SUBSCRIBE

The Human Touch

Technology is just one factor among many in successfully implementing customer relationship management processes across your organization

By Justin Kestelyn

“ Customer relationship management” (CRM) may be the three most motivational words in business and IT since “Year 2000 problem.”

With customer acquisition and retention costs apparently dictating the rise and fall of brick-and-mortar and dot-com businesses alike, no organization can afford to treat customer relationships as fungible commodities. But although the back-end technology integration challenges associated with implementing CRM and e-CRM processes are well known, those involving the front office — including cultural acceptance and IT-business alignment — are not.

As senior VP for customer marketing at U S West until last year, Dennison DeGregor was the chief strategist and architect behind that company’s CRM environment, World-Class Information Strategy for Data-Driven Marketing (WISDM), a pioneering system that integrates one-to-one campaign management (via Xchange Inc.’s Valex software) with data mining (built on SAS Institute Inc.’s Enterprise Miner) to support closed-loop business intelligence. Now, as VP for enterprise CRM at Merrill Lynch, DeGregor is responsible for developing the technology processes and data infrastructure that will support that company’s newly inspired cross-channel client relationship, marketing, and service strategy. He also has extensive database marketing experience from previous stops at Citigroup and Diners Club International.

IE:U S West was among the first companies to successfully bridge both halves of the CRM equation (back-end data analysis and marketing automation) in a production environment. Based on your experience there, what are your insights into making that equation work?

DeGregor: At U S West, we developed WISDM from the ground up as an automated, enterprise-level, CRM system that would be cross channel in nature. As a result, I gained a very important lesson from that experience: that in any CRM or e-CRM effort, the first thing you have to do is take a strategic view of the entire enterprise. That’s the biggest insight by far—implementing CRM is not just about deploying technology; it’s about integrating strategy, process, technology, and people in a comprehensive change management process. But even before that, you have to focus on client needs by taking a client-centric, client-in view of the enterprise. In contrast, if you take a technology-driven view, you’ll set your CRM strategy up for failure.

The documented failures in the industry, whether first-generation database marketing or attempts to migrate to a fourth-generation CRM model, consistently fixate on a technology-driven approach. They don’t start with getting the “what” right—specifically, what is right for the customer, what addresses the client’s needs, what customers should be targeted based on their profitability and their potential profitability, and what is the retention risk associated with that customer relationship. After you get the “what” right, then getting the right technology to support those business processes—the how—becomes straightforward. But rather than getting the business requirements right, many firms get hung up on answering technology-driven questions first, such as how do we deploy it? What is the right hardware platform? What is the right software? That’s classic cart- before-the-horse thinking.

You must start with a customer-driven perspective and then reverse-engineer your technology requirements to support those customer-facing processes, instead of the other way around. That approach will lead you to the correct long-term technology decisions.

IE: Can you provide an example of how that strategy came into play at U S West?

DeGregor: The business problem at U S West was that we needed to grow revenues within a shrinking client base, and the only way to do that was to increase share-of-wallet among existing customers. Therefore, we had to analyze our cross-sell rates for each individual household on a longitudinal or continuous customer marketing basis. In order to grow share-of-wallet in a positive manner, we needed to understand our marketing efficiencies and effectiveness against each individual customer.

Setting that business goal led us to the fact that we needed a longitudinal type of customer management technology—in that case, Valex. That’s an example of a business requirement leading to the correct technology decision.

IE: What are the broader corporate culture issues involved in reorienting business processes around clients?

DeGregor: In terms of change management in corporate culture, U S West is a case study in the extreme because we went from a regulated rate-of-return P&L environment to one in which customer profitability became a big issue. That cultural shift—getting people who work in a monopoly environment to think about customer-centric marketing and service—was like turning around an aircraft carrier. The change management effort required to transform an internally focused corporate culture into a customer-focused one was broad in scope as well as timeline.

In that environment, the only way we could succeed was to take a two-pronged approach. First, we sparked a CRM “wildfire” with our front-line employees by launching a grass-roots educational movement. Then, we came “over the top” with a series of senior management buy-in sessions and worked to bring those two efforts together in the middle. The result was that we achieved a lot of organizational consensus about which course our CRM aircraft carrier should take.

IE: What about cultural shock in the IT group?

DeGregor: Cultural issues pervade the development process as well, such as persuading the systems organization bureaucracy to get CRM functionality deployed quickly by adopting world-class, leading-edge tools—as opposed to building everything in-house. A rapid application deployment methodology is essential (vs. a “waterfall” approach). And getting the end users trained, up to speed, and ready to exploit those advanced capabilities was a steep learning curve. A lot of challenges were there just on a day-to-day tactical level. But we would never have been able to get through those barriers without alignment with the senior executives.

IT cultural change is a change management case study in itself. In most companies, CRM efforts often never get off the ground because they encounter such stiff resistance from IT. In some of those failed efforts, CRM is proclaimed a new cultural initiative. Similar to the TQM stuff in the 1980s, it sounds great in theory. But what happens in practice is that management encounters so much resistance to implementing CRM as a large change management initiative that it just fizzles out.

