Feb 23, 2021

Read Time 10 min

Q&A: Surprising Insights About Customer Health Scores and Its Impact on Recurring Revenue


Think customer health scores are only good for calculating renewal likelihood? Guess again.

When based on the right mix of behavioral- and event-based criteria, these leading indicators predict so much more. How are today’s Customer Success teams using these forecasting factors to get and stay ahead of their customers’ needs and wants?

To find out, we invited CSM Practice to share benchmarking data from their recent survey of Customer Success professionals on how they build, maintain, and measure customer health scores. During the webinar, we cover:

  • Who uses customer health scoring
  • The true impact of customer health scores
  • Current customer health score trends

If you missed the webinar, you can watch it on-demand.

Chief Customer Officer, CSM PracticeQ&A Recap

Speaker: Irit Eizips, Chief Customer Officer & CEO, CSM Practice

Q: What’s the maximum number of customer health score factors you recommend? What factors are the most common?

A: It really depends on what you’re trying to pull off. If you were to create a health score – that’s really a churn index – you want to have as many components that typically trigger a high risk for churn. Usually, that’s about 8 or 10 metrics at the most.

Now, if you’re going to create a health score, that’s an indication for the CSM to say, “What should I be focusing on?” That could include a lot more because it will always be true. A red [score] means you haven’t done X with a client. A customer who has a lower score doesn’t necessarily mean that they’re going to churn, but they’re not getting enough attention, or they’re not on the right track to become best-in-class. For those types of scores, I’ve heard executives having anywhere between 8 to 20 different metrics. If you do have a lot, you want to group them by sub-metrics (e.g., all metrics related to relationships or all metrics related to expansions), so you have sub–health scores that aggregate to the tool of the overall health of the customer.

But there’s a distinction between health and help. There’s a churn index and then there’s a health score to ask, “Are they becoming best-in-class?” I think that if you could do that, the best-in-class type of health score, the one that indicates if they’re getting enough value, is always true, then it doesn’t matter how many components you have in there. It’s not uncommon to have 20 components.


Q: Should you use different customer health score models for different customer segments?

A: Yes, especially if customers are using, for example, completely different products, it might make total sense to have completely different health scores. You can also have a different health score for onboarding versus ones for customers that have the system in place. This would be extremely important for those of you who have an enterprise application that’s a bit more robust, where the implementation process might take a little bit more time. It’s less important for the [customers] who self-serve. In that regard, you might want to have a health score for if you have a proof of concept or a trial period, or anything that highlights different phases in the customer life cycle.

You can also have a health score for different propositions or goals. You could have a health score that indicates if they’re ready for expansion; it could be an expansion index. What are you going to include in that? Setting up a health score that shows customers who are on their way to becoming best-in-class and “super green” doesn’t necessarily mean that they’re ready for an expansion. There might be other things in the usage data that tell you they’re ready for an expansion. For example, you can have a complete statistical analysis on your usage data and say, “Customers who have these three features usually opt into these five modules.” If you can identify all the customers who have these three features and did not opt into these additional [modules], they might be a good customer base to approach with an initiative to say, “Hey, other customers are using this [module]. We benchmark you against others. This is what we’re finding. Let’s take a look if this is applicable for you.” I want to break the notion that a green health score means expansion, and a red health score means churn; that assumption is dead. We already know it doesn’t work. Let’s build different health scores for different things, and let’s name them differently, and everyone here will have a much better way to be proactive, predictive, and efficient with their customer base.


Q: How should you evaluate the effectiveness of a customer health score?

A: Look at your key KPIs as a team. You want to set up a health score and ask, “What is the purpose of it?” If it doesn’t help you get there, then obviously something is wrong. If your health score is on occasion producing false positives or false negatives, then the customer health score doesn’t work either because it’s not a reliable metric. You want to create a metric that you can always rely on. If the metric components are just not producing consistent results, then it’s definitely something that you want to reconsider and reconfigure.


