Have you ever noticed those “How am I driving?” stickers on the back of semi-trucks? You know, the ones that provide a phone number to call if, for some inexplicable reason, you feel the need to critique a truck driver you’ve never encountered? Absolutely ridiculous! The folks who actually dial those numbers are the same ones who get kicked out of their kids’ soccer matches for yelling at the teenage referees.

Now, that’s not to suggest that feedback isn’t valuable at the appropriate time and in the right context. In reality, if you work in customer success at a SaaS company, consistently gauging your customers’ satisfaction is a crucial part of your role. Understanding a specific customer’s level of contentment and engagement allows your team to make well-informed decisions moving forward. Is this customer a good prospect for an upsell? Are they facing difficulties with certain product features? Are they at risk of leaving if you don’t intervene? The answer to these questions – and many others that are equally insightful – lies in customer health scoring. To delve deeper into this topic, I had a conversation with Phil Kowalski and Bobby Kittredge, two managers from our customer success team. By the time you finish reading this article, you’ll have a solid understanding of:

• The concept of customer health scoring
• Its importance for SaaS companies
• How to implement it successfully Let’s dive in!

Decoding Customer Health Scoring

Simply put, customer health scoring is evaluating customers based on their likelihood of achieving an outcome that’s important to your business. In essence, it allows you to anticipate potential shifts in your relationship with a given customer in the future, equipping you with the information needed to enhance your interactions.

Consider expanding on this. If this seems a bit unclear, it’s because different companies prioritize different outcomes. While Company A might focus on predicting customer churn (ending the relationship), Company B might prioritize predicting upselling opportunities (encouraging upgraded purchases). Keep in mind that there’s no rule limiting your company to predicting just one outcome. The key takeaway is that there’s no universal, one-size-fits-all approach to customer health scoring.

The Significance of Customer Health Scoring

It’s challenging to concisely summarize the importance of customer health scoring, but here’s my attempt: Satisfied customers whose expectations are surpassed become your most effective marketing force. Reflect on a time when you encountered exceptional customer service, whether from a SaaS company or another business. It likely made you want to shout your appreciation for that company from the rooftops, right? Your customers feel the same way. The higher the quality of service you provide – acknowledging that “good service” varies from customer to customer – the more likely you are to generate positive word-of-mouth. Customer referrals hold immense power, especially when you’re selling a high-priced product that demands a long-term commitment. Providing top-notch customer service not only enhances the retention of your existing customers but also boosts the acquisition of new ones.

Image source: ReviewTrackers. Customer health scoring is the cornerstone of providing exceptional service. Once your system is in place and you’ve collected the necessary data, you’ll have the insights required to address each customer’s specific pain points. In turn, you’ll benefit from reduced churn, improved retention, faster acquisition, and overall greater success. Sound promising? Let’s proceed to the five steps involved in establishing a robust customer health scoring system for your company.

Mastering Customer Health Scoring in 5 Steps

At its core, customer health scoring is about making informed predictions. This leads us to a critical question: How do you formulate these predictions? That’s precisely what we’ll cover in this section. To simplify, here are the five fundamental steps involved in customer health scoring. As a bonus, we’ll include a best practice for each step.

1. Pinpointing Your Outcome

Fundamentally, a customer’s health score hinges on the probability of a specific outcome – churn, renewal, upsell, etc. Ideally, a quick glance at a customer’s health score should tell you “They’re likely to churn soon” or “This presents a great upsell opportunity.” The first step in building your customer health scoring system is determining the outcome you want to predict. This choice depends entirely on your company’s unique circumstances. If you’re struggling to retain customers for more than a few months, predicting churn might be your focus. Conversely, if your retention rate is impressive and you want to generate extra revenue from existing customers, predicting upsell potential might be more relevant. Remember, you don’t have to limit yourself to a single outcome; I’m focusing on a single-outcome model for simplicity’s sake. You might find that starting with one outcome and gradually transitioning to a more intricate system works best for your company. Best practice: Organize (and maintain) your customer data. As you delve further into this article, the importance of data science in customer health scoring will become increasingly evident. Accurate and current customer data is non-negotiable; without it, your health scores won’t reflect reality. Therefore, ensuring data cleanliness is crucial before moving on to the next steps. Fortunately, you don’t have to handle this alone. Gainsight, a company we at nexus-security have partnered with, offers various products and services to centralize all your relevant customer data. We highly recommend checking them out!

Image source: Gainsight.

