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

Platform · Customer segmentation

The customer segmentation tool that sorts every customer by what they’ll do next.

Stop sending the same email to your best customer and someone who bought once two years ago. LiftSignals builds RFM and behavioral segments from your order history every night, ranks them by value and risk, and hands each one to the channel that should reach it.

No credit card · Read-only access · Works with Shopify, WooCommerce, BigCommerce, Amazon, Square or a CSV

A wooden tray divided into nine compartments of sorted gray pebbles, with one compartment of orange pebbles

Definition

Customer segmentation, defined for online stores.

Customer segmentation is grouping customers by shared traits so each group gets a different message, offer or budget. The most useful segments in ecommerce are behavioral: they group people by how they buy, not by age or ZIP code. RFM segmentation is the classic version.

THE RFM MODEL

Recency × Frequency × Monetary value, each scored 1 to 5

Example: a 5-5-5 bought last week, buys often and spends the most. A 1-1-1 bought once, long ago, and spent little. 125 combinations, usually grouped into about ten named segments.

Manual segmentsLiftSignals segmentation
Built fromFilters someone set up onceEvery order, scored nightly
BasisDemographics and guessworkBehavior: recency, frequency, spend
Forward-lookingNoAdds predicted value and churn risk
UpkeepGoes stale within weeksCustomers move between segments automatically

Why one list fails

When everyone gets the same message, everyone pays for it.

01

Your list fatigues the people who matter.

Blasting the whole list means your best customers get the most emails, and they’re the ones most likely to unsubscribe from noise.

02

Segments go stale the week you build them.

A customer who was “active” in March is lapsing by June. Static lists keep treating them like nothing changed.

03

Demographics don’t predict purchases.

Two 34-year-olds in Austin can be your best customer and your least profitable one. Behavior tells them apart; age doesn’t.

How LiftSignals does it

RFM segmentation plus prediction, rebuilt every night.

RFM scores

Every customer gets recency, frequency and monetary scores from 1 to 5, recalculated nightly from your order history.

Named segments

Scores roll up into plain-English segments such as Champions, Loyal, At risk, Can’t lose and Hibernating, so your team knows what to do with each.

Combined Insights

A grid that crosses predicted value with churn risk, so “high value, high risk” is one click, not a spreadsheet.

Contact rules

Set how often any customer can be contacted across segments, so one person in three segments doesn’t get three emails in a day.

Sync

Any segment can sync to Klaviyo, Attentive, Meta or Google and refreshes nightly.

COMBINED INSIGHTS · VALUE × RISK

LOW RISKMEDIUMHIGH RISKHIGH VALUE
610
180
140
MID VALUE
1,320
940
610
LOW VALUE
1,750
1,190
1,660

140 customers are worth the most and the most likely to leave. That’s this week’s win-back list.

Illustrative read-out for a store with 8,400 customers. Your grid is computed from your own order history.

What to do with it

Four segments worth acting on this week.

01

Thank your Champions, don’t discount them

KlaviyoShopify

Your 5-5-5 customers get early access and a note from the founder, not 20% off. Save the coupon for people who need a reason.

Customer lifetime value →

02

Win back “Can’t lose” before it’s too late

KlaviyoAttentive

Big spenders who used to buy often and have gone quiet get a personal win-back sequence while they still remember you.

Churn prediction →

03

Turn one-time buyers into second orders

Klaviyo

New customers with a single order get a second-purchase flow timed to your store’s typical gap between first and second orders.

RFM analysis tool (coming soon) →

04

Stop paying to reach Hibernating customers

MetaGoogle Ads

Exclude long-lapsed, low-value customers from retargeting and give that budget to lookalikes of your Champions.

Integrations →

Proof on every play

Then prove each segment’s campaign worked.

Every audience you send carries an automatic holdout: a random slice of those customers who don’t get the message. 30 days later you see what the campaign caused, not what it claimed, in dollars, with the confidence interval printed next to it.

Two glass columns of gray sand at the same level; the right column carries an extra layer of orange sand on top

Is it right for you?

Who behavioral segmentation is for.

A good fit if you

  • Email or text customers more than a couple of times a month
  • Have at least about 500 customers and 90 days of orders
  • Want segments that update themselves
  • Use Klaviyo, Attentive, Meta or Google and want better audiences in them

Sounds like you? Start free

Not the right tool if you

  • Need firmographic B2B segmentation by company size or industry
  • Want survey-based personas rather than purchase behavior
  • Have only a few hundred customers so far
  • Need a full customer data platform with web event tracking

FAQ · customer segmentation

Customer segmentation questions, answered.

Want the method? Read the guide to RFM analysis and RFM segmentation.

What is RFM segmentation?

RFM segmentation scores each customer on recency (how recently they bought), frequency (how often) and monetary value (how much). Customers with similar scores are grouped into segments such as Champions, At risk and Hibernating.

What are the main types of customer segmentation?

Demographic, geographic, psychographic and behavioral. For ecommerce, behavioral segmentation, built on what people actually buy and when, is usually the most predictive of what they’ll do next.

How is this different from segments in Klaviyo?

Klaviyo segments are filters you build and maintain. LiftSignals scores every customer nightly, adds predicted value and churn risk, and syncs the result into Klaviyo as ready-made lists. Many customers use both.

How often do segments update?

Every night. Customers move between segments automatically as they buy or go quiet.

Can one customer be in more than one segment?

Yes. Contact rules let you set how often any one customer can be reached across all segments, so overlap doesn’t turn into over-mailing.

How many customers do I need for RFM segments?

RFM scores work from the first sync, since they only describe past behavior. The predictive layers, value and churn risk, need about 500 customers and 90 days of orders before they appear.

Do I need a data team to use it?

No. Segments are named in plain English and each one has a suggested play. If you have analysts, they can export scores too.