How to rescue CRM: even dysfunctional CRM systems may be well positioned for future success. The trick is to step back and think about your real goals.
CRM has become a senior-management issue because it consumes staggering amounts of money and, notwithstanding the success stories, has mostly proved a disappointment. Companies around the world spend $3.5 billion a year on CRM software, (2) and that is only a fraction of the total expense; implementation, training, and integration outlays can be three to five times higher. All in all, a highly complex CRM installation can cost more than $100 million and take three years to complete. Yet when managers were asked to assess three key functions of CRM--marketing-campaign management, call-center management, and marketing analytics--no more than 35 percent of the respondents said that their expectations had been met in any function. (3) Moreover, while financial-services firms are among the biggest users of CRM, only 20 percent of the US retail banks that implemented it have raised their profitability as a result, and CRM investments during the big push of the 1990s didn't generate the expected returns for the bank s that made them. (4)
You know you have a problem with CRM if, after a year, you are not seeing at least half of the financial benefit you expected-and most managers expect revenue growth of at least 10 percent from their CRM investments. (5) Two other warning signs that may be evident much sooner should also be taken seriously. The first is cost overruns or missed deadlines; it isn't unusual for these projects to run two or three times over budget. Overruns usually result from needless complexity, whose cost can swamp whatever benefits it might produce. The second warning sign--complaints by employees about poor usability or unmet performance promises--often emerges in the pilot or early-rollout phase. Such complaints may point to serious underlying problems.
Some companies live with the performance shortfalls of these systems or leave them largely unused. A few have scrapped them altogether. But our experience with dozens of companies, from banking and insurance to travel and logistics, shows that faltering CRM efforts can be turned around. The key to uncovering the root cause of the problem is to reassess the business goals of the system and the organizational and technical support it receives. Every CRM breakdown we have seen stems from some combination of poorly defined objectives and organizational and technical weaknesses. By examining each of these areas, you can diagnose the problem effectively and develop an appropriate remedy.
Troubleshooting the system
A CRM turnaround team composed of business and technical managers, and led by senior executives, should take charge of the effort. The cross-functional mix is important because attempts to make CRM work encounter hurdles not usually found in other software initiatives. On the organizational level, for instance, many frontline service employees, who are not always highly skilled, use CRM applications. On the technological level, CRM offers more choices and forces more trade-offs than do most enterprise systems. Such problems as frequent crashes may have their roots in something as basic as badly written code or as complex as overly ambitious business goals. The diverse membership of the team gives it the expertise to identify specific problems and the authority to correct them.
A failure to establish clear business goals before launching a CRM effort is the most common and important source of these problems. Without clear objectives, how can a company decide among the dozens of possible CRM initiatives? Companies that fail to choose may implement "big-bang" solutions intended to cover every conceivable business contingency. As a project grows in scope, the system's development can take on a life of its own, incorporating new features that don't support business objectives but do add complexity and cost. IT professionals well know that the bigger a project, the harder it is to integrate and the more likely to miss deadlines or to be scrapped altogether (Exhibit 1).
Since the group must understand the goals that a CRM system was intended to promote in order to assess its performance, the first step in any turnaround effort is to check the vision that originally inspired the program. Unless the IT department and business managers have jointly developed such a vision, the former will view the business goals as hopelessly vague or unrealistic, while the latter will view IT as a roadblock.
The basic process for troubleshooting goals is the same for any company. Let's consider what a bank might do. Exhibit 2 illustrates some possible business goals and ways of achieving them. To capture more revenue from existing customers, for instance, the bank might select four levers. It knows from experience that the best way to get more money from its existing customers is to cross-sell products, such as mortgages or car loans. Since this lever was a priority for the bank, the turnaround team would specify the performance indicators of successful cross-selling campaigns and work out their cost and potential effect. To simplify the CRM system, the team would then pick the indicator with the greatest economic impact--in this case, increasing the conversion rate: the number of offers to customers the bank can turn into car loans or mortgages. Economic analysis might show, for example, that a 10 percent increase in the response rate to a marketing effort could raise revenue by as much as 60 percent, whereas a 10 percent increase in the number of customer contacts would raise it by only 10 percent. Improving the conversion rate itself would entail several supporting initiatives: more precise targeting of prospective buyers, more attractive terms, or a better follow-up system.
