Case Study: Account-Based Segmentation


To better optimize our client's marketing and sales activity, we developed an account-level segmentation using results from a quantitative survey.


Our client has a variety of customers, across a plethora of product lines. It asked KJT Group to help develop a consistent way to classify its customers, above brand, at the account level (e.g., private physician office rather than physicians).

The client’s specific research objectives were to:

  • Develop an account-level segmentation based on a variety hypothesized drivers.
  • Link segmentation results to its database using a typing tool to determine segment size.

KJT Group conducted a quantitative survey with a variety of respondent groups who were potential customers for our client. We recruited more than 700 respondents. These included:

  • Physicians from independent offices
  • Physicians from independent practice associations (IPAs)
  • Administrators/C-Suite from corporate systems
  • Physicians from offices who are a part of corporate systems

We recruited using postal mail, fax, telephone and email. Administrators/C-Suite were given the option to complete the survey over the telephone or online, while physicians were recruited to complete the survey online. We recruited only one respondent per account to ensure our final sample had unique accounts to analyze.

Administrators/C-Suite were also given the option to receive a high-level overview of study results to compensate them for their time for participating in the study.

We developed a 30-minute survey based on several hypothesized domains (e.g., capabilities, leadership strength and structure). The resulting analysis would divide customers into homogeneous groups based on differences in these domains.


The purpose of the study was to identify homogeneous groups that are maximally similar within, and maximally different from each other. We used latent class techniques to identify segments sharing similar behaviors, psychographic characteristics, and perceptions.

We evaluated various solutions, using different sets of questions and a different number of segments. The final solution (7 segments) showed strong statistical validity.


To elucidate the story of segmentation, we highlighted how segments performed within each domain. Based on this analysis, it became clear there were three segments that were market leaders, two segments that were average, and two that were followers.

We also provided detailed segment profiles, highlighting core attitudes and demographics for each segment that can be targeted/activated. These profiles included secondary brand data provided by our client, which brought out important differences for our client's marketing teams that would have otherwise been missing from the above-brand study.

When developing the segmentation, we made sure to focus on variables that could be connected to our client's secondary data. We provided a typing tool that allowed them to classify their customers into each of the seven segments.


Interested in learning more?