News Stay informed about the latest enterprise technology news and product updates.

Do you dare show a customer the door?

Weeding out low-profit, high-maintenance customers is all the rage in CRM these days. But columnist Peter A. Buxbaum argues that such a strategy is insane.

Open for

IN an effort to rescue CRM from its doldrums, vendors are emphasizing capabilities that allow users to analyze customer behavior and rank them based on their profitability. These features tempt CRM users to segment their customers based on their profitability, to treat them differently based upon that criterion, perhaps even to "fire" unprofitable customers.

Is that such a good idea? All companies, of course, seek to maximize profits. The question is whether differentiating customers based on profitability is a reasonable means to that end. The answer is not as obvious as it appears.

CRM has been under attack since its beginnings for its reported high failure rate. There are several reasons for this lack of success, but one overarching cause looms largest: the failure of companies to coordinate customer strategies to CRM deployments.

The best customer strategies focus company efforts on acquiring and maintaining customers whose needs are most tightly aligned with the company's products and services. Customer management makes little sense if a company doesn't know which customers it's going after and which it should fight hardest to retain.

Ranking and segmenting customers by profitability is now masquerading as such a customer strategy but its chances for success are dubious. For one thing, individual customer behavior is particularly difficult to predict. Historical data may be interesting, but it is limited as a guide to the future. It is next to impossible to analyze a customer base at the most granular level because much of individual customer behavior is random. This is especially true when sporadic behavior --such buying a car or applying for a bank loan -- is involved.

But what about the really nightmare customer, whose activities have not yielded a profit over time, who constantly makes impossible demands, and who has not responded to some coaxing to behave differently? Might not such a customer be a candidate for the boot?

Absolutely impossible customers sometimes need to be dealt with individually. But consider this: you don't know all of the reasons why a given customer is a low-profit pain in the neck. Maybe he's frustrated because his business is doing poorly. And maybe his business turns around, and he has plenty of money to spend, after you've already shown him the door.

The actuarial and credit card industries, among others, have decent track records analyzing aggregated customer data. Analyzing group data gives CRM users a better shot at predicting customer needs among a given cohort. Segmenting customers by profitability, say it proponents, would allow companies to give VIP treatment to its most profitable segment, provide ordinary service to marginally profitable customers, and, perhaps, urge unprofitable customers to take their business elsewhere.

While this seems reasonable on a superficial level, such a strategy is not likely to yield optimal profitability. Studies have shown that most customers don't maintain peak profitability in the long run, but rather trend toward the middle over time. This suggests that a profitability snapshot is not likely to yield a customer strategy that optimizes profitability.

CRM analytics, as they stand now, capture nothing more than a moment frozen in time. They're not going to work well analyzing evolving behavior, or in calculating the total lifetime value of a customer. Customer strategies and customer relationships need to be based on more than such a static snapshot.

Most companies are understandably reluctant to show customers the door. That intuition is confirmed by statistics that show that the cost of customer acquisition is much higher than the costs of customer retention.

And what about the costs of regaining a customer that has already been banished? What are the chances of that really happening -- at any price?

Peter A. Buxbaum has been writing about business and technology for more than 10 years. In addition to his regular contributions to, his articles have appeared in Forbes, Fortune, Chief Executive, InformationWeek, Line56, Computerworld and more than a dozen other publications. He has also developed and taught seminars on international business at Penn State University.

Dig Deeper on Enterprise business applications

Start the conversation

Send me notifications when other members comment.

Please create a username to comment.