Segments of One: Measuring the Loyalty of Mega Customers in B2B Markets


Share on LinkedIn

Every company in the B2B space, regardless of industry, has its 800-pound gorilla customers. Whether they are called Managed or Key Accounts, Enterprise Clients, Top Tier — or simply referred to as “our biggest customers” — these are the folks who are disproportionately large relative to the overall size of the business and the average customer. Depending on the size of the company, there might be 10 or 110 (or more) such clients.

These are the prized gem clients that expect and receive special treatment in some manner because of their value and importance. Companies often have dedicated staff managing and servicing these accounts – frequently onsite at customer locations. These clients typically demand customization in everything from product and service standards to contract terms and support. They receive extra attention from senior management.

In effect, these customers are “Segments of One”: individually managed outside of the standards and processes governing the company’s treatment of the other 95% to 99% of its clients.

The Measurement Challenge

Of course, the loyalty of Segments of One customers is far more important than that of the typical individual client. These mega customers, however, present unique challenges when companies try to include them in their loyalty or experience measurement programs.

* First, since these customers receive such highly customized treatment, the standard questionnaires used with other customers may not apply or may need substantial modification.

* More importantly, there almost never are sufficient numbers of mega customers to support traditional statistical analysis. Qualitative input can be instructive, but provides no substitute for quantitatively derived key drivers. Traditional analyses for identifying and quantifying the impact of key drivers typically require 300 to 400 responses.

* The 800-pound gorillas can be rolled in with other customers for analysis, but this leads to a potentially intractable problem regarding weighting. If the mega customers aren’t weighted to reflect their importance (typically based on sales value), their unique importance and impact is lost in a sea of far smaller customers. Conversely, if these customers are weighted to capture their economic value, they receive astronomical weights that can be distorting, and they drown out the voices of mid- and small-size customers. Damned if you do (weight, that is), damned if you don’t.

Recognizing these issues, what if you wanted to build a company-specific model of the drivers of loyalty or satisfaction? There might be one or a few dozen key decision makers and influencers. Or, what if you wanted to model the drivers for a handful or few dozen accounts? But there traditionally has been no reliable quantitative approach for modeling the key drivers for a single decision maker or a small group of decision makers or companies.

Key Customer Relationship Management (KCRM)

Recognizing the challenges of managing and measuring customers in this Segments of One category, we developed an approach explicitly designed to provide rigorous statistical analysis to identify and quantify key drivers for small samples. Rather than the 200 to 300 observations required for traditional modeling, KCRM supports modeling for much smaller groups, down to a sample of one. The result is the ability to derive key drivers for individual respondents and companies or small groups of individuals or firms.

KCRM accomplishes this by employing Analytic Hierarchy Process (AHP), a sophisticated technique for reducing complex decisions to a series of paired comparisons between alternatives. By systematically rotating through the trade-offs between pairs, AHP provides a vehicle for identifying and quantifying the relative importance of different criteria in driving an outcome, such as a measure of loyalty or another KPI.

The AHP approach has been validated as statistically superior to simple correlations and regression analysis when dealing with small samples, including samples of one to manage Segments of One. More importantly, it has been validated as a useful tool for managing Key Accounts and predictive of financial outcomes.

Paired comparisons are relatively real-estate intensive and require significant survey space. Attributes can be grouped to permit the inclusion of more performance ratings, but the number of attributes that can be included still is fewer than that which can be accommodated by traditional rating surveys. The scale is different than that used in most loyalty and “sat” surveys, so it may not be easy to display the results side-by-side with traditional driver models. The same KPI or outcome variable, however, can be used in KCRM and other customer loyalty and experience surveys.

Managing Segments of One

KCRM provides a company with an account-level or even individual contact-level road map of the drivers behind the relationship or some type of interaction or experience, and a measure of the company’s performance against those drivers. This also can include measures of the performance of competitors. This provides a company or an account manager and team with a meaningful way to set priorities and allocate resources to shore-up any weaknesses, capitalize on strengths, de-emphasize less important issues, and counter any advantages that might be held by competitors.

Quantitative measures should not by any means totally displace direct qualitative feedback from Key Accounts; but quant measures provide a critical additional level of insight. Decision makers and influencers are people. While their corporate hats may be draped with rational models of business decision making, they have all of the limitations and foibles of people making personal decisions, albeit in a more highly structured environment with other influences.

We know from countless studies of individuals at all levels of education and accomplishment that people are not in touch with how they make decisions, are lousy at understanding trade-offs, are poor at processing information and, most importantly, can’t really explain why they do what they do – although we’re great at rationalizing decisions after the fact. For these and other reasons, quantitative techniques are a critical complement to the context that can be obtained from qualitative approaches.

With KCRM, marketers and researchers have, for the first time, a means for identifying statistically derived key drivers on an account-by-account basis, for individual decision makers, or small groups of companies and individuals. As a result, the Segment of One customers now can be better understood and managed in a manner compatible with their distinct importance to a company.


Please enter your comment!
Please enter your name here