other insurance products, however.

Most of the top 50 banks in the U.S. are currently working to develop systems that can consolidate and analyze the data supplied by third-party processors, according to Desai. Some of the mid-sized and smaller institutions, by contrast, are using the processors to do that work for them.

Milwaukee-based Metavante Corp., for example, consolidates much of the data held by other parties with its own reports before handing over a complete customer profile to the bank. Typically, Metavante holds information about the core demand deposit accounts of a bank's customer. It then imports information from brokerage, insurance, credit card and mortgage processors.

"About 80% to 90% of the data that banks want is common data and we can get it using standard templates we have developed. In the other 10% to 20% of cases, we have to do customized work," says Steve Lane, Metavante's CRM strategy manager.

New York-based Bisys Group Inc. likewise introduced a relationship management solution that gathers and analyzes customer data held by other processors for banks. Participating banks must use Bisys for core processing, but the non-demand deposit accounts can reside with any other processor, according to Nanci Moore, senior vice president and product manager.

Whether done internally or outsourced, the up-front cost of data integration is not cheap — up to $3 million in one-time expenses for hardware, software and program development for a larger institution, according to Desai. But if a bank uses the compiled information to improve its cross-selling and to gain greater efficiencies — such as consolidating mailings that go out to the same customer — a bank can recover its investment within a year, he adds.

The best solution is to head off trouble before it starts by writing information-sharing procedures into the outsourcing contract. TowerGroup's Khirallah recommends that marketing and customer relationship management executives determine in advance the specific customer information they need for upcoming projects. They should also determine how timely that information needs to be — real-time, next day, weekly, monthly, or quarterly.

Absent such preparation, institutions must resort to less satisfactory post-facto fixes. SunTrust Banks Inc., for example, is notified only once by the insurance company representing the policies it sells that an annuity was sold by one of the bank's representatives. It does not receive any updates, even if the customer changes the terms of the annuity. "We would love to get more data, like knowing if the customer increased the size of the annuity later on, but our current contract limits what information is passed on," says Greg Holzwarth, senior vice president and director of customer information management.

Atlanta-based SunTrust is therefore considering changing the established processing contracts when they come up for renewal to allow the bank to get more data. "Presumably, it is going to cost us more, and we have to be able to present our case internally first before we can start making changes in the contracts," Holzwarth says.

As the SunTrust example indicates, banks need to think carefully before making contractual changes. While additional customer information may be helpful, there's a difference between need-to-know and want-to-know, according to consultant Desai. "Do you really need to know the customer's trading activity for the past 24 hours or what stocks the customer is trading? If the information needs to be updated, fine. But don't be requesting real-time data if it's not necessary."

Indeed, First Tennessee's Miller says it's often not worth the cost to get the most up-to-date information. "We try to balance the need for timely data against the cost of getting it. Everyone talks about real-time data, but the reality is that in most cases, last night's data is good enough. And for mortgage and investment data, monthly reports are often good enough."

In any case, the best cure for data disconnects is preventing them in the first place. That's something to remember as the tidal wave of outsource deals rolls on.

Ms. Giesen is a freelance writer based in Libertyville, Ill.

Copyright © 2003 by Banking Strategies, published by BAI.

back to top