LiftSignals is live. Every plan starts with a 30-day free trial — no credit card required. Start free →
Segmentation & RFM

Customer segmentation for ecommerce: types, examples and how to start

How to split customers into groups that behave differently, so each group gets the message, offer and budget that fits.

By LiftSignals editorial team · Published · 4 min read

Small clay figures sorted into four groups on a pale board, one group orange

Key takeaways

  • Customer segmentation groups customers who share traits, so you can treat each group differently.
  • For online stores, behavioral segments built from orders predict future buying better than age or location.
  • RFM (recency, frequency, monetary value) is the simplest behavioral segmentation to start with.
  • Segments only pay off when they change what you send, to whom, and when.

Customer segmentation is the practice of grouping customers who share a trait so you can treat each group differently: a different message, a different offer, a different ad budget. For an online store, the most useful segments are built from how people buy, not who they are. This guide covers the main types of segmentation, why behavioral segmentation suits ecommerce, a simple way to start with RFM, and eight segment examples you can build this week.

What is customer segmentation?

Instead of sending every customer the same email, segmentation lets you send your best customers early access, your new customers a second-order nudge and your lapsing customers a reminder. Each group gets what it's most likely to respond to.

The four types of customer segmentation

Type Based on Example
Demographic Age, gender, income Women 25–34
Geographic Country, region, ZIP code Customers in the Pacific Northwest
Psychographic Values, interests, lifestyle Eco-conscious shoppers
Behavioral Purchases, timing, spend Bought three times in the last 90 days

All four have uses, but they aren't equally predictive. Two customers of the same age in the same city can be your most loyal buyer and someone who ordered once on a discount and never returned.

Why behavioral segmentation works best for online stores

Behavioral segmentation uses what customers actually do: when they bought, how often, how much and what. Online stores have an advantage here, because every order is already recorded.

The best predictor of what a customer will do next is what they've already done.

Behavioral segments also update themselves. A customer who was “active” in March and hasn't ordered since will drift into a “lapsing” segment on their own if the segment is defined by recency. Predicted scores take the same idea a step further, and the predictive analytics for ecommerce guide explains where they help and where they mislead.

Start with RFM segmentation

RFM is the simplest behavioral segmentation and a good first step. It scores every customer from 1 to 5 on three measures:

  • Recency: how recently they last ordered
  • Frequency: how many orders they've placed
  • Monetary value: how much they've spent

A customer who ordered 12 days ago, has placed six orders and spent $640 might score 5-4-4. The RFM analysis tool page shows how to calculate scores in a spreadsheet.

Eight customer segmentation examples for ecommerce

  1. Champions. Recent, frequent, high spend. Thank them; don't discount them.
  2. New customers. One order in the last 30 days. Time a second-order nudge to your store's usual reorder gap.
  3. Rising customers. Order frequency or spend increasing. Invite them into VIP treatment early.
  4. At risk. Used to buy often, now quiet past their usual gap. Reach out personally.
  5. Can't lose. Your biggest past spenders who have gone quiet. Highest-priority win-back.
  6. Deal-driven. Mostly buy on promotion. The right home for discount codes. The discount addiction guide explains why codes belong here and not with everyone.
  7. One-and-done. One order, long ago. Cheap reminders only; exclude from paid retargeting.
  8. High return risk. Return a large share of orders. Leave them out of free-shipping offers.

In LiftSignals

LiftSignals rebuilds RFM segments every night, adds predicted value and churn risk, and syncs any segment to Klaviyo, Meta, Google and eight other tools, with a holdout on each so you can see which segments' campaigns actually worked.

How to make segments pay off

A segment is only useful if it changes an action. For each segment, write down three things before you build it: what you'll send, how often, and how you'll know it worked. If you can't answer the last one, run the first campaign with a holdout group that doesn't receive it.

Customer segmentation questions

What are the four main types of customer segmentation?

Demographic (who they are), geographic (where they are), psychographic (what they value) and behavioral (how they buy). Ecommerce stores usually get the most from behavioral segmentation, because order history is data they already have.

How many customer segments should a small online store have?

Start with four to eight. Enough to treat your best, newest, drifting and lapsed customers differently, but few enough that each one gets a distinct message you actually maintain.

What's the difference between a segment and a list?

A list is a fixed set of contacts. A segment is defined by rules, so customers move in and out automatically as their behavior changes.

Can customers belong to more than one segment?

Yes, and they usually do. Set a priority order or frequency caps so overlapping segments don't send the same customer several messages at once.

Put it to work: RFM analysis tool → · Customer segmentation →