How To Create A Customer Health Score Formula, Step By Step

Article by

Rochi Zalani

Former CS associate. Writer @



May 3, 2023

What if you had a magic 8-ball that could (almost always) successfully predict whether an existing customer will retain or churn, upsell or downgrade, stay happy, or complain?

Customer health scores help you do just that.

Want to get straight into the nitty-gritty? Feel free to skip down to the step-by-step for creating your customer health score formula.

What is a customer health score?

A customer health score is a number that helps you understand your relationship with your customers. It’s an examination of a customer’s health in relation to your brand — helping you predict (and avoid) customer churns, understand patterns, and upsell at the right time.

For example, let’s say your customer health score warns low product adoption for a customer. Knowing this, you can set up a call to understand how your product can fit better into their day-to-day operations. Maybe you also assess whether they need a product walkthrough for their team to grasp your product's features and its various use cases. Without the customer health score to warn you, this customer might’ve slowly descended to churn.

Customer health scores are simple to understand by definition, but they have many unique layers:

Customer health scores can be calculated and showcased in various ways

Your customer success team can use numbers, scales, grades, or even color coding to display your customer health score.

It depends on what your stakeholders care about, how your CS team is structured, and which customer success platform or other CS tech you use. At ClientSuccess, for instance, you have the option to add qualitative data and let it have an impact on your health score. At Catalyst, you have health-based automation — meaning you can automate specific actions when there’s a change in your health score.

You can have multiple customer health scores

Depending on your company's desired outcomes, you can have different customer health scores to measure different things. For instance, you might have different health scores for individual and enterprise customers because you solve different problems for them (and, by extension, they don’t use your product the same way).

Even if you use the same key metrics to determine both segments’ customer health score, you might put a different weight on each metric — avoiding churn for enterprise customers is more crucial while upselling to individual buyers is more important.

All this to say: customer health scoring isn’t one-size-fits-all. The formula your competitors use might not work for you. It’s best to create a personalized customer health scoring formula for your company, business goals, and CS model.

But is all this effort even worth it?

3 reasons why customer health scoring is crucial

Learning all the layers and complications of customer health scores, you must be thinking: is all this effort even necessary? Why can’t you just solve your customer support tickets, respond to buyers’ emails, do onboarding calls, and call it a day?

Turns out customer health scoring is essential for a successful customer success plan.

1. Customer health scores help you predict churns and upsell opportunities

Customer health scores can have a direct impact on your business growth. If your CS team can identify a potentially churning customer, they can redirect their efforts to retain them and address their pain points.

Similarly, monitoring your health scores can help you upsell to your customers at the right time. For example, if you sell employee management software and a customer keeps updating their account with large files — you know they’re in hiring mode and might need to upgrade their plan and purchase adjacent services you sell. But you have to know when a customer is spending much more time on your product (than usual) to catch this. Thanks to customer health scores, you can view this pattern.

At RFPIO Inc, customer health scores predict churn with a 95% accuracy, quarter over quarter, says Mackenzie Robinson, who works at Customer Success Operations in the company. She adds:

So not only is it [customer health scores] used by individual contributors on a daily basis for daily management and book of business management, risk indicators, etc., but it is also used by our exec team in various reporting to forecast our risk”.

2. Customer health scores enable you to identify your power users

Happy customers are the best marketers. Knowing the customers who swear by your products has many domino benefits: you can send them a gift to make them happier, ensure your CS team doesn’t lose focus on them, or ask them to leave a positive review.

Power users are also more likely to be a part of your customer research, testers of new features, and referral engines. Accurate health scoring can help you determine these healthy customers and make the best of this opportunity.

3. Customer health scoring is essential to providing impeccable customer service

36% of customers are unsatisfied with their customer service experience, and 55% of customer service issues go unresolved, according to the latest Global Consumer Trends report.

Customer health scores make it easy to always have a pulse on all your buyers’ satisfaction (or lack thereof). It enables your customer success team to address customer pain points and understand buyer behavior patterns proactively.

How to make your unique customer health score formula in 5 steps

Here’s a five-step process for creating a customer health score system in your organization:

Step 1: Define the outcome you expect out of calculating customer health scores

What should your customer health score do for your CS team? Does it alert them to potential churns or pinpoint upselling opportunities? Or does it do something else entirely? A health score can do multiple things — especially if you use more than one — but prioritize what results you want.

