What Is Behavioral Segmentation? Variables, Types and Brand Examples

Segments built on age, income or location lump together customers who behave very differently. Behavioral segmentation groups consumers by what they do (what they buy, how often they use a product, when they engage and how loyal they stay) rather than by those static traits.

It matters because what consumers have already done predicts what they will buy next far better than their age or income does. This guide shows consumer insights and marketing leaders how to build a first behavioral segment, with brand examples from Netflix to Nike along the way.

What is behavioral segmentation?

Behavioral segmentation is one of the four main types of segmentation, alongside demographic, geographic and psychographic, and it is covered in our guide to types of market segmentation.

The other three types of segmentation describe who people are. Behavioral segmentation describes what they actually did. Within it, marketers usually work with a handful of segment types:

• Purchase behavior
• Occasion and timing
• Benefits sought
• Usage rate and user status
• Loyalty status
• Engagement

How does behavioral segmentation work?

Behavioral segmentation runs in three stages. A team collects records of what customers did, groups customers whose actions look alike, and then targets each group with a different offer, message or timing.

The process breaks down like this:

1. Collect action data: Teams pull transactions, product usage, site and app events, and responses to past campaigns into one customer record.
2. Group into segments: Teams set rules (two or more purchases in 90 days, no app session in 30 days) or score customers on how recently, how often and how much they buy (a model called RFM, covered below), then assign each customer to a segment.
3. Activate: Teams send each segment a different campaign, offer or experience, and measure whether behavior changed.

Netflix runs the same three steps. It records what each member watched, how they rated titles and how long they stayed with each one. It groups members whose viewing patterns look alike, then uses those groups to decide which titles to recommend to whom and when.

None of this works without a record of what customers actually did, and that record almost always comes from a brand’s own systems rather than bought-in data. Most teams pull it together from four places:

• Transaction and point-of-sale data: What was bought, when, at what price and through which channel.
• Web and app analytics: Sessions, searches, product views, cart activity and feature use.
• CRM records: Contact history, service tickets, returns and email engagement.
• Loyalty program data: Tier, points balance, redemptions and member-only activity.

What are the common behavioral segmentation variables?

A behavioral variable is the specific action a team sorts customers by. Picking the right one is most of the work: a segment built on purchase frequency answers a different question from one built on the occasion people buy for, and a team usually needs only one or two per decision. These are the seven variables teams use most, with a plain example of each.

• Purchase behavior: What, how much and how a customer buys. An apparel brand separates shoppers who buy only at full price from shoppers who buy only during markdowns.
• Occasion and timing: When a purchase or use happens, from a regular weekly shop to a gift or holiday. A coffee chain treats morning-commute buyers differently from weekend afternoon buyers.
• Benefits sought: The payoff a customer is after, such as quality, speed, economy or convenience. A meal kit separates customers who pick the fastest recipes from those who pick premium ingredients.
• Usage rate: Light, medium and heavy users. A streaming service treats subscribers who watch daily differently from those who log in twice a month.
• User status: Nonuser, ex-user, first-time user or regular user. A fitness app runs one onboarding path for first-week users and another for lapsed members.
• Loyalty status: How strongly a customer sticks with a brand, from switchers to fully loyal buyers. A beauty retailer gives its most loyal members earlier access to launches.
• Engagement: Opens, clicks, sessions and app activity short of a purchase. The AMA’s marketing frameworks guide lists “purchase history, browsing habits, and engagement levels” as ways to segment by behavior.

RFM analysis

Most teams’ first behavioral segment combines three of the variables above into one score, and the standard way to do that is RFM analysis. RFM stands for recency, frequency and monetary value: how recently a customer bought, how often they buy and how much they spend.

Each customer gets three scores from 1 to 5: one for how recently they bought, one for how often they buy and one for how much they spend. A 5 is the best score. Teams rank customers on each dimension and split them into five equal groups, so each score covers a fifth of customers. Most tools collapse the resulting score combinations into ten or so named groups. The main ones are:

Vendors disagree on which score combination belongs to which name, so a team adopting RFM should fix its own definitions before reporting segment sizes over time.

How does behavioral segmentation compare with demographic, geographic & psychographic segmentation?

