The E-Business Software Weekly is a series profiling trends and developments in software and applications that support e-business, the Internet, and other electronic communication channels. Look for a new story each week in this space.
Making Customer Relationships Work, Part 3
There is an old saying, often applied to contract negotiations, that is particularly relevant to the implementation of a customer relationship management (CRM) program: "the devil is in the details." Perhaps more than any other corporate IT initiative, CRM programs must be grounded in a large-scale strategic vision that sets forth the effort's corporate purpose, goals, and success metrics. At the same time, however, because CRM programs typically cut across corporate organizational lines and usually involve highly complex business, technical, and database components, the success of any given CRM implementation can depend heavily on the details of how well this multiplicity of interconnecting elements is pieced together.
As mentioned in last week's column, I had the privilege a few months ago to conduct an in-depth survey of the "lessons learned" from the early years of implementing customer relationship management initiatives. Based on writings in the popular press, I had expected to find a mature management field in which certain key principles were well-established, not just philosophically but in practice as well. What I discovered was a great deal of hope and optimism, but a much more limited inventory of concrete and measurable successes. While some companies had accomplished a considerable amount with their CRM programs, many others were still struggling to meet their programs' goals-or, worse, were stalled at the starting gate, not yet able to get their programs up and running. The potential for great achievements was certainly there, but in most instances had not yet been realized.
The conclusion that this evidence supported, in my view, was neither one of failure nor futility. Rather, it was that customer relationship management-like most other e-business technologies-is no panacea to corporate competitiveness challenges. It is, rather, a potentially powerful tool for promoting both customer loyalty and bottom-line profitability, but only if well-understood, well-implemented, and well-maintained. And it is one that ultimately can generate the quantifiable and concrete successes desired if companies now undertaking CRM programs apply the very instructive "lessons learned" from the early efforts of their corporate counterparts.
In the previous column, I examined the first four of these lessons-the strategic and management lessons, if you will. They were:
Lesson 1: Think Strategically
Lesson 2: Put the Customer First
Lesson 3. Build Internal Support
Lesson 4: Manage the Implementation Effectively
In this column, I'll review three key organizational lessons-the details designed to keep the "devil" of CRM failure at bay. As before, I hope, at the end of this series, that readers will have a much better idea not just of the possibilities for customer relationship management programs, but of the often difficult but nonetheless necessary steps that must be undertaken if these initiatives are to truly be a success.
Lesson 5. Break Down Technical Barriers
"One of the major problems in implementing customer management systems is that the process rarely starts with a clean sheet of paper," says Merlin Stone, IBM Professor of Relationship Marketing at the Bristol Business School in the United Kingdom and a co-author of the IBM/QCi Scorecard, an in-depth 51-company survey that is one of the most exhaustive investigations of CRM program implementations ever undertaken. Such "technical baggage" is oftentimes present, he says, because many enterprises have developed their IT infrastructures organically over decades, adding systems to meet particular business needs as they arose. As a result, they now have systems in place that were installed in the 1980s or earlier running alongside systems developed in the 21st century. "The technologies and computer languages that these systems use are completely different," Stone explains, "and there are formidable obstacles that make it very difficult for these systems to communicate with each other." CRM software solutions have not always solved these problems. "Because companies over the past few years have introduced CRM systems to meet the needs of different channels or products," he says, "even these systems cannot [necessarily] share data easily."
Eric Ranta, Senior Product Manager for the Financial Services Group at Siebel Systems, the leading developer of CRM software, describes this integration problem as "the single biggest issue preventing the successful implementation of customer relationship initiatives," and it's easy to see why. As he explained to me earlier this year, a company may have an immense warehouse of data locked up in a back-office legacy system, "but if customer-facing employees can't get at it, that information doesn't deliver value." He noted, for instance, that data on portfolio trades is essential information for financial services relationship managers, but pointed out that this data is usually stored in a legacy database, and is often not accessible to relationship managers in real-time.
Vexing as such problems can be, they must be resolved if CRM initiatives are to prove worthwhile. The IBM/QCi Customer Management Scorecard puts the challenge bluntly: in order for customer management efforts to succeed, data simply "must be derived from these many incompatible systems and turned into useful information that can be employed to build a complete and intimate portrait of each customer." The challenge carries no small risk. The survey's researchers determined that such technical integration, with its associated data migration and integration requirements, can consume up to 50% of a project's time and cost and often can prevent projects from even getting off the ground.
