top of page

Creating accurate, data-driven customer segments

Why customer segmentation?

We’ve all been there: Trying to separate our customers into groups that will help us balance our teams’ time and resources, align the moving parts between Sales, CS and Account Management books of business, and ensure appropriate benchmarking. But how we do it is maturing with new knowledge, data, and technology.

Segmentation methods

Traditional customer segments

Traditionally, organizations have segmented customers based on some form of demographic data like revenue size (ACV), company size (number of employees or number of customers), or industry.

It’s a reasonable method! By considering size, spend or industry we are acknowledging that a small start up is unlikely to interact with products and services in the same way as a large conglomerate and that resource needs and behaviors that lead to success are likely different.

But what if these perfectly reasonable assumptions are wrong? If we assign the small start up to a low touch model because of their small size or minimal spend, we may miss complexities in their business and lose their growth and upsell potential by not investing and supporting early on. Alternatively, the large corporation we’ve assigned to a high-touch segment may not need as much support due to their own well-established internal processes and devoted resources, leading us to waste our team’s time. Assumptions can lead us astray when it comes to segmentation.

The future of customer segmentation

What if there was a better method of segmenting customers than arbitrary demographics? What if we could base segments on multivariate correlations that took each of the above pieces of information into account, along with other metrics, like product use, historical sentiment, location data, maturity, and more and then determined which customers are the truest apples-to-apples comparison? What if that could be done regularly and automatically, without a data science team running constant analyses, so that you could rebalance resources as needed?

How to achieve the data-driven future of customer segmentation

The only way to know, rather than guess at, customer segments, is to bring together a holistic and historical view of your customers’ data – as much data as possible – and then to run statistical clustering analyses to see which customers most closely align on the most variables. These analyses will help you see which customers have historically been most closely aligned on the most variables, allowing you to accurately benchmark behaviors and health going forward and to better assign CSMs, Account Managers, and Account Executives. If you run this analysis regularly, you may find that customers shift in their clusters, and rebalancing is needed to keep your organization sharp and well-allocated.’s Health Score Ingredients is a recently released feature that shows our customers exactly how their own customers are segmented, as well as how each segments’ behaviors are benchmarked. We handle the data unification and analysis, taking the lift out of accurate, predictive customer segmentation.



bottom of page