Behavioral segmentation tracks actions that change with every transaction or session, while the other three types of segmentation track traits that change slowly or rarely. Our guide to psychographic segmentation covers the values-and-attitudes side in depth. The four types side by side:

The AMA’s deep dive on behavioral segmentation puts the weakness of trait-based segments plainly: demographic and lifestyle segments give fairly crude insight, because inside any one of them some consumers love a brand and others never touch it, and behavior is what separates those two groups. Traits still matter for reaching people who have not yet bought, where no behavioral record exists.

If the question is who to reach and where, start with demographic and geographic data. If the question is what to offer a known customer and when, start with behavior. If the question is which message will move them, add psychographic research.

What types of behavioral segments do brands use?

Brands rarely segment on one variable alone. The examples below show how real companies in streaming, fintech, fitness, travel, retail and sportswear combine variables, and what each one does with the resulting segment.

Usage and engagement segments: Netflix

Streaming services segment on viewing behavior more finely than almost any other category. Netflix says its recommendations draw on a member’s viewing history, how they rated other titles, the behavior of members with similar tastes, and title information such as genre and cast. It also weighs the time of day a member watches, preferred languages, the devices they use and how long they stayed with a title.

Netflix has published what that targeting is worth. A July 2026 Netflix Technology Blog post reports that, compared with its current system, random recommendations would cut engagement by 16 percent and popularity-based recommendations by 12 percent.

Transaction-pattern segments: Chime

Fintech apps sit on continuous transaction data, which makes deposit and spending patterns natural segmentation variables. Chime’s SpotMe overdraft feature requires $200 or more in qualifying direct deposits each month. Eligible members start with a $20 limit and may later qualify for $200 or more, based on account history, direct deposit frequency and amount, spending activity and other risk factors.

Chime’s membership tiers work the same way. Chime Prime requires $3,000 or more in qualifying direct deposits in the past 34 days, so a member’s tier follows their deposit behavior month to month.

Activity-based segments: Peloton

Fitness brands segment on workout behavior: frequency, streaks, class types and performance. In fiscal 2026 Peloton launched Peloton IQ, which includes personalized plans, weekly insights and personalized recommendations. Its Q2 fiscal 2026 results report that nearly half of active members had engaged with their personalized insights and recommendations by the end of the quarter.

Booking-frequency tiers: Booking.com

Travel loyalty programs translate frequency into tiers. Booking.com’s Genius program has three levels: Level 2 requires five bookings within two years, and Level 3 requires 15. Level 3 brings 10 to 20 percent discounts on select stays and priority support. Trade publication Hospitality.today, citing company disclosures in March 2026, reported that Level 2 and 3 members make up over 30 percent of active users but generate more than half of total room nights.

Spend-based loyalty tiers: Sephora and Starbucks

Retail loyalty programs often tier on annual spend, which combines purchase behavior and loyalty status in one number. Modern Retail reports that Sephora’s Beauty Insider tiers run from Insider ($0 to $349 a year) to VIB ($350 to $999) to Rouge ($1,000 or more), with fewer than 10 percent of members in Rouge.

Starbucks reintroduced tiers in March 2026 with Green, Gold and Reserve levels, where Gold starts at 500 stars earned over 12 months. The company’s program announcement says Rewards members drove nearly 60 percent of U.S. company-operated revenue in fiscal 2025, more than $13 billion in spend.

Nike’s behavioral segmentation

Nike segments on membership, purchase patterns and activity data across its apps. It has not published internal segment names, so what follows covers only the programs Nike and the press have described.

The first split is member versus guest. When Nike launched its NikePlus membership (now Nike Membership), its press release promised members first access to products “matched to their individual preferences and buying patterns.” In a 2019 report, CNBC said Nike’s members spent 40 percent more than guest customers on average. On its fiscal 2021 fourth-quarter earnings call, Nike said it had more than 300 million members and that buying member growth was outpacing new member growth.

Within membership, Nike’s programs target behavior in three ways:

• Purchase and engagement: On its fiscal 2021 third-quarter call, Nike reported buying members up 80 percent year over year and monthly engaged users up more than 60 percent, led by the SNKRS app.
• Activity: Nike’s November 2024 release describes the Nike Run Club app’s member features: localized tips before a run, six training plans and about 300 audio-guided runs.
• Access: On SNKRS, Nike’s Exclusive Access page says traditional Exclusive Access “caters towards a limited number of community members,” while invite-only Exclusive Draws open coveted launches to more members.

