Audience segmentation

Customers differ in how they behave, what they need, what they are worth to the business and how likely they are to leave. Treating every customer the same way is inefficient and wastes resources.

Audience segmentation for customer retention means dividing the customer base into meaningful groups by shared characteristics, especially churn behaviour and value. Communication and actions can then be tailored to each segment.

Audience analysis shows what customers who have churned have in common. With that knowledge, you can pinpoint which of today’s customers behave in the same way, before they possibly decide to cancel.

Customers who are likely to churn probably do not form a single uniform group. There may be several subgroups with different behaviours. That is why communication needs to be tailored to each subgroup.

Why segment for retention?

Resources go where they do the most good

Without segmentation, every customer gets the same communication regardless of risk or value.

With segmentation: Resources are spent where they have the greatest effect.

Communication becomes relevant

Without segmentation, customers get generic communication that feels irrelevant to them.

Example

A new customer (three weeks) gets an email: “Thank you for two years of loyalty!”
→ Irrelevant, because they have been a customer for three weeks.

With segmentation

Messages are adapted to the customer’s situation.

Retention efforts deliver a higher ROI

Focus on the right customers

High risk + high value

These customers are the most worth keeping. Invest time and resources.

High risk + low value

These customers are worth trying to keep, but at a lower cost (an automated campaign).

Low risk

These customers need minimal effort. Do not waste resources.

The same retention budget then has a greater impact.

How to go from churn analysis to segmentation

Use the insights from the churn analysis

Next step: Use this information to segment current customers.

Pinpoint current customers who show churn behaviour

Example

Segmentation: Identify current customers with the same behaviour: → 300 customers match the profile → Segment: “High churn risk”

How the subgroups differ

Different subgroups need different communication.

Example

If everyone is treated the same, the wrong message reaches the wrong group, and the work becomes inefficient.

If customers are segmented, each group gets a message that fits, and more efforts succeed.

Five dimensions to segment by

1. Churn risk (likelihood of leaving)

Based on the churn analysis: A risk score is calculated for each customer.

Segments

2. Customer value (LTV or revenue)

Based on historical and expected revenue

3. Customer lifecycle (where they are in the customer journey)

Segments

4. Behavioural segments

Based on how the customer uses the product or service

5. Reason for churn (for at-risk customers)

Based on why the customer is likely to leave

Three segmentation models

RFM segmentation (Recency, Frequency, Monetary)

A simple and powerful model

Recency (R): When was the latest activity or purchase?

• High (recently): 1-30 days ago
• Medium: 31-90 days ago
• Low (long ago): 91+ days ago

Frequency (F): How often is the customer active or buying?

• High: Daily or weekly
• Medium: Monthly
• Low: Rarely or only once

Monetary (M): How much does the customer spend or use?

• High: Top 20 per cent
• Medium: Middle 60 per cent
• Low: Bottom 20 per cent

Combine into segments

How to combine churn risk and customer value in a risk and value matrix

What the matrix looks like

Behaviour-based segmentation

Power users, casual users and minimal users

How to segment the customer base step by step

Step 1: Define segmentation criteria

Choose dimensions: Which dimensions matter most to the business?

Step 2: Collect data

Each dimension needs data

Source: CRM, analytics, transaction systems.

Step 3: Create segments

Apply the criteria to the data

Example (RFM)

For each customer, calculate a score on a five-point scale instead of the three levels above:

• R-score (1-5 based on recency)
• F-score (1-5 based on frequency)
• M-score (1-5 based on monetary)

Example (risk and value)

Step 4: Validate the segments

Are the segments meaningful?

If not: Adjust the criteria or merge segments.

Step 5: Document the segments

Create a persona for each segment

Step 6: Implement in your systems

CRM tags and fields

Each customer gets a segment tag in the CRM.

Automation

Marketing automation tools (HubSpot, Mailchimp) can trigger campaigns based on segment.

Dashboards

Visualise the segment distribution and churn per segment.

Step 7: Update the segments continuously

Segments are not static: Customers move between segments as their behaviour changes.

Automate: Recalculate the segments weekly or monthly.

What segmentation can look like at a SaaS company

A hypothetical company with 5,000 customers and 8 per cent monthly churn

Step 2: Collect data

Step 3: Create primary segments (risk × value)

At-risk VIP customers (High risk + High value): 50 customers
Action: A personal email from the CEO, offer a one-to-one strategy meeting

Valuable customers to nurture (Medium risk + High value): 150 customers
Action: Proactive contact from an account manager, quarterly check-in

At-risk standard customers (High risk + Medium value): 300 customers
Action: An automated email series with onboarding tips and a discount offer

At-risk low-value customers (High risk + Low value): 150 customers
Action: A simple automated email, minimal use of resources

Champions (Low risk + High value): 300 customers
Action: Ask for referrals and case studies, give early access to new features

Step 4: Create secondary segments (feature use)

Within “At-risk standard customers”, the customers are segmented further:

Step 5: What the outcome could look like after three months

At-risk VIP customers

40 of 50 customers saved in three months (80 per cent).
Revenue retained: 40 × SEK 1,000 MRR = SEK 40,000 a month (SEK 480,000 a year)

At-risk standard customers

180 of 300 customers saved in three months (60 per cent).
Revenue retained: 180 × SEK 500 MRR = SEK 90,000 a month (SEK 1,080,000 a year)

Total impact

Churn would fall from 8 per cent to around 6.5 per cent a month (220 customers saved in three months, out of 5,000).

Tools for segmentation

CRM systems

Customer data platforms (CDP)

Marketing automation

Analytics

How to go from segmentation to action

Segmentation is the step between analysis and execution.

Process

1. Churn analysis → Identify who churns and why
2. Audience segmentation → Divide customers by risk, value and behaviour
3. Channel matrix → Plan which communication each segment gets and in which channel
4. Carry it out → Send campaigns and make personal contact
5. Measure → How is retention performing per segment?
6. Optimise → Adjust segment criteria and communication based on the results

Contact us about audience segmentation

Carnaby helps companies segment their customer bases for optimal retention, from data collection and segment definition to implementation in CRM and marketing automation systems.

Contact us →