News & Insights | Baesman

Loyalty segmentation: How to use purchase data to build smarter segments

Written by Giulia Panatta | Nov 17, 2026, 6:45:00 PM

Loyalty segmentation turns purchase history into decisions: who gets which offer, when and through which channel. It's where loyalty marketing gets sharper, because you stop treating members as one list and start grouping them by what they actually buy. This guide covers the data to pull, the segment rules to start with and how to use each segment across mail, email and SMS.

TL;DR: How do you segment loyalty audiences with purchase data?

Group loyalty members by what they've already done, then give each group one clear job.

  • Start with the transaction data you already own: recency, frequency, spend, categories, promo use and redemptions.
  • Build segments around how often each member buys and whether that's changing, not only tiers or totals.
  • Give every segment a goal, a trigger and a channel plan before it goes live.
  • Suppress as carefully as you target. Reminding someone to buy what they already bought costs you subscribers.
  • Measure against holdout groups, so you know which segments respond to marketing and which were going to buy anyway.

New to the basics? Start with What is RFM? A simple guide to recency, frequency and monetary value and why first-party data is what makes customer segmentation work.

What is loyalty segmentation?

Loyalty segmentation means grouping loyalty members by how they behave (what they buy, when and how often) so each group gets offers and messages that fit. It's a narrower and more useful kind of customer segmentation, because it runs on transaction data instead of demographics or guesswork.

Tiers are not segments. A tier shows what a member has earned. A segment shows what that member is likely to do next and what you should do about it. Two Gold members can look identical on paper while one buys more every month and the other hasn't been back since spring.

That gap matters because membership isn't the same as loyalty. In AnswerLab's March 2026 study of 1,500 U.S. consumers, only 24% said they feel loyal to almost all or all of the loyalty programs they belong to. Segmentation is how you find the members who are actually committed and the ones who still need a reason to be.

What purchase data do you need to segment loyalty members?

To segment loyalty members well, you need clean purchase history tied to one member ID, plus a few signals that show how each member uses the program. You don't need a perfect data warehouse to start. You need these fields, matched to the right person:

  • Recency: days since the last purchase.
  • Frequency: number of orders in a set window, usually 12 months.
  • Monetary value: total and average spend in that window.
  • Purchase cadence: the usual gap between a member's orders. This is the most underused field in loyalty data.
  • Category mix: which product lines each member buys, and how much of their spend goes to one of them.
  • Channel: store, online, app or a mix.
  • Promotion use: the share of orders that used a discount or coupon.
  • Program behavior: points balance, redemptions, tier status and email or SMS engagement.

The first three are the RFM basics. The rest is the customer behavior data that turns a scorecard into something you can act on. If member records are split across POS, ecommerce and email systems, fix the match before you build anything. A segment built on half of a customer's purchases will misfire.

How do you segment customers using purchase history?

To segment customers using purchase history, start with the decisions you need to make, write rules from transaction fields and test those rules against what actually happened. Here's the order we follow:

  1. Name the decisions first. Who gets the spring catalog? Who gets a win-back offer? Who should never see 20% off? Each segment should answer one of these questions.
  2. Set a lookback window. Twelve months suits most retailers. Go shorter for high-frequency categories like grocery or quick-service, and longer for furniture or other big-ticket purchases.
  3. Calculate each member's cadence. Find the average number of days between purchases. A member who usually buys every 30 days and hasn't bought in 50 is at risk. A member who buys every 120 days and hasn't bought in 50 is right on schedule.
  4. Write each rule in plain language. For example: "In the top 20% of spend over 12 months, bought in the last 60 days." If your team can't explain a segment in one sentence, it's too complicated to use.
  5. Size every segment. If a segment is too small to fund a campaign, merge it. If it's too big to say anything specific to, split it.
  6. Back-test against last year. Apply the rules to historical data and check whether the segments behaved differently afterward. If "slipping" members came back at the same rate as everyone else, the rule isn't finding anything real.
  7. Refresh on a schedule. Weekly or monthly is typical. Members move, and a segment is only a snapshot.

Which loyalty segments should you start with?

Start with a few segments that each lead to a different action: getting the second purchase, growing your best members, catching members who are slipping and winning back lapsed ones. The rules below are starting points. Adjust them to your own cadence and margins.

Segment Example rule Goal Activation
New members Joined in the last 60 days, one purchase Get the second purchase Welcome series that recommends a next item based on the first purchase, plus a points bonus that expires in 30 days
Rising regulars 2+ purchases in 90 days, buying more often than the prior 90 Build the habit Reminders of progress toward the next tier, early access to new products
Best customers Top 20% of 12-month spend, purchased in the last 60 days Recognize them and protect margin Member events, early access, personalized mail. Keep them off deep-discount sends
Slipping best customers Was a best customer, no purchase in 1.5 times their usual cadence Catch them before they lapse Personalized direct mail featuring the categories they buy, then an email follow-up
Single-category shoppers 80% or more of spend in one category Grow the basket An introduction to the category that similar members buy next
Points-rich non-redeemers Balance above the reward threshold, no redemption in 6 months Make the program feel real Show the specific reward they can claim today
Discount-dependent 70% or more of orders used a promo Protect margin Test perks like free shipping or early access against discounts
Lapsed No purchase in 12+ months Win them back or let them go One or two win-back touches, then suppress from paid channels

Segments will overlap, so set a priority order and give each member one main message per cycle. A slipping best customer should get the save offer, not the generic new-arrivals email.