ABN AMRO Bank N.A., the largest commercial bank in the Netherlands, struggled with just this issue early in its customer relationship management efforts. As a consequence of having multiple, nonintegrated systems that handled various customer interactions, Siebel Systems CEO Thomas M. Siebel relates, a customer might be required to speak to several individuals before getting a service request resolved. So dissatisfied were customers by this situation that some of the bank's largest customers were complaining directly to the company's president, while others were leaving the firm altogether. By implementing an integrated technical CRM infrastructure, the bank made it possible for customer service personnel to view customer transaction data in real-time. "As a result," says Siebel, "the bank experienced increases in both productivity and customer satisfaction, and has also improved its ability to cross-sell its products to its existing customer base."
As it turns out, a growing number of companies have joined ABN in recognizing this problem over the past several years, and have begun implementing the kinds of technical infrastructures that are conducive to effective customer relationship management. Some of the techniques that have proven most successful include:
Ensure that the IT infrastructure serves, rather than drives, the business model-that is, that the infrastructure is designed with specific and measurable business objectives in mind instead of being engineered in isolation.
Create a centralized database. Even if distinct CRM modules exist in different organizational units within a corporation, these modules each must be linked to and draw from a core information repository that maintains all customer data.
Establish consistent data definitions. Though a seemingly minor issue, if the data strings representing identical variables are defined differently within disparate CRM modules, it will be difficult if not impossible to correlate the data at the customer level, even if all of the data is stored in the same centralized database.
Employ standards-based technologies. Modern standards-based computer languages, like XML (extensible markup language), can enable vastly different systems to "talk" to one another with minimal IT integration.
Lesson 6. Break Down Organizational Barriers
The first modern customer relationship management applications, put in place beginning in the early 1990s, were generally "siloed" applications, like sales force automation (SFA) and customer service and support (CSS) systems, that served individual departments within a corporation. These silo-based solutions were a natural response to then-existing corporate organizational structures, since individual departments typically acted with a substantial degree of autonomy, each owning a small "slice" of the corporation's relationship with the customer. This pattern was especially pronounced in sectors like financial services, in which firms were typically organized along rigid product-centric lines (depository accounts, loans, investment accounts, institutional products, and so on).
Whatever its short-term organizational benefits, this silo-based approach to business operations, even when aided by advanced CRM tools, caused all kinds of customer relationship difficulties. "Due to clumsy handoffs from one system to another, critical data failed to get passed along or was missing," explains Stephen Shaw, executive editor of The CRM Journal. "Customers were treated inconsistently because no one was working from the same information base." Customers were understandably offended. As Shaw asks: "What could be more annoying than having to repeat the same complaint to different service representatives? Never receiving the product literature you asked for? Failing to hear back on a service request. And never getting a letter of apology for a billing error?"
The solution to this problem, appearing in the mid-1990s, was a suite of cross-functional or enterprise-wide customer relationship management tools-software that delivered what has popularly come to be termed a "360-degree view" of the customer. These enterprise-wide solutions proved ideal for managing all customer touch points. Such solutions, says Stephanie Hahn, General Manager of Customer Relationship Management Solutions for IBM, "merged the interactions with customers into a unified database and made the resulting data and analyses available to company participants across all customer channels." As a consequence, she explains, customer relationship management "at last became a tool for viewing and managing the company's entire relationship with a customer, not just bits and pieces of it."
Or so one would hope. Unfortunately, in implementing CRM initiatives, practice has not always measured up to the promise. In a great many firms, customer relationship initiatives remain heavily line-of-business focused, built around isolated silos of information-a situation that can be a significant roadblock to success in managing customer relationships. The IBM/QCi study, for instance, revealed that silo-based and line-of-business focused organizational structures "positively hindered" enterprise-wide customer relationship management in 59% of companies surveyed.
Neil Woodcock, an investigator at QCi, the British industrial research firm that co-led the IBM/QCi Scorecard, probed this organizational problem as part of the survey, and uncovered a number of reasons why silo-based and line-of-business focused operations persist despite efforts to implement enterprise-wide customer management solutions. One prevalent cause, he notes, is internal politics, such as a simple refusal by one department to alter its preferred view of customers in favor of the corporate view. A second cause: the inability of IT departments to keep pace with the company's proliferation of products and channels. Faced with delays and difficulties in securing IT solutions to support their new offerings, managers who are responsible for the new products and channels often declare independence from the corporate customer relationship management program and set up their own CRM database that is not well-integrated with other corporate operations. A third important reason is the non-compliance of customer themselves, ranging from their forgetting key data items (e.g., PIN numbers), to making errors in keying data in, to simply refusing to cooperate.