Only when the ream had fully investigated all of the activities that could promote the company's strategic objectives would it be able to assess the requirements of the CRM system. In this case, a full assessment of the business goals and supporting initiatives would lead the team to identify four high-impact system requirements, including the implementation of propensity-modeling tools to predict who would be the most responsive high-profit customers and of an on-line history of the offers made to them, so that call-center agents could follow up the next time they made contact (Exhibit 3). The assessment is likely to reveal that some of the solutions--for example, a new sales script that improves the presentation of offers to customers--have nothing to do with IT. Common to all these solutions is an understanding, based on clear business logic, of the root causes of the problems.
As a result of this exercise, many companies will realize that their business objectives are poorly focused, so they will have to clarify their goals before contemplating the next stages of the turnaround: organizational and technological issues. The few organizations that don't uncover problems with their business strategies can examine the two other possible sources of trouble.
Organizational issues are important because the success of a CRM system depends on its users, many of whom have direct contact with customers. A few keystrokes, for example, permit a customer service rep at one US telecom company to check the calling plan and call history of any customer who complains of being overbilled for an international call, to authorize a refund, and even to suggest a more suitable calling plan based on the customer's calling patterns. Lack of training is a possible culprit in poor organizational performance, for to use such features, reps must be trained.
A compliant, user-friendly CRM system is necessary as well, and to build one, the designers of a system must work with its users. One computer chip company launched a CRM initiative in hopes of reducing the time its salespeople spent handling paperwork and thereby freeing them to spend more time in the field. On the theory that field service technicians would know how to make such a system work, they were asked to develop it and roll it out. When the system was deployed, however, the sales reps rejected it; they hadn't been involved in the development effort, and the technicians had burdened it with data entry and other tasks that proved to be cumbersome, time-consuming, and of little clear benefit to the sales force.
Misaligned incentives can also undermine CRM systems. A production manager at a tool manufacturer, say, is rewarded for meeting schedules and quality standards. Inquiries about an order's status might require the manager to get information from the shipping, inventory, or purchasing departments, and expediting one order could put others at risk. People will take the time to share information, and accept the risk of doing so, only if they believe that this course serves their interests. Unless the right incentives are in place, however, it may not; thus frontline users at some companies bypass CRM systems altogether, undermining their effectiveness. It is important to ask whether the people responsible for making a system work have the incentive--or at least no disincentive--to enter data into it or to answer queries. Incentives for shared goals, such as an increase in customer satisfaction or spending, are essential.
Perhaps the most important diagnostic question is whether the executives who originally decided to implement the CRM system hoped to use it to effect more change than the organization could really support. Even with the right alignment of incentives and training--and it is rare for companies to get both right the first time--radical changes in work processes can take time and effort to gel. If the culture and practices of a company won't support its stated goals, the answer should be intuitive: scale back aspirations and build the organizational capacity needed to achieve early results.
Like organizational problems, technological ones are hard to fix if a company doesn't clearly understand what it wants from its CRM system. Even if it does, the IT department can stumble.
The problems can begin with attempts to select the right CRM architecture and to make it serve the company's needs. Many vendors promise best-in-class applications for a single CRM process, such as customer service support or sales tracking, and this approach lets companies cherry-pick solutions and avoid needless features. Because data must be pulled from various sources, however, these applications make it hard to integrate and manage the system; the symptoms may include incomplete or incorrectly routed information, too many passwords or log-ins for users, inconsistent data models (the frameworks used to capture, access, and display information), and inconsistent guidelines for entering data in different units (thereby forcing customers to restate information when they are passed from one department to another). Any of these problems can frustrate employees who use the system, to say nothing of customers.
