Consumer marketers have long known the benefits of database segmentation. Armed with extensive customer data, many have perfected the science of identifying the prospects that represent the likeliest buyers or most attractive future customers. Yet b-to-b marketing departmentsâoften judged on their ability to generate qualified sales leadsâare overlooking the value of such analysis, according to industry experts.
Broadly speaking, database segmentation is the analyzing and parsing of a customer list in order to find the most profitable customers. This could mean that one member of the marketing team tracks customersâ activity to identify the most and least valuable customers. Or, in more complex environments, it might involve an in-house analytics team that handles sophisticated data mining and modeling to predict customer behavior.
The benefits can be significant. A marketer might determine, for instance, that certain customers purchase its products four times a year and are highly responsive to promotional offers sent by e-mail. The marketer can use that information to better target the customer and prospects at similar businesses.
But database segmentation also presents challenges for business marketers. One problem, said Eric Schmitt, an analyst at Forrester Research Inc., is the dearth of current, relevant data on business customers. "Thereâs so much more data that exists on the consumer side," he said.
With the exception of big companies like Office Depot, he said, most b-to-b marketers have fewer customers than consumer marketers do. That means the value of analytics and predictive modeling is much lower. "The higher the number of customers, the greater your ability to slice and dice the data and look for meaningful patterns," he said.
Chris Berry, VP-business information at VNUâs Claritas, a business information provider in San Diego, said itâs difficult to obtain and freshen business information. "Even with segmentation
successes, b-to-b marketers are still challenged with ultimately identifying the key decision-maker or buyer," he said. "This is an area the data suppliers are always trying to improve, but [they] struggle with it based on the dynamic aspects of business information."
Claritas has a proprietary segmentation system on the consumer side but not on the b-to-b side, Berry said. The reason, he said, is that thereâs no clear return on investment for Claritas to have one.
Another problem is the organization in many b-to-b companies, which tend to be operations-driven, with a heavy reliance on field salespeople.
"Often the information that goes into the database is sales force-contributed, and itâs tough to get [the sales force] to be compliant and diligent in updating that data," said Gary Laben, president-CEO of KnowledgeBase Marketing, Richardson, Texas, a subsidiary of WPP Group.
Yet b-to-b marketers interested in database segmentation have a growing number of options to help them accomplish a task that historically has taken special expertise and considerable time. Most customer relationship management and campaign management software vendors now provide some level of segmentation capability. These programs have helped to reveal gaps in customer data.
"[Marketers are] finding out they are lacking critical data attributes and lacking a good classification of customers," said Kevin Cavanaugh, VP-technology at CRM software provider Unica Corp., Waltham, Mass.
Data providers such as infoUSA Inc. are also heeding the call for segmentation services. Last year, the company introduced Customer Analyzer and New Prospect Builder to address the needs of small businesses. The Web-based product analyzes a customer database against infoUSAâs database of 14 million businesses in order to profile and segment customers based on SIC codes and number of employees. The service costs $99.
Rakesh Gupta, president of infoUSA.com, said the product was developed for companies that want "a quick and dirty analysis" of the customers in their databases. "It is more affordable for smaller companies to do segmentation now," he said.
B-to-b marketers interested in segmenting their databases can also turn to consultants for help. They can provide basic segmentation tools such as RFM (recency, frequency, monetary) analysis, which is used to quantify a companyâs best customers by how recently they have purchased, how frequently they purchase and how much they buy. Such consultants can also provide more sophisticated tools such as regression analysis. This process takes into account RFM and adds additional information about the customer, such as number of employees and estimated spending in different product categories.
However they accomplish it, marketers that have committed to database segmentation have reported positive results.
Jim Roots, VP-marketing at Lake Forest, Ill.-based W.W. Grainger Inc., said segmentation has helped his company structure its marketing around a massive set of products and customers. "Not all of our 1.5 million customers buy all of our 500,000 products," he said. Grainger manages the segmentation processes in-house through its research and analytics staff of 10 people.
"When you get into b-to-b, large accounts with literally thousands of buyers under one relationship, [RFM] techniques donât work quite as well," Roots said. "You need to get more sophisticated."
The analytics staff also uses regression analysis and occasionally works with outside resources for data collection, survey work or to refine modeling techniques, Roots said.
"Weâve built models to help us refine our targeting, not only by type of companies to focus on but also by what type of individuals within the company to focus on," Roots said.
Currently, Graingerâs analytics group is exploring ways to refine the segmentation further to take into account the types of purchase processes that occur at each of its client companies, he said.
The key to success, of course, is to not just gather and segment data but to use it to better understand customers.
"Segmentation is a tool, and there are a lot of different types of segmentation you can do, depending on what you are trying to accomplish," Roots said. "What data are relevant and what data are not relevant" are the most important questions to answer, he said. "Knowledge out of the data is key."