The fanfare and promise that boosted CRM’s introduction – and raised managers’ expectations – have turned to disappointment and doubt. There is, as these authors point out, still hope for reviving and realizing hopes for high returns from CRM. To help them, managers would do well to apply the lessons that these authors write about.

Few business tools have been so over-promoted and have so underperformed as Customer Relationship Management (CRM). A recent Bain & Company brief summed up the situation: “In 2000, there were more than 14,000 media mentions on the topic, up from just one in 1989…451 senior executives placed CRM fourth from the bottom of 25 tools in Bain survey’s satisfaction ratings. In fact, a fifth of CRM users had abandoned the tool altogether.” (Darrell Rigby, Fred Reichheld, and Chris Dawson, Making CRM Work, Bain & Company Results Brief, February 2002.)

While estimates of CRM’s failure rate vary widely (some are as high as 90 per cent), one thing is certain: Something has gone terribly wrong with an idea that seems so sensible. While there are different opinions about what has gone wrong, we believe that there are five major reasons why CRM initiatives have been disappointing. We describe those five reasons and what we have learned from them below.

1. CRM is not properly understood, leading to unrealistic and inappropriate expectations.

Ask people to define CRM and one will typically get different answers, even in the same company. For some it is database marketing. For others it is sales contact automation or even self-service for customers through telecommunications technology. Yet, there is nothing new about the basic idea behind CRM. In fact, until recently we called it by another name: the marketing concept. Marketers have long urged their colleagues to “get close to the customer” and to then tailor the firm’s offerings (products, services, encounters) as closely as possible to their customers’ individual tastes and behaviours. And for exactly 100 years, we have been teaching this customer-centric notion in our business schools and professional meetings.

To be fair, the problem has always been how to implement the marketing concept economically for more than just a few large, important customers. This has meant that “relationship marketing” has largely been practised only in key account situations or by small businesses that have few customers and could therefore have individualized relationships. The rest of the time, we have had “mass marketing.” Thus, the basic strategic idea of CRM is to use information technology to create a highly personalized, learning relationship by moving ownership of the customer up to the firm level, from the department, channel or salesperson. If it is not economical to do this for all customers, then managers have been advised to give priority treatment to their most valuable customers. What really is new is the wide availability of information technology to identify, track, and interact economically and effectively with innumerable customers on an individualized basis. Unfortunately, many think that CRM is actually the information technology, rather than the reason the software was created in the first place.

But CRM is an attitude, not a technique or software. When CRM is misunderstood, the chances are very high that it won’t help at all.

Before any changes are made in the information system, a management team must become very clear about what CRM means to it and why it wants to proceed. For example, management at Kraft Canada has a clear vision of what it is doing with CRM-customizing relationships with customers rather than customizing products as an avenue to competitive advantage. This vision has made Kraft a leader in CRM implementation in Canada.

Any understanding of what CRM is must be accompanied by an understanding of what it is not. CRM does not replace, but rather complements, other marketing and customer service initiatives. For example, CRM does not supplant marketing research (MR). A good CRM system will provide detailed profiles on individual customers and longitudinal behavioural data, and enable an analyst to know what particular customers are doing. On the other hand, MR remains necessary to understand non-customers, customer behaviour by segment, and to predict departures from historical behavioural patterns. In other words, CRM and MR should be integrated for a more complete view of customers.

2. Too many CRM initiatives are designed to help the company, not the advantage.  This vision has made Kraft a leader in customer.

There are usually two implicit assumptions in the most enthusiastic proposals for embracing CRM: that the customer wants a relationship with the company, and that this system will make the customer better off. Neither assumption is accurate. How many relationships do most customers really want in their lives? Their behaviour indicates they do not want an intimate relationship with every company they buy from, nor every time they buy. For example, grocery stores have learned that good service at the checkout means being courteous, professional and speedy-not personal, conversational and dawdling. Customers want respect from sellers– respect for their time, individual worth and patronage. Customers want responsiveness-to their questions, concerns and problems. And customers want marketers to live up to their promises, to do what they said they’d do. Beyond these dimensions of a relationship, why would most customers want “intimacy” along the lines voiced by CRM advocates?

If a CRM initiative is really just an excuse to get information to promote cross-selling, if it is really just another trigger for a mass, direct mail exercise, if it keeps customers away from  contact with real people at the company, and so on, then customers soon learn there is no reason for them to share information with the marketer. If, on the other hand, the CRM system provides something extra for the customer (saves time or effort, recognizes past patronage, doesn’t ask for the same information twice, speeds up problem resolution, is matched to providing better information about the marketer’s products and their availability), then it is far more likely to elicit co-operation. In short, does it add value to the relationship for the customer, not just the company?