Q: How should you include ROI factors into a customer health score?

A: To understand what ROI is for your customers, you have to send them an outline with five or six different outcomes your customers can achieve. Then ask yourself, if I was the executive on the client side, what are the key metrics that I’m concerned about and how does this outcome relate to my key metrics?

Let’s say, you sell into the finance office, and their key metrics are to complete their accounting cycle quickly or without delays. Then, you can say, “I have an application that can help them with that. Are there certain modules that can support the cycle to be more accurate and faster?” Then, you want to say these are the modules that we recommend for this KPI. Now, you can solidify a ROI discussion with these customers. The best way to do it with your CSM is to identify those key use cases and ask, “Has the customer actually had a moment of value with us? Were we able to enable something for them, or change something in their organization, so that they can get this outcome?” It’s most likely a manual metric in that regard if it’s a high-touch model.

If it’s a low-touch model, some applications have this notion, let’s say, a marketing application where you have a ton of leads, and your goal is to get them to have more revenues or more conversations with prospects. You can certainly count that: how many views did you have for every lead page? How many phone calls did you get based on that lead page? You can show ROI really quickly. It all depends on your solution and whether you could do it automatically.


Q: Should a customer health score have more qualitative data (which may be subjective from a CSM) or more quantitative data?

A: It depends on the strategy. For a high-touch model, qualitative is one of the ways to go. Again, if you’re thinking about a health score as an indicator of the customer becoming best-in-class, then qualitative is not a problem. The trick is to make sure that you have clear definitions, so it doesn’t become whatever I think, but these conditions have been met, therefore, this client is in green, yellow, or red. Then for quantitative, it would be great if it’s data that’s really hard to access. You have to manually run a report. It’s time consuming. If you can get it from a system, put it in your Customer Success software, create a quick metric that’s calculated for the CSM, and introduce productivity around that. There’s nothing wrong with that if it helps the CSM focus on the right things.


Q: Aside from renewal management, how else can you use a customer health score?

 A: I’m a big fan of using it for the onboarding phase. You definitely want to have a health score for that. The other way to think about it is by asking “What does the adoption phase look like?” That’s the most predominant use and where you want to maximize value for the client and make sure you have a strong partnership with them. For that phase, you would define it based on a value score. How much value are they getting? Are they in my user community groups? Are they attending my webinars? What’s the level of engagement? Do I have the right relationships? You want to include those kinds of metrics, besides the usage, of course. Have they taken training? Do they have a dedicated admin? Start outlining all those things that ask, “Is this the best-in-class customer?” That should be your primary health score.

On top of that, you could identify all the different triggers, where if these incidents happen, that means customers at 70% probability or higher are definitely going to churn. That should be your churn index, or your code red flag that’s being calculated separately. Customers may be best-in-class and using the system and getting all the resources, but you still might lose the executive sponsor. All of a sudden, you get a new executive that’s gung-ho about a different solution. You want to prepare for these things, and make sure that you tackle all angles by having the appropriate metric.


Q: What qualitative factors should you include in a customer health score to measure a relationship?

A: I highly recommend watching Ziv Peled’s video on our channel because you actually get the screenshots of everything that he included. I’m just repeating; he’s the thought leadership here. He asks, “How deep is the relationship? Do I know their family, or did I just meet them once? On a score from 1 to 5, how well do I know them? Can I score the level of the relationship?”

At the tactical level, am I buddies with the technical admin? The technical admin doesn’t influence the decision for renewal or own the budget. Do I have an executive alignment? This could be another way to score a relationship. But again, I highly recommend hearing it from the guru’s mouth and listening to that video. There’s a lot of great insights that Ziv shares around this topic.


Q:  How should you deal with null values for a customer health score factor?

