The Value of Customer Segmentation in B2B Sales and Marketing

Lisa Fiondella

September 12, 2014

The following scenario is probably relatable for those of us in B2B markets that have struggled to gain greater perspective to clients and prospects but just can’t put our finger on the real, tangible behavioral characteristics that drive performance – good or bad. If this sounds familiar…read on.

Maximum Medical sells its products to hospitals, laboratories and physician groups in North America. Since the company launch eight years ago, it has created a solid revenue stream and customer base. In the last year company growth has slowed as new entrants have created greater competition. Company leaders feel its time to refocus sales and marketing efforts to improve business performance.

There has been a great deal of internal debate about the best approach but the presence of some unanswered questions have brought discussions to a standstill. These questions include:

Within our market, what are our best customer segments and why?

What are the profiles or characteristics that describe our best customer segments?

Company leaders believe if they can obtain answer to these questions, they can create the best marketing plan and determine an appropriate sales structure (including sales model, organization and process) for the business. They believe these insights will enable the company to better service its current customers and help identify and prioritize those that represent the best future ones.

The marketing department is asked to conduct a customer segmentation analysis and the project is assigned to a marketing manager who goes about the initiative in the following manner.

  • 1. Interviews are conducted with sales managers and representatives to gain their perspective based on past sales experience.

2. A customer survey is conducted that yields a .4% response rate. Sales representatives are asked to call customer and encourage participation by offering a restaurant gift card.

3. An analysis of customer care cases from the last 90 days is performed to look at recent case reasons.

4. SEO and other lead generation sales results are reviewed to identify the lead types with the highest close rates.

5. A report is generated from the financial system to rank customer types by sales dollars in the most recent six-month period.

6. The marketing manager evaluates the results and assumptions are derived based on the findings. A summary report is then circulated to company leaders.

The problem with this approach may be apparent to some but let's not assume anything. In reality, there are at least two critical flaws in the above method.

First, this is a subjective analysis. It relies on opinions and experiences from a handful of people (internal and external) to develop its findings. While a subjective analysis is certainly better than none at all, it introduces a level of bias that may in fact mask the real insights the business seeks to uncover.

Second, the analysis relies on small and/or point in time slices of data. Sampling can be an effective data technique but only if it captures the right data and is done in such a manner as to remove most forms of bias. For example, the customer survey is too small and is likely biased based on company/client relationships. The finance report looks only at sales in the last six months and doesn't take into consideration customer profitability. And just because a customer is captured in the top buyers doesn't mean they represent the best types of future prospects.

In my experience, the best approach to customer segmentation (B2B or B2C) is through predictive analytics. This approach removes the bias of a subjective analysis and allows the business to leverage the maximum data to “tell the customer story.” A predictive analytics approach will leverage:
1. A design process that clearly articulates the desired outcome. This is where business leaders agree upon the specific objective of the analysis.

2. The maximum data available to drive model performance including internal business data (think sales, product, marketing and financial) and external third-party data as applicable.

3. Proven predictive modeling techniques that cluster and rank order customers and prospects into meaningful groups.

4. A simple, easy to understand and usable deliverable.

The result is a statistically sound segmentation and customer profiling system that can be used for process and organizational planning, lead acquisition and within existing business applications (such as CRM and marketing automation systems) to guide and prioritize sales and marketing activity.

High value customer segmentation is not a pipe dream for B2B companies. The tools, data and skills exist and if applied correctly, the resulting insight will improve sales and marketing programs and most importantly, company revenue.