Consultants Don Peppers and Martha Rogers expressed the essence of a good CRM approach in their seminal work on one-to-one marketing: “One-to-one marketing occurs when you and your customer interact directly, the customer tells you something about how he or she wants to be served, and then you change your behaviour with respect to this individual customer based on that interaction” (, Dec. 4, 2001).

Too often, advocates of CRM also seem to forget that crossing the privacy line to fill their database is not just intrusive; it is also likely to result in customer backlash. Critical to any relationship is trust. Branding is all about creating trust in a customer relationship. So is the use of customer information. Permission is critical to CRM, and anyone intending to implement a CRM system should read Seth Godin’s important book, Permission Marketing (Simon and Schuster, 1999).

3. It is difficult to choose the appropriate CRM solution.

While it is possible to have a client system that is paper-based and can be accessed by everyone in the organization, companies are increasingly using linked databases to allow information to flow from one department to another. In this way, the database of customer relationships becomes the “corporate memory.” Software is necessary to link together a company’s various databases through a central database server. As of today, there certainly is no shortage of CRM system service providers. A recent Gartner Group study found that while there are 500 or more companies claiming to make CRM software, only about 200 of them could legitimately make that claim. This means that it is difficult to know which vendor is right when shopping for CRM software. The wrong technology may not only be very costly, but it could well cripple an organization. There are many examples of companies who chose the wrong application. In 1998, rolled out a high-end software package to provide its telephone sales force with instant information on prospective customers, with the hopes of expanding the company around the world. However, the new system was so slow that telephone reps were unable to help customers. Salespeople in the field were locked out of the company’s customer database for a full year. This cost the company millions of dollars.

Many CRM vendors are launching industry-specific suites (referred to as vertical tools) that enable companies to deploy the software faster than generic packages, and without the costly customization. Bo Manning, CEO of Pivotal Corporation, has stated that, “Customers of generic CRM suites typically spend at least four months customizing 25 to 30 per cent of the functionality; those who buy vertical suites spend about two months customizing 10 to 15 per cent of the applications.” (Jennifer Maselli, “Industry-tuned CRM Suites Save Time,” Informationweek, Nov. 5, 2001.)

The technology for CRM is constantly evolving. In the fall of 2001, CRM software vendor Aspect Communications Corp. added voice-recognition technology to a self-service application that allowed customers to process transactions by telephone. The software identified customers by their “voiceprints,” eliminating the need for them to enter personal identification numbers or other passwords to access account information or pay bills. Banks and other security-conscious companies apparently are particularly interested in this.

4. Implementing CRM is complex and difficult

Implementing a CRM system is a major challenge in most organizations because there are so many ways in which the initiative can go wrong. (For an excellent review, see John Calhoun, “Driving Loyalty by Managing the Total Customer Experience,” Ivey Business Journal, July/August 2001, pp. 69-73.) While resistance may come from salespeople who do not wish to give up “ownership” of customer information or from IT people who do not wish to change legacy systems, experience has shown that for a CRM program to be successful, above all else it must have the full support and ongoing commitment of top management. They must view CRM as a strategic initiative, not as a technological tool or tactical program.

Top-management commitment can be secured by demonstrating that automation supports the business strategy (i.e., automation delivers the information required to make the key decisions which enable business strategy to be realized); that automation measurably impacts and improves results (e.g., improved win rates, improved margins, higher sales revenues, and higher customer satisfaction ratings); and that it significantly reduces costs (e.g., lower general sales costs) and thereby pays for itself over a specified time period. (Barton Goldenberg, “Customer Relationship Management: What Is It All About?”:, Dec. 4. 2001.)

For companies focused on products or services, this means that they must align themselves with the customers’ needs, something that might itself mean a radical change in a company’s culture and systems, a kind of corporate rewiring. All employees-especially those in marketing, sales, service and any other customer-contact functions-need to have knowledge of the customer’s total experience with the firm. An early step is to consider how to identify and track individual customers, not transactions or accounts. This means aggregating customer information across all points of customer contact, including multiple channels, if they exist. The success of any CRM system depends upon collecting complete, clean, consistent, current and accurate data from all customer touch-points across the company. The company must apply strict rules to incoming customer information. Data are collected from numerous sources, e.g., point-of-sale transactions, web-site usage and third-party databases. It can be transaction- based data, customer-profile data or behavioural data.

