If you ask Customer Success Managers (CSMs) what the main benefit of their CS tool is, I would bet that they say “insight in the health of my customers” or “I want to become more proactive”. The issue is that health scores are mostly generic and lead to even more passiveness, as CSMs wait for triggers to govern their time.
How most health scores are created?
Mix the most generic usage data that you can find i.e. number of log-ins with the good old NPS score. Shake it twice and sprinkle some feature usage events in there. Hey, why not add the number of tickets that support receives? That surely influences health negatively. Voila, a health score is born.
The main issues with health scores:
1. Wrong KPIs are used or KPIs are falsely interpreted
Here are few goldies to illustrate my point:
- NPS is not correlated with retention
- The number of tickets is reversely related to retention
- The number of sessions is vanity metrics
2. The health score is generic and not focused on the desired outcome of the customer
We call this very common phenomenon the six-pack problem.
Imagine you are the owner of a gym and various types of people come to enroll. You have the sporty person (trying to look buff in the summer), the larger teen (trying to lose weight), the elderly person (coming for entertaining classes), and the urban young professional (coming for relaxation).
Everyone goes to the same gym aka. they use the same product, however, they have different goals.
Now imagine if you have so many people sporting in your gym that you want to apply a health score to your customer base to understand them better. The equivalent of assessing everyone with the same health score in SaaS is to measure every gym customer by the weight that they have lost. As we presume that every customer wants to have a six-pack. Thus, this health score logic only applies to the larger teen and even here it is questionable if the target is really well defined.
Assume now that we start to react to our generic health score and offer weight watcher lessons etc. to the sporty person or elderly man/woman, who is more interested in social interactions than sport. You will alienate them.
An unfocussed health score will not only be irrelevant, but it will also provoke dangerous messages and actions by the gym owner (CSM) that are in conflict with the real desired outcome of the client.
3. The right data is unavailable
Generic product events are not redundant to track. However, it is paramount to measure the right events that contribute to the desired outcome of a customer. The main issue is that most companies only get a mature event tracking in place in the Series A stage (>50 - 60 employees) and CS doesn’t have the internal lobby to hassle the product team to prioritize the product tracking higher on the roadmap. So for a crucially long time in the early stages of a company, most CS organizations are flying blind even when they try to apply the best practice of health score building.
4. Reacting to health scores, doesn’t make Customer Success teams more proactive
Most CS teams are still reacting to customer requests, rather than influencing the fate of the customer proactively in a positive way. Hence, the first step is always to become more proactive. The panacea seems to be the health score that tells you when to call a client that is in need before they call you.
That’s like saying: “Let’s try to predict car crashes, to be able to send ambulances before the accident participant actually calls 911.” It doesn’t solve the problem.
If the desired outcome of a client is to be crash-free, a better tactic to be proactive is to i.e. 1. steer traffic through traffic lights to avoid jams 2. educate drivers 3. make public transport more affordable, or 4. control speed limits more rigorously.
Hence, try to aim for 50% of your health score to be made up out of strategic value-adding moments that add to the achievement of the desired outcome. Once those proactive steps are taken, you can still measure the health of your client base, according to how well they are progressing towards their desired outcomes.
How to define a good health score?
First and foremost, there very rarely is only one health score, think plural. Most tools can serve multiple purposes.
1. One health score per desired outcome
First and foremost, there very rarely is only one health score, think plural. Most tools can serve multiple purposes. Imagine the case of a car lease company: persona A leases a car to commute faster; persona B leases a car to drive a great car that they couldn’t afford to buy, and persona C simply to save costs. All of these desired outcomes demand a different health score. Granted there will be a lot of overlaps, however, you cannot apply the same formula of ‘health’ to ‘persona A’, compared to ‘persona B or C’. Each desired outcome needs to have its own set of ‘outcome packages’, services, or products that bring the persona closer to their desired outcome.
If you do discover that you have a lot of different desired outcomes (>4) that you need to cater towards within your product, then you have found a symptom of a non-customer success focussed company.
Probably management has prioritized the acquisition of new logos over retention and making logos successful, which leads to having numerous rivaling desired outcomes that are claiming CS & product management attention.
2. Health scores must follow customer journeys aka. success journeys
If you watch the news in the summer you are probably familiar with segments about the Tour de France and the map of the course that the cyclists have traveled in that day’s ‘stage’.
If we know that our customer’s desired outcome is to finish the stage, we know what it takes to be successful and which milestones need to be passed roughly at what time. Hence, we can plot our health score according to the success journey of the customer including activities (log-ins etc).
Beware that purely passive check-in’s according to the success journey is obviously not proactively helping the customer. We must proactively support the customers in various stages of the journey, just like cycling teams provide information & nutrition along the way.
3. Analyze and repeat
Just as your product changes one must revisit the desired outcomes of the customer as well. The desired outcomes of customers will evolve and change over time, which then demands a recalibration of their health score. Moreover, in your quest to optimize the journey according to i.e. Time to value in onboarding, you will learn new accelerating methods to make the customer successful quicker. Our recommendation is to revisit the customer journeys and health scores at least once every 6 months, to see if they are actually predicting health correctly.
Do you want to be coached through the journey of building relevant health scores?
This is the SaaSCollective’s model of the two-sided coins. Why two-sided? Because a health score must be 50% proactive & 50% reactive. Why coins not coin? Because almost every company needs more than one health score, depending on the number of desired outcomes found in their customer base.
- Start with gaining insights into the usage patterns and define clusters
- Define what success looks like in your company (MRR; renewals; easy on support; good culture fit etc.). In short, within each cluster - which logos do you want to copy-paste.
- Get on the phone with them for 45min to ask them about their success journey (not optional)
- Analyze the data and define desired outcomes and success journeys that the customer has taken
- Aggregate in a health score
- Rinse and repeat for other usage clusters or desired outcomes