Market leaders such as Oracle, PeopleSoft, SAP, and Siebel Systems sell integrated suites, which provide for one-stop shopping, are easier to manage and integrate, and offer consistent data models for information about customers. Such features are powerful but may entice companies to deploy more modules than they need to accomplish their primary goals. These suites also may not mesh well with other applications (say, a traffic-management program in the call center) and could therefore lead to dropped calls, long queues, or an uneven distribution of calls, so that some reps would be overwhelmed and others underutilized. Companies experiencing problems of this kind can expect lower productivity and dissatisfied customers--and much finger-pointing among vendors.
Besides considering these architectural choices, companies should ask whether their IT managers are trying to deliver too much information across the enterprise. One potential benefit of CRM software is its ability to provide a universal view of the customer, with the same information (such as sales and service histories) available at every customer touchpoint--the Internet, call centers, retail locations, and employee desktops. In this way, a large company can serve its customers and prospects seamlessly and consistently. But a long and expensive development effort is necessary, and the results can be uneven.
Slow or unreliable system performance, for example, may suggest that unnecessary information, from too many sources, is traveling through the system; members of the turnaround team should find out, say, if a user request generates a ten-screen report when one or two screens would do or if information is going to more people than the system can really support. To balance the costs and benefits of such decisions, the team must identify the information, the data sources, and the features needed to achieve the small number of business goals that add the most value to the company.
Making it work
Once the full turnaround team has uncovered the system's business, organizational, and technological breakpoints, a senior-management group from within and beyond the team--including high-ranking executives, board members, and key department heads-should lead the remediation effort. Although implementing the exact fix may require the help of organizational-design specialists, systems integrators, or other outside experts, this group will be able to suggest the appropriate remedies. It must resist the temptation to go on developing a system simply because a lot of money has already been spent to do so. The point is to build up the pieces that could work and to drop the rest.
Many companies must narrow the scope of a CRM effort to improve its effectiveness. The private equity arm of a large bank, for example, had invested in several CRM initiatives in the vague hope that better information about its customers would help boost sales, only to discover that the system was barely used. A cross-functional team of business, IT, and sales managers decided to retain only those parts that would support one overarching goal identified in the diagnostic process: increasing the bottom line. The remediation team therefore set out simply to increase the number and profitability of products sold to the company's best customers. To that end, it discarded the CRM system as a whole but extracted a single data-mining application that captures customer profiles and demographics and tracks the profitability and estimated lifetime value of each product in a customer's portfolio. This data-mining program is less comprehensive than the original CRM system but supports the company's immediate goals.
The bank's private equity arm, having narrowed its goals, used a "pilot and rollout" approach led by a director to relaunch its CRM initiative. To improve the usability and accessibility of the data-mining application, the remediation team pulled all of the relevant data from the original CRM system and created a simple Excel-based application, designed with input from salespeople, that organizes and ranks sales leads, matches products to customers, and shows revenue growth from targeted ones by tracking newly generated sales. The remediation team also created an on-line CRM hub where senior sales managers demonstrate the usefulness of the system to the unit's regional sales managers and staff and where selected regional sales managers promote its use in their territories. As positive results began to appear--and as the sales reps' incentive pay rose with them--support for the program spread throughout the firm. The turnaround team expects to keep this effective but bare-bones program in place for two years a nd then to add new features as demands on the system increase.
IT problems too may need a radical fix. A large insurance company, for instance, wanted to use CRM to cultivate high-value customers and either to phase out customers with a lower lifetime value or to persuade them to spend more. After a two-and-a-half-year implementation effort, the application still didn't work. Although the objectives were clear, the IT department got lost in an attempt to integrate more than 20 legacy systems, some more than two decades old, and compounded the mistake by trying to do everything at once. Millions were spent in the attempt to build a single data warehouse that would link all these systems and reconcile differences in each data source.