Since most companies cannot afford to automate or integrate all CRM functions at once, they should begin by examining all their business processes and determine which ones need to be automated first and what technical features are required. This is particularly true when many legacy IT systems are involved. The best way to decide on priorities is to ask the people who will actually be using the technology (e.g., sales representatives, customer support staff) what they need to do their job better. While an IT group or outside service provider can offer technical information such as which IT infrastructure may be best, these people should not lead a CRM implementation. Those who will be using the CRM system must be involved. Further, successful CRM depends on coordinated action by all departments within a company rather than action that is driven by one department. Many firms have struggled with the challenges of integrating data across organizational silos and of data ownership.

A CRM system is not independent of the people who operate it. They must be trained properly. This includes demonstrating how to gather information correctly (with minimum intrusion on the customer) and how to access and use CRM information. It has been estimated that over the life of a sales-and-marketing automation system, training ends up costing an average of 1.5 times the cost of the sales and marketing automation system hardware/software. As part of their training, CRM users need to be shown the importance of CRM to their specific job and to the company, especially since CRM crosses organizational departmental lines. Many CRM implementations have failed because some people weren’t sold on the system. For example, a salesman with Crane Engineering, a Wisconsin-based industrial equipment distributor, caused his company’s server to crash when he fell 190,000 transactions behind when using his company’s CRM software. The salesman neglected to update address changes, read messages and input customer appointments for several days. (Sharon McDonnell, “Putting CRM to Work,” ComputerWorld, March 12, 2001.)

5. It is easier to demonstrate the costs of CRM than the benefits

Measuring some of the costs of implementing CRM can be straightforward. According to SAS Institute Inc., an application service provider, the average cost for a full analytical CRM package is $500,000 (U.S.). Wendy Close, a research director with The Gartner Group, has said that in a typical CRM implementation, 28 per cent of the total cost is for buying software, while 38 per cent is for services such as software customizations, application integration and training. Twenty-three per cent of the cost goes to hardware and 11 per cent goes to telecommunications.

However, there are costs that are not so straightforward to estimate. A company can build its own CRM strategy or customize an off-the-shelf solution. A recent AMR report shows that for every dollar a company spends on CRM software, it must spend $2.50 to $5 on integration and implementation. (Rene White, “Integrating e-CRM,” ebiz, website:, Dec. 3, 2001.)

Large companies in particular have difficulty in finding one software package to serve all their needs. For example, Bank of America is using CRM solutions from Synchrony Communications Inc., Siebel Systems and Youcentric Inc. Synchrony will integrate the other applications. A March 2001 Gartner Group research note entitled “The Emergence of CRM Infrastructures,” indicated that the average Fortune 1000 company stores customer data on a minimum of 10 different systems. To improve customer relationships, these systems have to be connected.

Measuring the payoff of a CRM system, especially in its early stages, can be difficult and contentious, which can lead to poor judgments by management. Of particular concern is measuring the impact of, “What would have happened if we hadn’t adopted a CRM system?” Consulting companies caution that up to 70 per cent of CRM projects don’t produce measurable business benefits. However, it is not clear if this is because CRM approaches were inappropriate, badly implemented, or badly measured.

Companies can expect returns on investment in CRM in two major areas: reductions in operating costs/increased efficiency (e.g., reduced call-centre talk time), and in the increased spending/visit activity of the customer base (and increased loyalty). Measurement criteria should be established based on the objectives of the CRM program (such as: maximize customer revenue, decrease cost-to-serve, maximize penetration, maximize loyalty, maximize sales, maximize marketing effectiveness, increase employee satisfaction, etc.). These measurements can be used to manage and continuously improve strategic CRM performance.

The best candidates for successful CRM programs are those companies that have differentiated customers, that offer differentiated value propositions to those customers, that have customers with potentially high lifetime value, that collect data on individual customers and transactions, and that are in direct contact with customers. Applying the lessons we have learned from CRM initiatives over the past few years should enable business leaders to realize greater returns from CRM in the future.

About the Author

Michael R. Pearce is Professor of Marketing at the Richard Ivey School of Business.

About the Author

Yvette Mahieu was Ivey Research Assistant to Professor Pearce during his tenure as Eaton/NSERC/SSHRC Chair in Retailing.

About the Author

Yvette Mahieu was Ivey Research Assistant to Professor Pearce during his tenure as Eaton/NSERC/SSHRC Chair in Retailing.