2. Identifying Predictive Signals

After defining the outcome you want to predict, you’ll need a way to gauge the likelihood of that outcome. This is where predictive signals come into play. In essence, a predictive signal is any customer behavior linked to your chosen outcome. Let’s continue with our example of predicting customer churn. To identify the most effective predictive signals, ask yourself: What behaviors might suggest a customer is at risk of churning? For now, let’s stick with three possible indicators:

• Infrequent product use: A customer who’s barely used your product in recent weeks might be at risk of churning.
• Limited feature use: A customer who only utilizes a small fraction of your product’s features might be at risk of churning.
• Unsatisfactory results: A customer who isn’t seeing a significant positive impact on their business from your product might be at risk of churning. Choosing the right predictive signals is crucial: The stronger the correlation between your signals and the outcome, the more accurate your customer health scores will be. Imagine each signal as an ingredient and your health scoring system as the final cake. Just as the wrong ingredients result in a bad cake, the wrong signals make your health scoring system ineffective. Best practice: Scrutinize your assumptions. Selecting predictive signals inevitably involves making assumptions. For example, you assume a positive correlation between a customer’s product usage frequency and their overall health score. While assumptions are unavoidable, you can minimize their potential negative impact by staying on top of your data. If you notice a significant churn rate among frequent users, it’s time to reassess your signal choices.

3. Assigning Weights to Your Signals

One crucial aspect that might not be immediately apparent after selecting your signals is that they don’t all hold equal weight in predicting the outcome you’re tracking. For instance, while the breadth of features a customer uses during each product session might be somewhat indicative of churn, the impact your product has on their business outcomes is likely highly predictive of churn. Treating signals of differing importance as equals will compromise the accuracy of your customer health scores.

This is where assigning weights to each signal becomes essential. Think of your signals as ingredients and their weights as the measurements. Just as a cake requires precise measurements for a delicious outcome, your health scoring system needs carefully considered weights for meaningful results. So, how do you determine these weights? Best practice: Consult your customer success representatives. As the primary point of contact with your customers, they have valuable insights into which signals carry more or less significance. If some representatives have clients who use few features but love the product, while others report the opposite experience, it’s safe to infer (for now) that the breadth of product usage is only a moderate predictor of churn.

4. Establishing a Health Score Scale

You’ve successfully defined your desired outcome and outlined how to predict it. Surprisingly, you’re now equipped to assign health scores to your customers. However, without a benchmark for comparison, these scores remain meaningless. For example, a score of 70 carries different implications if the highest possible score is 100 compared to 1,000. This is where developing a health score scale becomes essential – to make your customer scores meaningful and actionable. Using a familiar 0-100 scale is a safe choice. This scale also simplifies the process of categorizing customers into different buckets. For example: Scores between 0-39 could be labeled unhealthy, 40-79 as neutral, and 80-100 as healthy. This scale can be made even more intuitive by assigning colors to each bucket: Red for 0-39, yellow for 40-79, and green for 80-100.

Despite its simplicity, this approach significantly impacts your health scoring system. Suddenly, each customer’s score communicates a clear message about their relationship with your company. Best practice: Emphasize specific customer segmentation. For simplicity, we used a traffic light model (red, yellow, green) as a scale example. While this might work for some, the general rule is that more segmentation leads to better outcomes. “Segmentation” refers to the level of detail used to categorize customers based on their health scores. Since you’ll be creating strategies for each score bucket, more buckets allow for more targeted strategies. This, in turn, increases the likelihood of your customers feeling understood and their needs met.

5. Strategizing Based on Scores

The ultimate goal of customer health scoring is to empower your team to improve customer relationships in a logical and mutually beneficial way. Therefore, planning your response to each customer’s score is crucial; otherwise, your hard-earned data becomes useless. For instance, you could offer discounted pricing or free access to premium features to incentivize “unhealthy” customers to stay. On the other hand, you could gauge the interest of “healthy” customers in upgrading to a premium subscription.

Best practice: Regularly review and update your system. Since your initial health scoring system relies heavily on assumptions, adjustments will likely be necessary. Expect things to deviate from your initial plan – your system’s success hinges on your willingness to adapt. Moreover, new customers bring new challenges and needs. As your company grows, so will the diversity of your customer needs. Stubbornly clinging to your original system will leave a significant portion of your customer base behind. Alternatively, regular updates (annually is a good starting point) will ensure its continued effectiveness.

Make Customer Health Scoring a Priority This Year

If you’re questioning the necessity of customer health scoring, ask yourself this: Is long-term success for my company important? The answer is a resounding yes! Your SaaS company’s longevity is directly tied to the success of your customers, which, in turn, depends on your ability to provide solutions to their problems. Implementing a customer health scoring system significantly enhances this ability, giving you the insights needed to provide the best possible service to your customers. A huge thank you to Phil and Bobby for their invaluable contributions to this post.