What you value depends on your unique business circumstances. If your retention rate is solid, upselling might be more of a priority. But if customers sign up and stop the subscription after two months, knowing when a customer’s likely to churn is more important.

Defining the outcome you want from customer health scores also includes solidifying what a “healthy customer” means for your business. For Rachel Jennings, Global Head of Customer Success at Assignar, a healthy customer means a buyer who uses their product five days a week. She similarly defines what an at-risk customer looks like, which is a bit more complicated:

“There’s a lot more variables because not just one thing will say a customer’s at-risk. But typically, it’s a reduction in their usage — the key modules they use slowly starts to drop.”

Get a picture of your healthy and unhealthy customers and select what outcome you want from your customer health score — this will help determine the metrics you value most.

Step 2: Determine your predictive metrics

What metrics will help you determine whether you’re close to your goals for scoring (set in step one) or nowhere near it? Here are five common industry scores you can use or build off from:

1. Product usage: How often is your product used by your customers? As a general rule of thumb, people tend to use the products they love and the ones they don’t collect dust. You want to monitor how many times your customers logs into your product, whether they use all the features, and any other specific usage metric that might signal product adoption.

2. Customer engagement: How frequently does a customer reach out to your CS team for assistance, questions, or general feedback? Many companies define customer engagement as the “last touchpoint” — meaning how many days it has been since you last spoke with the customer. If a customer barely responds to your emails and has below average frequency of engagement, it’s a red flag.

3. Customer sentiment: What does the CSM think of the customer’s health? While interviewing CS managers and directors, many voiced a “gut check” from the CSM as one of their core customer health score metrics. The manager on the frontline, handling all communication with a customer has a good idea of the tonality of the feedback a customer provides and how happy they are. If you want to quantify this, you can do a sentiment analysis of the nature of support tickets.

4. Customer advocacy: How likely are your customers willing to talk positively about your product to their peers? NPS scores are often used to evaluate whether or not a customer would refer your product to someone else in the industry. If a customer is happy to recommend you, it’s a positive sign that they’re happy with your product and service.

5. Customer ROI: Are customers seeing a positive ROI in their operations since buying your product? If yes, it’s a net positive for your customer health score. ROI is an important metric because even if your customers are using your product every day and still not seeing a positive return on their spend, they’re likely to churn.

These are the five common customer health scores used across various companies. Use them individually or assign weight to each score to calculate an overall health score. For example, if your goal is to reduce churn, the question becomes, “Which metrics can help you predict that a customer is likely to churn?

Low product adoption is one — signaled by a weak DAU/MAU ratio. Unhappy customers in NPS score is another. It’s vital you pick the right metrics to track — ones closely related to your outcome.

How do you find the right predictive metrics? Evan Rich, former Director of Customer Success at NS1, recommends approaching the process like a doctor. He scoured through customer history and found key differentiators between healthy and unhealthy customer behavior for his company. One of them was annual versus monthly contracts: monthly contracts enable customers to exit the contract at any time, while a 12-month contract shows a greater commitment to the product. Similarly, if a customer used advanced features, that was a positive sign that they loved the product.

⚠️ Warning: don’t rely only on your gut feeling to find out predictive metrics. You might think a low MAU correlates to a high churn rate, but ensure the data agrees with you. Thomas Smeallie, the regional director at Planhat, also suggests looking at scoring data in context:

Don’t assume that ‘more tickets’ means less healthy. Often it's the customers who open the most tickets who are trying their hardest to adopt your product, and are most likely to end up seeing success and achieving their goals.

Step 3: Create a scoring system by assigning weights to your signals

This is the step to create your own unique formula for calculating your customer health scores.

If you’re creating a new formula entirely: decide what actions count as “negative” and which count as “positive.” For example, if a customer proactively notices new features to use and understand, it adds one point to their customer health score (positive), but if they complain of delayed response from CSMs, it reduces one point from their customer health score (negative). The critical thing is you get an overall score from the metrics you chose.

If you’re building off existing formulas: even if you aren’t creating a new formula from your chosen predictive signals from scratch, not every metric is equally predictive of your desired outcome. At Mailmodo, Dhiraj Patel, the Revenue Operations Lead, uses Vitally to assign a 20% weight to engagement status, 30% weight to product adoption, and more.