Think about it: the ability to transform a firm to a true customer-centric operating model that is enabled by CRM technology is very threatening to traditional product silo groups. CRM technology enables an organizational structure where the product silos should become product factories or manufacturers, whose new organizational role is to feed product to a marketing group who is in turn responsible for moving the right mix of products and services to customers through multiple channels in a way that maximizes profitability. Therefore, by definition, the CRM executive is a change agent. And, whenever you try to bring change to a corporation, people feel threatened because their day-to-day lives and careers will be affected in ways they can’t anticipate—and uncertainty drives unpredictable behavior, for sure.

As I’ve learned, the best way to coun- ter that resistance is to spark a grass-roots movement, because the people who are going to do the work will become evangelists for CRM. When you combine that movement with the top-down support from senior management, you can pull it off.

IE: After a slow start, many of the traditional financial-services companies, such as Merrill Lynch, are now taking the e-trading channel very seriously. What are the business goals of your “e-client” relationship management strategy there?

DeGregor: For confidentiality reasons, I can’t speak about Merrill Lynch specifically, but it’s common knowledge in financial services that customer lifetime value (LTV) is the primary influence here. My experience at Citigroup confirms that view as well.

The traditional definition of LTV is fairly generic across all financial services firms: the number of products, the size of those product relationships in terms of revenue generated, and the duration period of those product relationships. Those are the three main drivers of LTV, although many other drivers exist as well—such as the number of referrals those customers give you that you can track, and the “cost-to-serve” based on how they interact with you. Duration of the customer relationship is perhaps the most crucial driver of all, so maximizing LTV is mainly about the economics of retention; it’s only partially about cross-selling.

One practical difficulty of arriving at a true LTV is obtaining the actual cost-to-serve. Most cost methodologies today are based on top-down cost allocation, which gives you a false LTV. Ideally, you would do transaction-based costing and get an actual cost of serving that customer in all channels.

The relevance to e-CRM here is that, even though IT and new distribution channels are changing the landscape, customer economics stay the same—whether you’re in the direct-mail channel or the e-channel. And the fundamental economics are that it costs seven times more to acquire a new customer than to retain your existing one, and that a dollar of revenue from an existing customer has a higher margin than a dollar of revenue from a new customer because acquisition costs are already amortized. Because of these issues involving basic economics, customer retention rates and the associated tactics really become the primary components of an e-CRM strategy.

IE: Let’s turn to e-CRM technology strategy briefly. Dynamic scoring—which lets you model and score individual customer records, as opposed to entire databases at once—was an important part of your environment at U S West. Can you discuss?

DeGregor: What we wanted to do at U S West—a goal that eventually gave us a lead relative to our competitors—was to migrate to a realtime marketing engine as quickly as possible. The core of that engine was a statistics-based dynamic scoring system. The idea was that when you log into a session with U S West, whether with a human or in the electronic channel, we would take additional information from you in real time and then respond within that session with an appropriate cross-sell offer.

For example, if you were to give us profile information regarding the number of children you have, which is predictive of telecommunications consumption, we could revise that cross-sell or retention offer in-session. The result is realtime customization of offers. That’s in contrast to the static approach in which a company would do a batch process every night so that any business rule associated with customer marketing or service would be “in the can” for the opening of the next business day.

IE: Thus, dynamic scoring is a key enabler for personalized marketing over the Web.

DeGregor: Yes. For example, at U S West we wanted to begin the process of migrating to a Web-based permission marketing model. We wanted customers to self-identify that they wish to receive offers and then provide profile information that let us engage in a learning exchange about that customer’s unique needs, which is the approach taken by everyone seriously engaged in e-CRM. At the core of that approach are behavioral scoring engines that analyze clickstream data and combine it with other back-office marketing analytics.

IE: We’re seeing a lot of packaged solutions emerge that claim that capability. But I expect that you would agree with the premise that if you want to build competitive advantage on top of those processes, you have to do it in-house.

DeGregor: Yes. I don’t believe there is a generic behavioral scoring engine available today that could satisfy even 50 percent of the complex targeting needs at U S West. There are other technologies that are easier to use that will give us 80 percent of what we need. But the mechanics of clickstream analysis—and the richness of the products and services that customers can click through—are so unique that we have to develop behavioral scoring engines on a proprietary basis. There’s just no package out there today that could get us there.

Of course, an enterprise package or solution is very appealing to IT—the lure of just plugging it in and developing a few APIs seems irresistible. But the downside to that approach is the package probably won’t deliver the business functionality you need. If you engineer the data models correctly initially, you’ll be OK when you integrate best-in-breed applications.

IE: No discussion of CRM would be complete without a question about fair information practices. What trends do you see here?

DeGregor: The privacy issue is the largest single issue facing every company using customer data today—not just financial-services companies, but all of them. Recently a federal judge ruled that Trans Union Corp.’s Performance Data division cannot sell certain data and that it is illegal for a company to acquire that data as well. [See “Privacy Watch” in News & Analysis, April 28, 2000.] That’s a huge event for the industry.

The lack of self-policing has resulted in legislative scrutiny, so we’re going to go through a period in which marketing practices are under a microscope. But with a couple of fundamental changes to data security practices, such as practicing permission-based marketing with customers, we will eventually reach a consensus about what is acceptable.

But because a lot of mistakes have been made, in the short run we’re going to see restrictions on the sale of data. If these restrictions continue, we’re just going to have to get smarter and smarter in the ways that we leverage the data available to us. That’s what we’re paid to do.

Rate This Article Rating: 1 (best) 2 3 4 5 (worst) Comments: Optional e-mail address:

RESOURCES