SaaS companies live and die by customer retention, but most teams are flying blind when it comes to knowing which customers are actually thriving, and which are quietly drifting away. Automated customer health scoring is supposed to solve that—if you set it up right. If you’re using Churnzero, this guide will walk you through setting up a health scoring system that’s actually useful, not just another dashboard nobody trusts.
This is for SaaS folks who want clear signals about their customers, without drowning in nonsense metrics or fiddly setups. You’ll get step-by-step instructions, a dose of skepticism, and some hard-won tips that save time and headaches.
Step 1: Get Clear on Why You’re Scoring Health (and What to Ignore)
Before you even open Churnzero, nail down what you actually want to know. Don’t fall into the trap of tracking a dozen barely meaningful numbers, just because the tool can.
Start by asking: - Who’s this for? (CSMs, execs, product managers?) - What do you want to act on? (Renewal risk, upsell potential, onboarding red flags?) - What’s actually predictive of churn or success in your business?
Ignore: - Vanity metrics (logins, pageviews) if they don’t drive outcomes. - Stuff you can’t get reliably or consistently. - Overly complex formulas that nobody will maintain.
Pro tip:
Talk to your CSMs and sales teams. They usually know the warning signs before any tool does.
Step 2: Gather the Right Data (Don’t Overcomplicate)
Churnzero is only as good as the data you feed it. You want a handful of key signals, not a laundry list.
Common data sources: - Product usage (feature adoption, usage frequency) - Support interactions (tickets opened, NPS scores) - Account milestones (renewals, plan changes) - Billing status (late payments, upcoming renewals)
How to get data into Churnzero: - Use their native integrations for CRMs like Salesforce or HubSpot. - Set up API connections for product usage/events. - Import CSVs for one-off or historical data.
What to skip:
Don’t try to automate everything at once. If you can’t get clean, reliable data for a metric, leave it out. Bad data is worse than no data.
Step 3: Define Your Customer Health Segments
Now, decide what “healthy,” “at-risk,” and “unhealthy” really mean for your business. Churnzero lets you define health scores using rules and weightings.
Tips for defining segments: - Keep it simple—3 to 5 segments is plenty. - Make the difference between segments clear and actionable. - Use a mix of positive and negative signals (e.g., high usage = good, support escalations = bad). - Don’t obsess about the exact cutoff points at first. You’ll tweak these later.
Example: - Green (Healthy): High feature usage, on-time payments, no recent support escalations. - Yellow (Caution): Drop in usage, late payment, or negative NPS. - Red (At Risk): No logins in 30 days, multiple support tickets, overdue invoice.
Pro tip:
If you can’t explain what makes an account “red” in a single sentence, your definition is too complicated.
Step 4: Build Your Health Score Logic in Churnzero
Time to get into Churnzero and actually set this stuff up.
4.1. Create Health Score Categories
- In Churnzero, go to Health Scores > Categories.
- Create categories for your main signals (e.g., Product Usage, Support, Billing).
- Assign a weight to each category. Don’t overthink the math; just reflect what’s most important.
Example weights: - Product Usage: 50% - Support: 30% - Billing: 20%
4.2. Add Health Score Rules
- For each category, add rules that define what’s “good,” “neutral,” or “bad.”
- E.g., “Logged in at least 10 times in past 30 days = Good”
- “NPS score below 6 = Bad”
- Set point values for each rule. These add up to the overall health score.
Honest take:
The default rules Churnzero suggests are fine as a starting point, but they’re generic. Customize them for your actual business patterns or you’ll just get noise.
4.3. Map Scores to Health Segments
- Set score thresholds for each segment (e.g., Green = 70+, Yellow = 40-69, Red = below 40).
- Don’t chase “perfect” thresholds—pick something reasonable and plan to revisit.
Step 5: Test, Tweak, and Reality-Check Your Scores
You’re not done just because the dashboard lights up. Test your setup:
- Review health scores for a handful of accounts you know well. Do they make sense?
- Ask your team: “If you saw this account marked Red, would you believe it?”
- Track a few accounts over time—does the score move the way you expect?
What to ignore:
Don’t obsess over making the scores “look” good to execs. If the truth is ugly, it’s better to know now.
Pro tip:
Schedule a regular review (every quarter or so) to adjust your rules and thresholds as your product, customers, or strategy changes.
Step 6: Set Up Alerts and Workflows
A health score is pointless if nobody acts on it. Use Churnzero’s automation features to make the scores actually drive action.
- Automated alerts: Set up notifications for CSMs when an account drops into “Red” or “Yellow.”
- Playbooks: Trigger predefined actions (email templates, check-in calls) based on health changes.
- Reporting: Build simple dashboards for the team—don’t build a report nobody reads.
What works:
Automated alerts for sudden drops are genuinely useful. Playbooks for “Red” accounts can help CSMs focus. Don’t try to automate away all human judgment, though.
What doesn’t:
Don’t flood your team with alerts for every tiny change—alert fatigue is real.
Step 7: Communicate and Train Your Team
No matter how slick your setup, scores are useless if nobody trusts or understands them.
- Walk your team through what the health score means and how to use it.
- Be honest about what it does not capture (e.g., executive relationships, upcoming competitor launches).
- Encourage feedback—what’s confusing, what’s missing, what’s off base.
Pro tip:
Make it easy for CSMs to flag accounts where the score “feels” wrong. Sometimes, people spot things the algorithm misses.
What to Expect (and What Not to)
A good health score in Churnzero won’t predict every churn, but it can save your team hours of guessing. Don’t expect magic. You’ll still need to talk to customers and use your judgment.
What works: - Simple, actionable health scores that drive real action. - Regular reviews and tweaks. - Clear alerts for real risk.
What to ignore: - Overly complicated formulas. - Dozens of micro-segments. - Scores that make sense only to the person who built them.
Final Thoughts: Keep It Simple, Iterate Often
Don’t try to build the perfect health score in one go. Start with the basics, see what works, and tweak as you go. The whole point is to make it easier to spot risks and opportunities—if your team dreads looking at the health dashboard, you’ve missed the mark.
Set it up, use it, and keep refining. Simple beats clever every single time.