Assign a weight to each metric you’ve shortlisted in step two on a scale to quantify its impact on the overall score.

Once you’ve created a scoring system, templatize it so every CSM can use it without a hiccup. Where are they supposed to take the data from? Where are they supposed to enter it? How often should they calculate health scores and present them in standups? Make the process crystal clear.

Depending on your tech stack, everything from entering inputs to calculating a customized customer health score can be fully automated.

Step 4: Categorize your customer health score(s)

What score means a “healthy” customer, and what score means an “at risk” customer? Step four is categorizing your scores into different ranges.

Your scoring method is unique to you and how much weight you give to each metric — it can be percentile based, where a health score in the 25th percentile signals high-churn risk. Or it can be grade based, where an A signals a healthy customer. Color-coded, numbered, graded — choose whatever floats your boat to make customer health scores meaningful and useful.

Liz Russell, founder at Russell Street Consultancy and Director of Customer Success, says building binary scores that don’t take segmentation into account is one of the biggest mistakes CS teams make:

“Try to understand what customers' actual goals and needs are and create segmentations that group customers together based on what *they* want to get out of the software. I typically call these “personalization segmentations.” Then, make different health score metrics and weights for these segmentations so that you have a more accurate definition of what “good” looks like and can better calibrate your actions based on those scores.”

For example, a company signing up to use only a particular feature of your product might have low usage rates — but that doesn’t mean they’re unsatisfied. Russell encourages you to discard what you think ideal customer health should look like and focus on your customers’ varying expectations instead.

This is also the stage to visualize your customer health scores in charts, graphs, and more to make it easy for your team members and stakeholders to understand them.

Step 5: Develop your CS playbooks to react to the health score data

Knowing customer health scores is pointless if you don’t know what to do with them. What is the assigned CSM supposed to do if an active customer goes into red mode (aka low customer health score)? What strategies should they deploy to retain them?

Sonam Dabholkar, Director of Customer Success Operations at Gong, uses customer health scores in three ways in their day-to-day operations:

“First, our CSMs use health scores to prioritize their accounts. This is particularly valuable for our CSMs in higher volume segments, as they aren't able to proactively engage with every customer on a regular basis, so they look at the health scores to provide direction on where they should focus their efforts, which are typically customers with red health scores and/or low engagement.

Second, Gong have an internal risk management program and process for each of their CRM teams.

The health score being flipped to red can enter the customer into the risk management workflow, and the CSM then works with the account team to create a risk mitigation plan that is then reviewed by management during a risk meeting. In addition, the health score data from Gainsight is piped into the tools that other customer-facing teams use daily (ex: Sales, Support), so that they have visibility into the health scores in their own work with customers, and can stay in sync with the CSM and the broader account team on the health and status of customers.

Finally, the exec team uses health score data to prioritize which customers require their attention.

"Our executive team utilizes health score data in our very own Deal Boards in Gong to prioritize customers that require exec-level support and outreach, for example an executive alignment conversation to support an at-risk renewal.”

Notice how Gong doesn’t limit using health scores to the CS team? Executives and cross-functional teams can also gain valuable insight using customer health scores. For example, your marketing team can approach a good-health customer for a case study for your business. Whatever it is, ensure you map clear action items against your hard-earned customer health scores.

Customer health scores aren’t a one-and-done deal

Refresh your scoring system regularly. The first iteration you do would largely be based on your gut and what you “feel” is the right predictive signal — and you might be 100% right or way off. Expect to make adjustments frequently based on the feedback you receive while practicing scoring. Thomas Smeallie, Regional Director at Planhat, agrees:

A health score must be tested every quarter to ensure low health correctly correlates to churn rates. If you see too many "healthy" accounts churning or too many "low health" accounts renewing, then you're likely missing some key factor or assigning too much weight to a factor that really doesn't impact the likelihood of renewal.”

And even if your first-time scoring formula was off-the-charts: new customers bring fresh pain points. Businesses evolve to serve another set of buyer personas. Branding shifts, and a new feature becomes your unique selling point. All these changes happen as your company grows and expands — which means you must keep updating your customer health scores with external changes too. Sticking to your original method will mean futile results over time. Mark it on your calendar — at least annually — to update how you measure your customer health score.

Remember, customer health scores are crucial warning signals to run your business. You don’t want to wait until things get too dire to start focusing on your customers.