Best Practices

How to Build a Customer Health Score

Which of your customers are unhealthy? Which ones are thriving?

If your company is small or just starting out (or both), you may be able to answer these questions off the top of your head. But what happens as your organization scales? If you have hundreds, even thousands, of customers, how will you identify the unhealthy ones?

At this year’s Pendomonium conference, the team from cybersecurity firm Rapid7 told attendees how they faced this exact challenge. The more customers they onboarded, the larger their blind spot regarding customer health grew. To address this problem head-on, they made the decision to create a customer health score against which all customers could be quickly and easily measured. 

The Rapid7 team made this decision back in 2017. Now they consider their customer health score to be their most important customer-focused metric. During their Pendomonium presentation, three members of the Rapid7 team explained how they got started building this metric, what went into theirs, and how other companies could create a CHS of their own. 

Getting Started: Health Score Data Bingo

When the Rapid7 team took on the customer health score project back in 2017, they quickly discovered that building this metric would require a multi-departmental effort. As presenter Kalley Collete put it, it took a village. The first step was to establish a steering committee for the initiative, which included team members from multiple departments, including product, marketing, and customer success

Next, the team needed to select the data points that should be included in the CHS metric. The first exercise the team tried was “Health Score Data Bingo,” which was basically a sticky note session where they put every metric they could think of on a whiteboard. Then they winnowed down and prioritized the metrics. They ended up breaking their list of numbers down into “initial,” “medium-term,” and “long-term” categories, based on how quickly they felt they could incorporate them into their overall customer health score model.


  • Product adoption
  • Support experience
  • Purchasing behavior


  • CSM input
  • CSM interactions
  • NPS


  • IT ecosystem
  • Security maturity
  • Product certifications

A Customer Health Score Model

For their initial pass at a customer health score, the steering committee decided to start simple and include only their “initial” metrics in the model. Over time, they planned to incorporate additional data points in the overall score. 

Rapid7’s first customer health score was an amalgam of the following: product adoption (40%), support experience (35%), and purchasing behavior (25%). If data on customer success interactions was available, that accounted for 35% of the score in place of support experience.

Each of these three components was itself a mashup of multiple data points. For example, the support experience piece of Rapid7’s CHS score included stats like NPS, time to close, and number of escalations. And the purchasing behavior component included the number of products, number of renewals, and lost vs. open opportunities.

Based on their score, a customer would be classified as “very unhealthy,” “unhealthy,” “neutral,” “healthy,” or “very healthy.” Plans of action would then be put into place according to that score, with unhealthy customers receiving more attention than healthy ones. 

Putting CHS Into Action

So far, Rapid7 has experienced pretty high levels of success with this model and scoring framework. Recently, they celebrated the milestone of having their customer health score available directly in Salesforce.

In addition, they’ve begun seeing real benefits to having a single metric against which every customer can be assessed. The CHS framework allows Rapid7 to easily identify unhealthy customers and “treat” them accordingly. It also makes prioritization much easier, with the less healthy customers moving to the top of the pile. Finally, the team can see how a customer’s score is trending over time. Hopefully, unhealthy customers move toward health over time, and healthy customers stay that way. If not, the team can quickly address it. 

Some Lessons Learned

Since Rapid7’s initial adoption of a customer health score, they’ve come away with some key insights. For one, they’ve discovered that some unhealthy customers may already be lost. At a certain point, no amount of additional support will bring them back to health. For another, healthy customers are the best cross-sell candidates. They’re the most likely to buy your product and more of it!

Neutral customers, however, are where you should spend the greatest amount of time. These customers are satisfied but “untapped.” With a little extra attention, they can become your company’s biggest advocates.

Finally, creating a customer health score model requires a lot of trial and error. If your team wants to build one, just start doing it and iterate as you go!