Most of us know that, periodically, even the healthiest people need to go and get a “just in case” checkup. It’s why life insurance companies require a physical.
However, when it comes to customers, many SaaS companies rely on an “if it ain’t broke, leave it alone” model. As a SaaS founder who instituted health scoring as soon as we acquired our 25th customer, it’s hard for me to wrap my head around the sheer number of companies who aren’t tracking the health of their customer base.
If your customers have gone years without a routine checkup, here are the steps you should take to get back on track.
What does a healthy SaaS customer look like?
This varies for everyone, but deciding what a healthy customer (and an at-risk customer) looks like is your first step toward accurately assigning a health score. Take a look at the various activity, actions and usage that indicate customer health. Here are some to consider:
- Customer pays (for freemium products)
- Customer uses product with X frequency
- Customer uses X feature(s)
- Customer account has X number of users
- Customer account includes multiple key stakeholders
- Customer account includes executive-level relationships
- Customer has desired outcomes met
- Customer is using support resources
- Customer is active participant in community
- Customer has explicitly stated their satisfaction (not a fan of this one, but…)
It’s also worthwhile to include some “at risk” flags for scoring as well. Here are some examples:
- Lost customer stakeholder
- Lack of adoption of key features
- Business event occurred (such as customer being acquired or an executive team being replaced)
- No login in past X days
Check their vitals
There’s a reason all doctor’s visits start with the same blood pressure and temperature check. It’s impossible to predict what might be wrong if there’s no baseline for health in the first place. The same goes for customer health scores; metrics act as your vitals. If those are off, then you can’t start searching for bigger problems. However, choose your metrics carefully because testing for too many variables at once could lead to a misdiagnosis.
Pick no more than 3-5 metrics to determine your customer health score. Anything more than that will be untenable and unscalable to track. Make the score itself super simple—you can actually use a score, from 1-5, but I personally love red (at risk), yellow (unsure or midline), green (healthy) as indicators of health.
This is about a total picture of the customer in just a few metrics, so you need to pick your metrics wisely. That’s why companies can get paralyzed at this step. Which health metrics should they use to score? How should they weight each? Do they each need to be weighted the same? Do all metrics need to be objective or can some be subjective?
Just get started. Pick some key metrics, and then score the customer base. Take a look at the results. Do they seem accurate? Spot check some accounts you are really familiar with. Is the score representative of the customer’s health? If you score the customer base and then look at the results and they seem pretty accurate, then run with it. If they don’t seem right (somehow your most deeply adopted customer who just signed a 3-year renewal ended up with an “at risk” flag, for example), then just tweak it.
Keep customer scoring simple, above all else, so you can manage and maintain it on an ongoing basis.
Know your one key metric for health
When it comes to your product, you’ve got to think like a specialist rather than a general practitioner. You’ve definitely got one product-specific metric that’s more important than all the others. If this metric isn’t there, you are dead in the water in terms of product adoption and customer retention.
The one key metric is unique to each product, so you will need to figure out what yours is. It’s the metric that matters more than any others—it’s pass/fail. It could be that a user sends an invoice if you are a financial application (not that they create an invoice, but that they actually send it). It could be receiving end-user traffic if you are a web-publishing application (not that a page or site is created, or that it goes live but that it actually receives traffic).
You will have to figure out what your one metric is, and when you do, everything starts from there—it’s the thing you design for and drive your customer towards, because without it they will never achieve their desired outcome(s). You, and your entire company, must know your ONE metric and monitor it religiously. It’s the metric that is pass/fail. Without it, the customer health score is zero, no matter how many other elements of health they may have.
Don’t just diagnose the problem. Treat it!
Imagine if a doctor said, “Yep, looks broken,” and simply sent you home. If your customer health scoring works, chances are, you’re going to spot some problems. Make sure you have a plan in place for intervention and correction for your yellow and red customers, and ways to reward and engage your green customers even further. The health statuses are just that—a status.
Yes, it is important to know things like “80% of our customer base is green, 10% is yellow and 10% is red”, but what is more important is to actually do something about it. What’s your plan to move your yellows to green or to red? Because yellow means customers are sort of hanging out in no man’s land—neither successful or unsuccessful, but somewhere on the bubble. What’s your action plan to tackle your red customers and bring them back to health? Your company needs to know how to proactively keep customers in the green, and also how to handle it when a customer isn’t healthy. A customer success/retention playbook can help with this.
Use your customer scores to stay healthy in the long run
Ideally, your customer health score isn’t just a score that helps you see who is healthy and who is at risk. That’s important for retention, sure. But your scores can help you improve your business and your product when you look deeper than the surface. Let’s say you are tracking how long it takes a customer to activate a certain feature, as a health metric. And that metric, in general, is longer than you deem acceptable. Now that you are tracking it, you can seek to improve it. Why is the feature taking longer to activate? What’s the barrier to adoption? You can make product or process changes to help shorten that. Use your scores to track health but also use them to uncover areas of opportunity for improvement in your product and your processes.