Revenue Operations intermediate 2-3 hours

How to Build a Customer Health Score System

Learn how to predict churn and identify expansion opportunities by building a data-driven health score for your Australian small-to-medium business.

James 2 February 2026

A Customer Health Score is essentially a 'weather report' for your client relationships. Instead of guessing who is happy or waiting for a cancellation email to drop into your inbox, a health score uses real data to tell you exactly which customers are thriving and which ones are at risk of leaving.

For Brisbane business owners, this is the difference between being reactive and proactive. If you can spot a 'red' customer two months before their contract ends, you have a genuine chance to save the relationship and protect your recurring revenue.

Prerequisites: What You’ll Need Before Starting

Before we dive into the spreadsheets and data, make sure you have the following ready:
  • A list of your current customers: Ideally in a CRM (like HubSpot or Salesforce) or a well-maintained Excel/Google Sheet.
  • Access to your data points: This includes things like last login date (for software), last order date (for retail/wholesale), or the date of your last strategy call (for service businesses).
  • A basic understanding of your 'Ideal Customer Profile': What does a 'perfect' customer actually do? (e.g., they order once a month, they pay on time, they attend your webinars).

---

Step 1: Define Your 'Health' Categories

You can't measure everything, or you'll end up with 'analysis paralysis.' Most successful Australian SMEs focus on four key categories of health.
  • Product/Service Usage: Are they actually using what they bought? (Frequency and depth).
  • Financial Health: Are they paying their invoices on time? Do they have an active ABN? (Checking the ABN lookup is a great pro tip for B2B—if a company is de-registered, that’s a massive red flag).
  • Relationship Quality: How often do they talk to you? Do they respond to your emails?
  • Customer Sentiment: What was their last NPS (Net Promoter Score) or CSAT (Customer Satisfaction) result?
Pro Tip from the field: Don't try to track 20 things. Start with 3-5 metrics that actually correlate with a customer leaving. In my experience, 'Last Login Date' or 'Last Purchase Date' is usually the single most predictive metric.

Step 2: Select Your Specific Metrics (The 'Signals')

Now we get granular. For each category above, pick one or two specific signals.
  • Usage Signal: "Logged in at least 3 times in the last 14 days" or "Ordered more than $500 of stock this month."
  • Financial Signal: "Zero invoices overdue by more than 7 days."
  • Engagement Signal: "Attended our last quarterly review meeting."
Screenshot Description: Imagine a simple table with columns for 'Metric Name', 'Data Source', and 'Frequency of Update'. This helps you see where the data is coming from (e.g., Xero, HubSpot, or a manual log).

Step 3: Assign Weights to Each Metric

Not all metrics are created equal. Getting a negative feedback comment is bad, but a customer stoping usage entirely is usually worse. You need to assign a percentage weight to each metric so they add up to 100%. Example Weighting:
  • Product Usage: 40%
  • Invoices Paid on Time: 20%
  • Relationship/Meeting Attendance: 25%
  • NPS Score: 15%
Total: 100% Note: This is where most people get stuck. Don't stress too much about getting the percentages perfect on day one. You can always tweak them later once you see how the scores play out in real life.

Step 4: Create Your Scoring Scale (0-10)

To make the math easy, score every metric on a scale of 0 to 10.
  • Example for Usage:
* 10 points: Used the service today. * 7 points: Used the service in the last 3 days. * 3 points: Used the service in the last 7 days. * 0 points: Hasn't used the service in over 14 days. Example for Finance (Australian Context): * 10 points: All invoices paid within terms. * 5 points: One invoice 7 days overdue. * 0 points: Any invoice 30+ days overdue (This is a 'Red Alert' territory in the current economy).

Step 5: Build the Calculation Engine

You don't need fancy software for this—a Google Sheet works perfectly fine for your first version.

For each customer, you will multiply their score by the weight.

The Formula: (Usage Score x 0.40) + (Finance Score x 0.20) + (Engagement Score x 0.25) + (NPS Score x 0.15) = Final Health Score Screenshot Description: A Google Sheet showing Customer Names in Column A, their individual scores in Columns B-E, and a final 'Health Score' in Column F using a simple SUMPRODUCT formula.

Step 6: Define Your 'Traffic Light' Thresholds

Now, give those numbers a colour. This makes it easy for your team to glance at a list and know who to call first.
  • Green (8.0 - 10): Healthy. These are your advocates. Ask them for a Google Review or a referral.
  • Amber (5.0 - 7.9): At Risk. Something is sliding. This is where your Account Manager should reach out for a 'coffee catch-up' (virtual or at a nice spot in Eagle Street Pier).
  • Red (0 - 4.9): Critical. High churn risk. Immediate intervention required from a senior manager.

Step 7: Decide on Update Frequency

Data goes stale fast. If you're a high-volume business, you might want to update this weekly. If you're a boutique consultancy, monthly is usually plenty. Honestly, the interface of most CRMs doesn't help with this naturally, so you might need to set a recurring calendar reminder for your admin or RevOps person to 'Refresh Health Scores'.

Step 8: Create an Action Plan for Each Colour

A health score is useless if you don't do anything with the information.
  • When a customer hits Red: Trigger an automated internal task for the founder to call the client.
  • When a customer hits Green: Trigger an automated email thanking them for their loyalty and offering a 'VIP' bonus or asking for feedback.

Common Mistakes to Avoid

  • Making it too complex: If it takes four hours to calculate the scores, you won't do it. Keep it simple enough that a spreadsheet can do the heavy lifting.
  • Ignoring 'Ghosting': Sometimes the lack of data is the data. If a customer hasn't opened your emails in three months, that’s a signal, even if they are still paying their bill.
  • Setting and forgetting: Your business changes. Review your weighting every six months to ensure it still reflects why customers actually leave.

Troubleshooting Common Issues

"My data is all in different places (Xero, Mailchimp, CRM)." This is the most common frustration. Use a tool like Zapier to pull key dates into a single Google Sheet, or simply have a team member spend 30 minutes once a month doing a manual export/import. It’s worth the effort. "The scores don't feel right—my best customer is showing as Amber." This usually means your weighting is off. Perhaps you're penalising them too much for a late invoice when they are actually your biggest brand advocate. Adjust the weights and see if the 'vibe' of the list matches reality. "I don't have enough data to score them." If you're just starting out, use 'Proxy Metrics'. Instead of 'Product Usage', use 'Email Open Rate' or 'Last Meeting Date'. Something is always better than nothing.

Next Steps

  • Draft your metrics: Spend 15 minutes listing the top 3 reasons customers have left you in the past year.
  • Build your MVP (Minimum Viable Product): Create a basic spreadsheet for your top 10 customers today.
  • Review with your team: Show them the scores and ask, "Does this look right to you?"

Building a health score system is one of the smartest things you can do for your business's long-term stability. It turns 'gut feel' into a repeatable system for growth.

If you need help connecting your CRM data to a dashboard or setting up these automations, we’re here to help. You can reach out to the team at Local Marketing Group here: https://lmgroup.au/contact.

Revenue OperationsCustomer RetentionCRMBusiness Strategy

Need Help With This?

Our team can help you implement this and more. Book a free consultation.

Book Free Consultation