Every SaaS company needs a customer health score, and almost all of them have one in some form or another. Building a truly effective customer health scoring system is not a one-time event however, it's an ongoing practice. Also, as the SaaS industry advances, new best practices for customer health management are developed, and it's important to incorporate them into your organization to stay competitive, which is why this article asks...

What do companies need to know about health scoring in 2020?

Irit Eizips is the Chief Customer Officer & CEO at CSM Practice, a Customer Success consulting firm that has helped companies increase net retention, improve customer satisfaction, grow customer advocacy, and grow bottom lines. Irit has been named one of the top Customer Success and CX Strategists globally, every year since 2013, and in this interview, she imparted a wealth of advice around customer health scoring in 2020 and moving forward.

Watch the interview below, and then read some of our key takeaways about Customer Health scoring in 2020.

What is a customer health score?

A customer health score is a system made up of a series of metrics that indicate how customers are progressing towards a desired outcome. Some companies may be more focused on using a customer health score to identify customers at risk of churning, while others might want to focus on identifying which customers are eligible for upsells. Generally, companies try to build health scores that show them the likelihood of multiple outcomes.

Why does customer health scoring matter so much?

There are many benefits, however some key ones to note would be:

• Ability to identify top performers
• Ability to identify upsell/expansion opportunities
• Ability to perform accurate revenue forecasting
• Ability to identify churn risks far enough in advance to prevent them
• Ability to use data to drive desired outcomes for the company and the customer
• Ability to assess the key trends of best-in-class customers
  

Why are many customer health scores outdated?

As the Customer Success industry has progressed, it has become clear that having just one customer health score causes confusion, frustration, and is often ineffective. Some of the reasons for this are that companies don’t update them regularly, they include vanity metrics in their scores which overcomplicate them, and they don’t take customer maturity modelling into account. 

What is a customer maturity model?

In the same way that riding a bike or doing a somersault is impressive for a 3 or 4 year old, but not impressive at all for a 20 year old, how companies define customer health changes with customer maturity. A customer maturity model will tell the company whether or not a customer is healthy considering which stage they’re in. A customer who has been with the company for 1 year, should not be measured with the same exact yardstick used to measure a customer that has been with the company for 5 years.

A customer maturity model is also important to show to customers, because it gives them a roadmap of where they can and should aim to be. It allows them to benchmark themselves against comparable companies and understand if they’re doing well or not, and what they can be doing better.

If a CSM can tell their customer champion “Look, your team is doing better than our comparable customers when it comes to product utilization, but where you lag behind our other customers is in executive sponsorship. Our top performers have an executive sponsor actively supporting this, so if you’re looking to derive even more value, I think that’s what we should aim to work on.”, it’s an extremely helpful conversation for both the CSM and the customer.

Generally people can be quite competitive, so sharing the data around “these are the 6 or 7 things we see our best in class clients do” can be an excellent motivator, and works much better than merely nagging or pressuring the customer to do things.

What is a customer health score (as defined by this new framework)? 

This version of a customer health score is similar to the previous ones that have been around for a long time but with a few key differences. Instead of the old way, which aimed to be everything to everyone, the new version should be designed specifically to make CSMs more proactive, allowing them to quickly assess client needs and take the necessary actions. It should be something the CSM can check daily and know exactly what they need to focus on.

Customer health scores should also stay internal. Specific metrics can be pulled out to share with clients, but generally, customer health scores are built for CSMs and the metrics they care about. What customers care about could be entirely different, so it’s best to carefully plan what should or shouldn’t be externalized, and validate it with a customer champion before sharing with a larger group.

As mentioned previously, the customer maturity model is much more ideal to share with customers so that they can improve and take responsibility for becoming best-in-class, whether that means adding additional resources, increasing internal communication, or even changing internal processes.

Everything externalized to the customer should be positioned as something being shared for their benefit and their improvement.

What should be used for revenue forecasting?

Both a customer health score and customer maturity model can be extremely useful for forecasting revenue. Depending on the size and complexity of the company, additional tools, like a dedicated renewal probability score and even a dedicated expansion probability score might be called for.

Irit recently interviewed Taylor Johnston, Director of Customer Success at Salesloft about how she built a revenue forecasting system that is accurate within 2% (click here to watch) and because of the size and complexity of her organization, they have separate pipelines to manage renewals and expansions, on top of their regular customer health scoring practice. 

What are some common mistakes people still make now with health scores?

One common mistake companies make is merging risk indicators with a true health score. People do this to try to keep things simple, however it leads to high numbers of false positives and false negatives. It also leads to the obscuring of root causes which makes it harder to improve customer health. It needs to always be perfectly clear that if a customer’s health is “yellow” there is an easily-identified root cause. It is easy to set up a calculator and call it a health score, but it’s much harder to set one up that is always true and gives clear takeaways.

Another key mistake that is quite common is trying to manage long tail customer segments with limited or inaccurate data. It is impossible to manage these customers manually, so companies need to rely on data, and at scale, it can be difficult to pull in specific data points like “how much value is the customer getting?” In addition to difficulty getting data points, companies often have inaccurate data which can completely negate the effectiveness of their health scores. It is critical for these companies to use only validated data points, invest in the technology or engineering resources needed to map data points to specific accounts, and automate plays wherever possible.

What are 3 tips companies can start thinking about right now?

Having a health score in place is great, but it’s not enough. 

Companies should:

• Create playbooks that clearly state what action would be taken for each metric if it turned red. Different causes require different solutions.
• Determine how leadership can best incorporate all this new information into quarterly strategic planning sessions. 
• Decide what the strategy will be for updating/changing the health score
  

When and how should companies change/update their customer health score?

Changing a health score is not something that needs to be done after a specific amount of time. It needs to happen whenever something is not working (i.e. false positives, false negatives), or when something is no longer relevant (i.e. metrics that keep getting greener and greener, after a time become irrelevant to customer health).

The way to change a customer health score is by either adding or removing metrics altogether, or by changing the weighting system. If a specific metric is almost always green for all customers, it should be weighted less than metrics that fluctuate and might be stronger indicators of churn risk. Whenever it becomes clear that a specific area needs improvement, the health score should be adjusted to reflect that.

What is a quick way for a CS team to improve their customer health score right now?

An easy way to make meaningful change to a health score is by analyzing current metrics and asking the question “out of these metrics, which ones, if they’re red, mean the customer is at least 70% likely to churn?” If metrics are only 50% likely to predict churn, they aren’t going to be very helpful, so they can be removed, or the weighting system can be adjusted accordingly.

Conversely, CS teams can ask “out of these metrics, which ones, if they’re green, really reflect a best-in-class customer?” and make their desired adjustments. 

Building and maintaining a customer health system is no easy feat, but hopefully this article provides the information needed to either get the ball rolling, or catalyze further progress towards building and improving the systems already in place. 

For more information make sure to subscribe to Catalyst and to CSM Practice. Thank you to Irit Eizips for her contributions to this post and to the Customer Success industry in general, which she has played and continues to play a significant role in shaping.