The answer was to abandon the data warehouse and to link the systems by creating a layer of middleware, a software fix that provided most of the required features and cost less than a total redesign. The company implemented the project in manageable stages and, instead of trying to do everything, provided only the data and analytics with the greatest business impact. This approach--retiring unnecessary or overbuilt pieces of a system and working around obstacles that prove too costly to dismantle and reconstruct--is typical of many architecture fixes.
Once the underlying problems have been identified and the remedies implemented, the original turnaround team must monitor progress against its reaffirmed or revised goals, such as higher revenue from customer retention campaigns. The team should also set internal benchmarks, such as increased use of the system or compliance with project deadlines and budgets, and remain alert to any recurrence of the original signs of trouble, such as complaints from users. A well-executed turnaround ought to yield early wins and measurable progress on most goals within 6 to 12 months.
We believe that many companies with CRM systems--even those that have yet to show results--are well positioned for future success. The key is to step back and do what best practitioners do long before installing CRM: articulate goals, narrow the focus to meet them, and provide the necessary organizational and technological support. Only then can a company expect to see a return on its CRM investment.
EXHIBIT 1 The harder they fall IT project size, function points (FP) (1) Probable outcome, percent On time Delayed Stopped or early 100 81 12 7 1,000 62 18 20 10,000 28 24 48 100,000 14 21 65 IT project size, function Average duration of IT points (FP) (1) project, (2) months Expected Deviation from duration expected duration at completion 100 9 1,000 22 8 10,000 36 14 100,000 48 26 Examples Enhancement ot existing application (for example, major feature in word-processing application) = 100 FP Corporate word-processing application = 5,000-6,000 FP Insurance-administration system = 15,000 FP 'Big-bang' customer-relationship-management systes = 10,000-100,000 FP (1)Function points measure size and complexity of software by counting and weighing functions visible to user (such as inputs or outputs); they are useful in determining relative productivity of software. (2)Excludes stopped projects. Source: capers Jones, Patterns of Software System Failure and Success, London: International Thomson Computer Press, 1996; McKinsey analysis Note: Table made from bar graph EXHIBIT 3 Identify the initiatives that matter Importance of IT to Requirements of customer- Initiatives to improve initiative relationship-management system Targeting of profitable Very high Introduce propensity modelling--for customers example, rule-based systems to improve capture of most important customer variables Ability to deal with High Create campaign library with follow-up contacts history of offers made to targeted customers; make available on-line to call-center agents Convenience of response High Enable response through convenient channels such as e-mail or, for mobile-phone users, Short Message Service (SMS) Delivery of offer High Automate delivery of successive messages to customers who have not yet responded Presentation of offer Low Introduce sales scripts for (sales pitch, material) call-center agents; employ support systems that help customize scripts by customer segment Attractiveness of offer None N/A
(1.) AMR Research, 2001.
(2.) Giga Information Group, Market Overview: E-Business/Enterprise Software Applications in 2001 to 2005: Giga Planning Assumption, March 5,2002.
(3.) International Data Corporation, Demand-Side Survey: A Reality Check on CRM Software, a 2001 study of 300 companies.
(4.) Retail banking was one of eight sectors studied in the McKinsey Global Institute's report US Productivity Growth 1995-2000, 2001.
(5.) According to the 1999 IDC/Cap Gemini publication Customer Relationship Management: The Changing Economics of Customer Relationship, a survey of 300 large US and European companies, 68 percent of all managers surveyed expected at least a 10 percent increase in revenue.
Manuel Ebner is a principal in McKinsey's Zurich office; Art Hu and Daniel Levitt are consultants in the Silicon Valley office, where Jim McCrory is an associate principal. Copyright [C] 2002 McKinsey & Company. All rights reserved.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||customer relationship management software|
|Author:||Ebner, Manuel; Hu, Arthur; Levitt, Daniel; McCrory, Jim|
|Publication:||The McKinsey Quarterly|
|Date:||Dec 22, 2002|
|Previous Article:||Who's accountable for IT? Business leaders--that's who.|
|Next Article:||Intellectual property: partnering for profit; companies could earn up to 10 percent of their operating income from the sale of patents and...|