Customer Segmentation

Not All Customers Are the Same

Sarah from Portland orders organic skincare products every six weeks like clockwork. She always pays with Apple Pay, never uses discount codes, and has a lifetime value of $2,400. Meanwhile, Jake from Miami bought one item during a Black Friday sale 14 months ago and has never returned. Should you spend the same marketing budget on both of them? Obviously not — but when you have 50,000 customers, identifying the Sarahs and the Jakes manually is impossible.

This is where AI-powered customer segmentation transforms your business. Instead of treating your entire customer base as one group — or manually splitting them into "new" and "returning" — AI analyzes purchase patterns, browsing behavior, and engagement signals to create precise customer clusters. Each cluster gets a tailored strategy, and your marketing dollars go where they generate the highest return.

RFM Analysis: The Foundation

The most proven framework for customer segmentation in retail is RFM — Recency, Frequency, Monetary value. It works because it is based on actual purchase behavior, not demographics or surveys.

Each customer gets a score of 1-5 on each dimension (5 = best). A customer who scores 5-5-5 is your champion. A 1-1-1 is effectively lost.

RFM in Practice

Customer: Sarah K. (Portland, OR)
Last Purchase: 8 days ago (Recency: 5)
Purchases in 12 months: 9 (Frequency: 5)
Total Spend: $2,400 (Monetary: 5)
RFM Score: 5-5-5 — Champion

Customer: Jake M. (Miami, FL)
Last Purchase: 422 days ago (Recency: 1)
Purchases in 12 months: 1 (Frequency: 1)
Total Spend: $34.99 (Monetary: 1)
RFM Score: 1-1-1 — Lost

Open data/customer-rfm.csv in the code panel. This file contains anonymized RFM data for 5,000 customers of a mid-size DTC home goods brand — customer ID, last purchase date, purchase count, total spend, and pre-calculated RFM scores. This is the starting point for AI segmentation.

AI Segments: Beyond Basic RFM

While RFM gives you scores, AI clustering goes further by identifying natural groupings in your data that you might not think to look for. Here are the segments AI typically discovers in Western retail customer bases:

What AI Adds to RFM

Traditional RFM treats all purchases equally. AI adds layers:

Segment-Specific Strategies

Champions (Your Top 8-12%)

These customers love your brand. The worst thing you can do is send them the same generic 20% off email as everyone else.

Strategy:

Do not: Discount to them. They already buy at full price. Discounting trains them to wait for sales.

Potential Loyalists (Your Growth Engine)

These customers bought recently but have not yet become regulars. They are your highest-ROI marketing target.

Strategy:

At Risk (Win-Back Before It Is Too Late)

These were once good customers. Something changed — maybe a bad experience, maybe a competitor, maybe life got in the way.

Strategy:

Open data/segment-campaigns.json to see 20 campaign templates organized by segment — email subject lines, SMS copy, offer structures, and expected conversion rates based on industry benchmarks.

Bargain Hunters (Manage, Do Not Ignore)

These customers only buy during promotions. They are not bad customers — they just respond to different incentives.

Strategy:

Building Segments with AI

The Simple Way: AI Chat

Paste your customer data into Claude or ChatGPT:

Here is my customer transaction data for the last 12 months:
[Paste: customer_id, purchase_date, order_value, items,
payment_method, state/country]

Tasks:
1. Calculate RFM scores (1-5) for each customer
2. Assign segments: Champions, Loyal, Potential Loyalists,
   At Risk, Bargain Hunters, One-and-Done
3. Show segment size (count and % of total)
4. For each segment, recommend:
   a. Best channel (email, SMS, social, paid ads)
   b. One specific campaign with subject line and offer
   c. Expected conversion rate
5. List the top 20 customers most likely to churn in the
   next 30 days

The Platform Way: Built-In Tools

Most e-commerce platforms now offer AI segmentation:

Privacy and Compliance

Customer segmentation requires handling personal data responsibly. Key regulations:

Measuring Segmentation Success

Track these metrics monthly to know if your segmentation is working:

Key Takeaways