All of these are serious problems, many seeming to defy resolution, and they bring with them some very serious implications. As Paul DeVriendt, Senior Director of Institutional Finance Products for Siebel Systems, lamented to me earlier this year, without a full, 360-degree view of the customer, you "can't fully understand customers or their value to the firm. And you can't service customers well if you don't know everything about them. This," he explained, "ultimately has an effect on the bottom line: you can't sell more to a customer if you don't know what they need." For instance, "you may have fully satisfied their needs in your own line of business, but without a full customer view, you won't know enough to cross-sell into other lines of business."
On the other hand, companies that do manage to collapse these organizational silos soon witness the substantial economic value they have been missing. One major U.S. bank recently realized the power of this 360-degree viewpoint when it succeeded in implementing a truly enterprise-wide customer management system. In mining its cross-functional customer database, the bank discovered that 75% of all customers who eventually left the institution for a competitor held only one product for a full year before their departure. Armed with this information, the bank began routinely contacting customers whose product holdings fell, enabling the firm to retain customers it otherwise would have lost.
But effective enterprise-wide customer management solutions cannot stop with the breaking down of organizational and data silos. "Integration involves more than unlocking data that is imprisoned in legacy systems," says Merlin Stone. "Just as important is the ability to move current data around the business. Particularly now, with customers being offered and choosing to use a variety of touch points, it is important that some forms of data [be] updated instantly."
Alberta Treasury Branches (ATB), a locally focused major Canadian bank, became aware of this need in trying to sell its financial products in the field. "Our relationship managers had nothing but paper reports that were a few days old to prepare them for a client visit," says Ken Casey, ATB's Vice President of Corporate Services and Operations. Once they reached the clients' offices, "something inevitably had changed or a question was asked that they didn't have the answer to, which left them in the situation of having to say, 'I'll call you back.'" This problem disappeared, however, once ATB implemented Siebel eFinance, one of the few truly enterprise-based CRM solutions for financial services firms. ATB's new enterprise-wide view of its customers, reports Casey, "helps our representatives to be more knowledgeable, saves our customers time, and greatly improves our service."
One particular organizational challenge in the Internet age is the integration of Internet operations with more traditional customer contact points. For instance, say proponents, if a customer places an order on the Web and then telephones the call center later with a question about the order, the customer service agent should be able to view all the details of the Web-based order. Similarly, the customer should be able to view, via a self-service Web page, details of an order that was made through the call center.
IBM itself faced the challenges of integrating online and offline activities in the early years of its Internet operations. "We did not have a complete view of what a given customer was doing across the entire IBM company at a particular point in time," says Doug Maine, General Manager of ibm.com. "We had many standalone customer relationship management applications operating inside the company, which inhibited the effectiveness of sales and marketing activities, leading to missed revenue opportunities and lower customer satisfaction."
To address these problems, IBM decided in 1999 to implement a system that would integrate all customer information and customer-facing operations across the enterprise. One example: from within its customer extranets, IBM's Gold Service customers could click on a "call me" button that triggered an immediate phone call from a dedicated IBM call center sales specialist who knew the Gold Service customer's history and could access additional details on a real-time basis. Once they experience this fully integrated service through IBM's Gold Service program, Maine says, IBM's customers invariably show "a measurable increase" in the amount of business they do with the company.
Other firms have built their entire business model around the multi-channel capabilities that enterprise-wide customer management solutions make possible. A prime example: Charles Schwab. Unlike its Internet-only competitors, Schwab interacts with its customers through multiple channels-branch offices, the telephone, and the Internet. The channel relationships are actually quite complex: while 70% of new accounts come into Schwab through its 360 branch offices, customers execute nearly 90% of their stock trades over the Internet. At the same time, the company's call center handles more than 10 million customers per month. All told, some 95% of Schwab's customers use all three channels to interact with the company, making robust, enterprise-wide customer management solutions a must.
Lesson 7. Capture the Right Customer Data
Thomas M. Siebel, the CEO of CRM leader Siebel Systems, knows the customer relationship management process as well as anyone. And so it is significant that the first rule he sets forth in his article, "The Eight Essential Principles of E-Business" is this: "Know your customer." The real power of e-business and the customer relationship management processes that underlie it, he says, "comes from the ability to capture and leverage information in order to better understand the customer and, hence, better anticipate and serve customer needs."
