How to Build a Client Health Score in Your CRM

Why You Need a Client Health Score

You probably know which clients are happy. The ones who reply quickly, pay on time, and keep coming back. But what about the quiet ones? The clients who used to order every month but have gone silent for six weeks? By the time you notice, they may already be talking to a competitor.

A client health score solves this problem. It takes the signals scattered across your CRM, such as purchase history, email engagement, meeting frequency, and payment patterns, and combines them into a single number. That number tells you, at a glance, whether a client relationship is strong, stable, or slipping.

With Q2 now underway and the new UK tax year just begun, this is the ideal moment to build one. You have a full year of client data behind you, and a fresh quarter ahead to act on what you find.

What Goes Into a Health Score

A health score is only as good as the data behind it. The metrics you choose should reflect genuine engagement, not vanity numbers. Here are six signals that work well for most small businesses.

Purchase Recency

How recently did the client last buy from you? A client who placed an order two weeks ago is in a very different position to one whose last purchase was eight months ago. Recency is often the strongest single predictor of future behaviour.

Purchase Frequency

How often does the client buy? Regular purchases indicate a healthy, embedded relationship. A sudden drop in frequency is one of the earliest warning signs of churn.

Communication Activity

Are they opening your emails? Replying to messages? Attending meetings? Track inbound and outbound communication in your CRM. Silence is rarely a good sign.

Revenue Trend

Is the client spending more, less, or the same as last year? A client whose order values are gradually declining may be shifting budget to a competitor without telling you.

Payment Behaviour

Late payments can signal financial trouble, dissatisfaction, or both. A client who always paid within 14 days but has started stretching to 45 is worth a conversation.

Support or Complaint History

Frequent complaints are an obvious red flag. But the absence of complaints is not necessarily positive. Some clients simply stop complaining and leave instead. Track the pattern, not just the count.

Choosing Your Scoring Model

There is no single correct way to score client health. The best model is the one you will actually use. Here are two approaches that work well for small businesses.

Simple Traffic Light Model

This is the fastest to implement. Score each client as Green (healthy), Amber (needs attention), or Red (at risk) based on a quick review of their recent activity.

Colour	Criteria	Action
Green	Purchased in last 30 days, responds to communication, payments on time	Maintain relationship, look for upsell
Amber	No purchase in 30 to 60 days, reduced communication, or one late payment	Schedule a check-in call within a week
Red	No purchase in 60+ days, no communication, multiple late payments, or complaints	Immediate outreach, escalate internally

This model works best if you have fewer than 50 active clients and can review them manually each month.

Weighted Points Model

For a more precise score, assign points to each metric and weight them by importance. This scales better and removes some of the subjectivity.

Metric	Weight	Scoring
Purchase recency	25%	Last 30 days = 10, 31 to 60 days = 7, 61 to 90 days = 4, 90+ days = 1
Purchase frequency	20%	Above average = 10, average = 7, below average = 4, none = 1
Communication activity	20%	Regular replies/opens = 10, occasional = 6, silent = 2
Revenue trend	15%	Growing = 10, stable = 7, declining = 3
Payment behaviour	10%	Always on time = 10, occasional late = 5, consistently late = 1
Complaints	10%	None = 10, resolved = 6, unresolved = 2

Multiply each score by its weight, then add them up. The result is a number between 1 and 10 that represents overall client health.

How to Set It Up in Your CRM

You do not need specialist software to run a health score. Most CRMs with custom fields and basic reporting can handle it. Here is a step-by-step setup.

Step 1: Create a Custom Field for the Score

Add a number field (or a dropdown for the traffic light model) to your contact or company records. Name it something clear like “Health Score” or “Client Health.” This field becomes the single point of reference for every client relationship.

Your CRM’s tags and custom fields guide covers the mechanics of setting this up if you have not done it before.

Step 2: Define Your Metrics and Thresholds

Write down the four to six metrics you will track and the scoring criteria for each. Keep this in a shared document so everyone on the team applies the same standards. Consistency matters more than precision at this stage.

Step 3: Score Your Existing Clients

Run through your active client list and assign each one a score based on the last 90 days of data. This initial pass will take time, but it gives you a baseline to work from. Sort the results from lowest to highest to see your at-risk clients immediately.

Step 4: Set Up a Monthly Review Cadence

Block time on the first Monday of each month to update scores. Pull your monthly CRM reports and use them to refresh each client’s score. Track how scores change over time so you can spot trends, not just snapshots.

Step 5: Create Alerts for Score Drops

If your CRM supports automation, set up a notification that triggers when a client’s health score drops below a threshold (for example, from Green to Amber, or below 5 on the points scale). This ensures at-risk clients get attention without relying on someone remembering to check.