A: That’s more common. Nobody talks about it – the old N/A. Sometimes, you’ll have customers and the best practices for them is to activate or mandate X feature. So, you had a conversation with them, and they have a different tool for that specific feature. There’s no way they’re going to integrate something, because they already have something else in place by the time they purchased you, and they’re not interested at this point. You’re going to put them in red. You know that this isn’t the use case for them, or it’s not applicable for them. We encourage our customers to put things in N/A so they don’t get major red scores where it’s not even applicable. You already validated that point. It’s not relevant for this customer. Put it on N/A or a white score, so that it doesn’t impact your formula.

But I will say that it needs to be used sporadically. It’s not something you want to encourage all your CSMs to do, especially if you think that maybe next year you should open the conversation up again. If you have a lot of white scores, by the way, that means that you might need to have a different score for a cohort of clients. Let’s say, you mark white for all your financial institution clients because they’re banks and they don’t need that feature. You should have a different formula for different industries.


Q: Is there a correlation between CSM rotation and a customer’s health score?

A: The health score is not just to predict renewals, and it’s not just to give the CSM an area of focus. If CSMs leave a lot, that means that the health score is being abused a little bit. To say, “CSM, you’re not doing your job properly” versus looking at it as an opportunity to actually give yourselves, as team leaders, perspective on where should we, as a company, invest more? Let’s say, we don’t have a lot of executive relationships. Maybe that’s an area that we can work on as a company. At the executive level, we need to think “Should we build an executive alignment program? How can we support the CSM at the top level to get the results that we want?” It usually is an indication that there’s a lack of policies or processes that we can embed in the system, or tools that we can give the CSMs, or additional webinars to create awareness that we need to address at a higher level and add budget to improve overall. When you’re seeing a high correlation between CSMs leaving the company and a lot of red accounts, that means there’s a lack of ownership for these trends from the leadership team. They’re not using the customer health score in an appropriate manner.


Q: Should you create different customer health scores for low- and high-touch accounts?

A: Yes, because for the high-touch accounts, you have qualitative metrics that you don’t want to miss. For the low-touch accounts, you want to put more emphasis on their engagement with your digital assets. The way that you would use the health score for the digital assets, the self-serve cohort, or the SMBs, is to say, “What are they participating in? What are they not? Where can I make more investments? What works and what doesn’t?” Whereas, for the high-touch accounts, it would be “How can we make them best-in-class? Do we have the right relationships?” It’s a little bit different questions. At the end of the day, you want everybody to get maximum value. You’re just doing it differently. It’s not uncommon for the SMBs, the longtail customers, to have a separate team. To make sure that cohort is successful and supported, you’ll probably hire different types of CSMs to handle that cohort as well. I’d definitely recommend different health scores for those segments.


Q: For customer health score management, where do CRMs fall short compared to Customer Success platforms?

A: Where do I start? CRM systems are nice. You set it up. You get some dashboards. But it’s not easy to maintain. If you need to make any changes, you need to call IT to make the changes. They don’t have a sophisticated algorithm. If you wanted to have some sort of a rule behind it that automatically calculates the score in a more sophisticated way, the CRM systems are going to fall short. Customer Success systems are much more adapted in making sure that your customer health score is adjusted to what you need. They can do sub–health scores. They can update the metrics and show you where your formula fails. In CRMs, the health score is very basic; it’s a starting point until you get Customer Success software.

To learn more about the customer health score benchmarks from CSM Practice’s survey:

Customer Success Around the Web

Fighting Churn is a newsletter of inspiration, ideas and news on customer success, churn, renewal and other stuff and is curated by ChurnZero





Subscribe to the newsletter   

ChurnZero product release notes: Q1 2023

Here’s a list of ChurnZero’s major product releases in Q1 of 2023.   Work faster with Customer Success AI (CS AI™) CS AI is the first generative artificial intelligence built into a Customer Success platform, and the first designed for the CS industry. CS teams have...

Seven customer community building tips for SaaS CSMs

This article is by Peter Adams, customer enablement team lead at ChurnZero, and Branden Maue, community manager at ChurnZero. There’s a closely held growth secret among large B2B professional services firms, like consultants, accountants and law firms, that Customer...