If you work in customer success, you’ve probably heard a lot about “customer health scores.” Tracking them sounds great—who wouldn’t want a magic dashboard showing which accounts need love? But getting useful, real-world health scores in Totango (or anywhere else) isn’t magic. It’s a pile of practical steps, some trial and error, and yes, a bit of patience. This guide is for anyone who wants to actually do this—without drowning in buzzwords.
Why bother tracking customer health scores?
Before you start, be honest: are you trying to predict churn, prioritize outreach, or just check a box for management? Health scores are only useful if you set them up with your real goals in mind. If you’re just copying someone else’s formula, skip it. Done right, though, health scores help you:
- Spot at-risk customers before it’s too late
- Focus your team’s time on accounts that actually need help
- Measure if your customer success efforts are working
If that sounds good, let’s walk through how to track customer health scores using Totango.
Step 1: Get your data in order (don’t skip this)
Totango is only as good as the data you feed it. Garbage in, garbage out. Before diving into score setup, make sure you can reliably get the right data into Totango:
- Core account data: Customer name, renewal date, segment, assigned CSM, etc.
- Usage data: Are your customers actually logging in and using your product? What features?
- Engagement data: Meetings, support tickets, survey responses.
- Financial data: ARR, contract value, payment status, etc.
How to get it in:
Totango can pull data via CSV uploads, API connectors, or integrations with CRMs like Salesforce. Be realistic about what you can automate and what you’ll need to update manually. If your team can’t keep a field up-to-date, don’t base your health score on it.
Pro tip:
Don’t wait for “perfect” data. Start with what you have, but make sure it’s current and trustworthy. It’s better to track one reliable thing than ten half-baked signals.
Step 2: Decide what “healthy” actually means
Every company’s definition of customer health is a little different. Don’t just copy Totango’s default settings or some blog post you read. Have a quick gut-check meeting with your team and ask:
- What really makes a customer likely to renew?
- What are early warning signs they might churn?
- What can we actually measure with the data we have?
Some signals that often matter: - Product usage frequency (e.g., logins per week) - Adoption of key features - Support ticket volume or satisfaction - NPS or CSAT scores - Days left until renewal - Overdue invoices
Stuff that usually doesn’t matter as much as people think: - Number of emails sent to the customer - CSM “sentiment” (unless you’ve got a clear rubric) - Vanity metrics like “number of users invited” if they never log in
Pro tip:
Start with 3-5 inputs. Simpler is better. You can always add more later.
Step 3: Build your health score model in Totango
Now it’s time to get your hands dirty.
3.1. Create health score criteria
In Totango, health scores are made up of “criteria”—basically, the different inputs you picked above.
- Go to Settings > Customer Health.
- Click + Add Health Metric.
- For each criterion, set:
- Data source – e.g., “Last Login Date,” “Feature Usage,” etc.
- Thresholds – What counts as “good,” “OK,” or “bad”? Totango uses colors (green/yellow/red), but you set the numbers.
- Weight – How important is this metric compared to the others? Totango lets you set this as a percentage.
Example: - Product usage > 3 logins/week = Green (40% weight) - NPS > 7 = Green (30% weight) - No open support tickets = Green (30% weight)
Honest take:
Don’t obsess over the weights. Get them roughly right, then see what happens. You’ll tweak them later.
3.2. Combine criteria into an overall health score
Totango automatically rolls up the individual metrics into an overall health score for each customer. You can customize how strict or lenient this is (e.g., does one red make the whole account red?).
- Decide if you want an average, weighted average, or something stricter.
- Test it with a few real customers—do the scores match your gut feel?
- If not, adjust thresholds or weights.
Pro tip:
Don’t hide the formula. Make sure everyone on your team knows what goes into the score and why.
Step 4: Test, tweak, and get buy-in
Here’s where a lot of health score projects go off the rails. If the scores don’t feel right to the people using them, they’ll get ignored. So:
- Spot-check real accounts. Does the health score match what your CSMs think? If not, figure out why.
- Show the formula to the team. Invite feedback, but don’t let it become endless debate.
- Tweak thresholds and weights. It’s normal to adjust a few times before things settle in.
What to ignore:
Don’t try to make everyone happy. You want a score that’s directionally useful, not perfect.
Step 5: Put health scores to actual use
A health score isn’t magic—it’s just a tool. Here’s how to make it matter:
- Trigger alerts or tasks: Set up Totango to notify CSMs when an account turns yellow or red, or when a key metric drops.
- Drive workflows: Use health scores to prioritize calls, outreach, or QBRs.
- Measure over time: Track how health scores change after interventions. If you reach out to a “red” account and it turns green, that’s a win.
- Report to leadership: Keep it simple. “X% of accounts are green, Y% are at risk” is more useful than a 10-slide deck.
Honest take:
If your team isn’t acting on the health scores, revisit your setup. Either the formula is off, or you’re tracking the wrong things.
Step 6: Avoid common traps
- Overcomplicating things: More inputs don’t always mean more accuracy. Too many signals can muddy the waters.
- Ignoring data quality: If a data feed breaks, your scores are worthless. Set a calendar reminder to sanity-check your data every month.
- Treating the score as gospel: Customer health is a guide, not a guarantee. There are always exceptions.
- Never updating: Your product and customers will change. Review your health score model every quarter or so.
Step 7: Keep it simple and iterate
Getting health scores right is a process, not a one-and-done project. The real value comes from using them, seeing what works, and making small improvements. Don’t chase some mythical “perfect” model—just ship something useful, and get feedback from the people actually talking to customers.
Bottom line:
Start simple. Track what matters. Adjust as you go. And don’t let the perfect be the enemy of the useful.