Segmenting Clients by Health Score

Once every client has a score, you can segment your database by health status. This unlocks targeted actions for each group.

• Healthy clients (7 to 10): These are your advocates. Ask for testimonials, request referrals, and explore upsell opportunities. They are also the clients whose behaviour you should study to understand what a good relationship looks like.

• Attention clients (4 to 6): Something has shifted. Maybe they are buying less, responding less, or paying slower. A personal check-in often reveals an issue you can fix. Do not wait for the score to drop further.

• At-risk clients (1 to 3): These clients are on the verge of leaving, if they have not mentally left already. Prioritise direct phone calls over emails. If you learn they have already moved on, your win-back strategy becomes the next step.

Combining health scores with client lifetime value creates an even clearer picture. A high-value client with a declining health score should be your top priority. A low-value client with a strong score might be ready for an upsell conversation.

Common Mistakes to Avoid

Building a health score is straightforward, but there are a few pitfalls that trip up small businesses.

Tracking Too Many Metrics

More data does not always mean better insight. If your score relies on 15 different signals, it becomes impossible to maintain and difficult to interpret. Start with four to six metrics. You can always add more later.

Ignoring the Score Once It Is Built

A health score only works if you act on it. If you calculate scores in January and never look at them again, you have wasted your time. Build the review into your monthly routine alongside your other CRM metrics.

Relying on Gut Feel Instead of Data

The whole point of a health score is to replace “I think they are fine” with “the data says they are fine.” If you override the score based on instinct, you undermine the system. Trust the numbers. When your gut and the score disagree, investigate rather than dismiss.

Scoring Without Context

A client who has not purchased in 60 days might look at risk, but if they always buy quarterly, they are perfectly healthy. Tailor your scoring thresholds to your business cycle. A consulting firm with long project timelines needs different recency thresholds than a supplies business with weekly orders.

A Real-World Example

Here is how a fictional marketing consultancy might score three of its clients at the start of Q2 2026.

Metric (Weight)	Client A: Greenfield Ltd	Client B: Maple & Co	Client C: Hartley Group
Purchase recency (25%)	Last order: 2 weeks ago. Score: 10	Last order: 11 weeks ago. Score: 1	Last order: 5 weeks ago. Score: 7
Purchase frequency (20%)	6 projects in 12 months. Score: 10	2 projects in 12 months. Score: 4	4 projects in 12 months. Score: 7
Communication (20%)	Replies same day. Score: 10	No reply to last 3 emails. Score: 2	Opens emails, replies within a week. Score: 8
Revenue trend (15%)	Up 20% year on year. Score: 10	Down 40%. Score: 3	Flat. Score: 7
Payment (10%)	Always within terms. Score: 10	2 late invoices. Score: 5	On time. Score: 10
Complaints (10%)	None. Score: 10	1 unresolved. Score: 2	None. Score: 10
Weighted total	10.0	2.6	7.6
Status	Healthy	At Risk	Healthy

Client B needs a phone call today, not next week. The data is clear: declining revenue, silence on emails, an unresolved complaint, and payments slipping. Without a health score, this pattern might go unnoticed until the client formally terminates.

According to Harvard Business Review ↗, acquiring a new client costs five to 25 times more than retaining an existing one. A health score gives you the chance to retain before you need to replace.

Linking Health Scores to Revenue

Health scores are not just about retention. They connect directly to revenue planning.

Bain & Company ↗ found that a 5% improvement in retention can increase profits by 25% to 95%. If your health score helps you save even two or three clients per year that would otherwise have left, the revenue impact is significant.

Track the correlation between score changes and actual revenue outcomes. After six months, you should be able to answer questions like:

• What percentage of clients who dropped below 4 eventually churned?
• What is the average revenue of clients in each health tier?
• Did clients who received a check-in after a score drop stay longer than those who did not?

This turns your health score from a monitoring tool into a predictive one. Gainsight’s guide to health scores ↗ covers predictive models in more depth if you want to take this further.

Start Building Your Health Score This Week

You do not need perfect data to get started. You need enough data to be useful, and if you have been using a CRM for more than a few months, you already have it.

Here is a quick-start checklist:

1. Pick four to six metrics you can track reliably.
2. Choose the traffic light or weighted points model.
3. Add a “Health Score” custom field to your CRM.
4. Score your top 20 clients this week.
5. Expand to your full client list by the end of April.
6. Set a recurring monthly review in your calendar.

The start of a new tax year is a natural reset point. Use it. Build the habit now, and by the time you reach Q3, you will have a clear, data-driven view of every client relationship in your business.

If you want to explore which numbers matter most for growth, our guide on CRM metrics that actually matter is a good next step.