As it happens, many customer-facing enterprises have little problem in capturing vast amounts of data about their customers. After all, many such companies, particularly in the financial services arena, know every transaction their customers make. While at first glance beneficial, this opportunity, notes Siebel Systems' Paul DeVriendt, can beget serious problems, as such companies may have a "tendency to capture too much information, especially during the early stages of a CRM deployment. Every transaction, every phone conversation, is a new piece of data to be logged and captured." This flood of data oftentimes "overwhelms end users," leaving them uncertain which, if any, customer data to record.
Such uncertainty about data-capture priorities was repeatedly observed by the IBM/QCi researchers. Among the 51 blue-chip companies they surveyed, only 11% had a comprehensive customer information plan in place that addressed such critical data-capture issues as information value, acquisition priorities, and information management and usage. Only 15% had developed a priority matrix for capturing customer information that would help to manage the customer relationship. And just 13% sought out and recorded a "reason for loss" code on the customer database for every known customer loss.
Even where the right classes of data are collected, the data itself can be so error-prone that it fails to enhance, and may even impede, customer relationship management. This is particularly the case with respect to historical data stored in legacy system databases. While these databases can be a rich source of customer information, explains Merlin Stone, "generally speaking, all of the different sources of data are likely to have been constructed at different times, for different purposes, and often using different technologies. Conflicts can be as basic as having the same data presented in different length fields in different databases." More seriously, as noted earlier, "data with the same meaning may be recorded differently in various systems, resulting in varying degrees of reliability. Some data may not even math their original document description."
The results of this "semantic drift" and "data mutation" can be calamitous. "Frequently," says Stone, "projects fail to deliver the anticipated business benefits, simply because the quality of the data they produce is too suspect to be useful. The problem is especially pervasive in the financial services industry. "Consolidating information into a single view of the customer is very challenging," Tim Angst, the Vice President and General Manager of Siebel Retail Finance, told me in an interview this past summer. "However bad we think data quality is going into a project, it's usually worse." Yet most companies are not well-prepared to correct this problem. In 75% of the firms surveyed by the IBM/QCi researchers, no individual bore any explicit incentive or sanction for maintaining customer data quality.
Perhaps the most serious threat to data quality, especially in the Internet age, is the behavior of customers themselves. As Louis V. Gerstner, chairman of IBM, said in a speech last March, when customers go to a Web site and "are asked to fill in their name, address, age, income levels, and all that, they say they're Albert Einstein with an income of $5 and an email address of E=MC squared. Worthless data." What are customers really saying when they do that? he asks. "They're saying they don't trust the security of the site, and they don't trust that the owner of the site is going to respect their privacy and not abuse or sell their personal data." It's a confidence issue, not a technical issue, and, says Gerstner, it "is not going away."
Such challenges as these are among the most difficult facing companies seeking to better manage their customer relationships. But there are ways to address them. Some steps are obvious. As suggested earlier, establish and enforce consistent data definitions from the start. Put in place procedures to ensure and enforce data integrity. And decide which data need to be captured and retained-and which don't.
Beyond these more technical fixes, companies also need to give both their customers and their customer-facing employees the appropriate incentives to supply and record the desired data. Putting themselves in the shoes of customers and frontline employees, CRM system developers should ask themselves such questions as: "Why should I supply or capture the requested data? What's in it for me? How do I get more out of the process than I put in?" If customers feel they have something to gain by supplying the data-more savings, more responsive service-they will usually comply. And the same goes for the employees who administer the system.
Finally, once these procedural issues are resolved, the customer relationship management system has to make it easy for employees to accurately collect the data the company needs. Country Companies Insurance Group of Bloomington, Ill., found this requirement lacking in its data-capture protocols prior to its implementation of a new CRM system last year. Although the firm's 3,000 agents had ample incentive to capture customer data, they would track calls on paper and then type the notes into a database, resulting in numerous errors and considerable losses of valuable data. After implementing an enterprise-wide CRM solution, however, data-collection became a seamless, one-step process, as data could be entered directly into the database the instant it was acquired. The payoff, reports Christi Smid, Supervisor of Country Companies' Client Acquisition Center, was greater "speed and accuracy that will result in better service for our customers and more sales for us."
Moving On
Mastering the organizational details of a CRM implementation is one of the most challenging, but nevertheless most essential, aspects of a customer relationship initiative. But as with the initiative's strategic and managerial context, it is only part of the solution. Equally important are the implementation's technical parameters-a final area of consideration that I will explore in the concluding column of this series next week.
ADVERTISING FEATURE