How do you turn loyalty segments into campaigns?

To turn a loyalty segment into a campaign, give it a trigger, an offer, a channel sequence and a suppression rule, so the segment's data decides what gets sent and when. Two examples show how that works.

For example, a specialty retailer's best customer usually buys every five weeks. At week eight she moves into "slipping." That triggers a personalized postcard built with variable data, showing the two categories she buys most and a member-only reward. An email follows a week later. She's left out of that week's storewide promo, so the brand doesn't teach her to wait for a sale.

For example, a home goods brand finds a large group of members who only ever buy bedding. Instead of sending them every new-arrivals email, the brand looks at what bedding buyers most often buy next (bath towels, say) and builds a short series around that one category, with bonus points for a first purchase there.

Choose the channel based on how valuable the segment is and what it would cost to lose it. Email and SMS work well for frequent, lower-stakes nudges. Direct mail earns its cost with high-value members who are starting to drift, because it reaches people who've stopped opening email. Our guide to retention marketing channels covers how to match each channel to its job.

Suppression matters as much as targeting. In Attentive's 2026 personalization survey of 1,050 shoppers, 67% said they're more likely to unsubscribe if they keep getting reminders about a product after buying it. Getting it right pays off: 47% of the same shoppers named a brand remembering past interactions, such as purchases, as a top reason they keep shopping with it.

How did American Girl use customer behavior data to shape its loyalty program?

American Girl built its loyalty program on an analysis of real customer behavior instead of assumptions about what members might want. One challenge. One idea. One accountable team.

Challenge: American Girl wanted to launch a loyalty program that would make sense for its customers and for the business.

Solution: Before designing anything, we ran an in-depth analysis of customer behavior data, looking at personalization potential and financial viability together. That analysis showed which customers were most worth activating and how to reach them in ways that matched how they already engaged with the brand.

Impact: The program reached 20% of its full year-one enrollment goal in the first month.

The lesson carries over to any program. The offer mattered less than knowing which customers to build it around.

Results reflect specific client engagements; averages are noted where applicable.

What mistakes weaken loyalty segmentation?

Most segmentation problems come from building segments the team can't act on or never revisits. Watch for these:

  • Segmenting on tier alone. Tiers lag behind behavior, so a member can keep top-tier status for months after they've stopped buying.
  • Using one lapse rule for everyone. A 90-day rule flags weekly shoppers far too late and flags quarterly shoppers far too early. Use each member's own cadence instead.
  • Building more segments than you can message. If there's no unique creative or offer for a segment, it's just a report.
  • Forgetting suppression. A segment should decide who doesn't get a message as well as who does.
  • Never refreshing. Segments built once a year end up describing customers who've already moved on.
  • Skipping holdouts. Without a control group, you'll give the campaign credit for purchases that would have happened anyway.

How do you know if your loyalty segments are working?

Your loyalty segments are working when members in a segment change their behavior compared with a holdout group that didn't get the campaign, and when more members move into valuable segments than out of them. Track these:

  • Incremental lift: the difference in purchase rate and spend between the targeted group and its holdout.
  • Segment migration: the share of members moving up or down between segments each month. We cover this in tracking customer movement for loyalty analysis.
  • Time to second purchase: how quickly new members come back after their first order.
  • Save rate: the share of slipping members who buy again within one normal cadence.
  • Margin per member: especially for discount-dependent members, where a higher response rate can still lose money.

Final takeaway: Segment by what members do, then give every segment a job

Your purchase data already shows which members are growing, which are drifting and which are only there for a discount. Loyalty segmentation turns that into decisions: who gets which offer, through which channel and who gets left out.

Start with five to eight segments, write each one in plain language, base the timing on each member's own cadence and measure against a holdout. Strong customer loyalty programs aren't built on a perfect data model. They're built by acting on clear segments every month and adjusting as members move. If you want a partner to connect your customer data and analytics to the campaigns that act on it, see how we approach loyalty program design and management, or let's talk.

Frequently asked questions about loyalty segmentation

What is customer segmentation in marketing?

Customer segmentation in marketing is dividing customers into groups with shared traits or behaviors so each group gets messages and offers that fit them. Loyalty programs usually segment on purchase behavior because it predicts what members will do next better than demographics do.

What are the main types of customer segmentation?

The four common types are demographic, geographic, psychographic and behavioral segmentation. Loyalty marketing relies mostly on behavioral segmentation, which groups customers by what they buy, how often and how they engage.

What is RFM analysis?

RFM analysis scores customers on recency (how recently they bought), frequency (how often) and monetary value (how much they spend). It's the most common starting point for segmenting loyalty members by purchase history.

How many segments should a loyalty program have?

Most programs do well with five to eight active segments, each tied to a distinct action. Add more only when you have the creative, offers and budget to treat each one differently.

How often should you update loyalty segments?

Update loyalty segments weekly or monthly, depending on how often your customers buy. High-frequency categories like grocery need more frequent refreshes than big-ticket categories.

What customer behavior data matters most besides purchases?

Redemption activity, points balance, email and SMS engagement, and channel preference matter most after purchase data. Together they show whether a member is actively using the program or only enrolled in it.