Nike’s recent results show the limits of a member-first strategy. Its fiscal 2026 results report a 12 percent decrease in NIKE Brand Digital revenue, and CEO Elliott Hill told CNBC in 2025 that the Covid-era shift toward digital commerce “ended up hurting the brand” because some consumers want to shop across channels. Behavioral targeting inside an app works best as one part of a channel strategy.

What are the benefits for retention, personalization & ROI?

Teams gain most from behavioral segments where revenue depends on repeat behavior.

Retention is the oldest argument. In Loyalty Rules! (2001), Bain’s Frederick Reichheld wrote that a 5 percent increase in customer retention increases profits by 25 to 95 percent. The range is wide and varies by industry, but the direction has held: teams that identify at-risk customers early can protect the most valuable revenue.

Marketers use behavioral segments to personalize campaigns and improve their performance. McKinsey’s explainer on personalization reports that faster-growing companies derive 40 percent more of their revenue from personalization than slower-growing peers. It also says personalization leaders drive 5 to 15 percent revenue increases and 10 to 30 percent gains in marketing-spend efficiency, “predominantly by deploying product recommendations and triggered communications.” Those triggered messages (an abandoned-cart email, a win-back offer after 60 quiet days) are behavioral segments in action.

Lifetime value follows from the same mechanism. The Starbucks and Booking.com figures above show how a minority of high-frequency customers carries a majority of revenue, and tiering keeps that minority visible and rewarded.

If a segment changes no offer, message, price or send time, leave it out of the activation plan.

How do teams build a first behavioral segment?

A first behavioral segment should answer one business decision with one or two variables. Teams that start with a dozen variables usually end up with segments nobody acts on.

The steps:

1. Name the decision. State what the segment will change, for example how to win back first-time buyers who have not made a second purchase within 90 days.
2. Choose one or two variables. Pick the behaviors closest to the decision. For the second-purchase problem, user status and recency are enough.
3. Assemble the data. Join transactions, site and app events, CRM records and loyalty data on a single customer ID.
4. Set thresholds or scores. Write explicit rules (one purchase, last order 90 to 180 days ago) or apply RFM scoring, and document the definitions.
5. Check size and reachability. Remove customers who have unsubscribed or opted out, then confirm the segment meets platform minimums. Google Ads requires 100 active users in the past 30 days before a list will serve on Display, Search or YouTube.
6. Activate one campaign. Send one offer or message through one channel so the result is readable.
7. Hold out a control group. Keep a share of the segment out of the campaign, as with Braze’s global control group, and measure the difference in behavior between the two groups.
8. Set a refresh schedule. Re-run the segment rules on a fixed cadence so customers move between segments as their behavior changes.

Where behavioral data stops

Behavioral data shows what people did, and that creates four practical limits:

• Missing motivation: A spike in cancellations shows that customers left, not which experiences pushed them out. Nielsen Norman Group notes that analytics cannot reveal each user’s goals or expectations, or the detail of their thoughts and feelings.
• Sparse data for new consumers: A first-time buyer has one transaction and no pattern. Teams fill the gap with product attributes until history builds up.
• Drift: Behavior shifts with seasons, prices and life events. Bain advises against treating segments as carved in stone, because individual customers move from one segment to another.
• Over-targeting: A 2025 Gartner survey found personalized marketing created negative experiences for 53 percent of customers, who were 3.2 times more likely to regret a purchase and 44 percent less likely to buy again.

The first limit is the one qualitative research addresses directly: a lapsed-buyer segment tells a team who to talk to, and a conversation with those buyers tells the team why they stopped and what would bring them back. In an AI-moderated interview, an AI moderator talks with a real participant, voice to voice, and follows up on what that person says. On Strella, teams run these conversations as AI-moderated interviews with consumers from segments such as lapsed or high-frequency buyers, and every finding links back to each participant’s own words so a team can check the verbatim behind a theme before changing a campaign.

If a segment is about to drive a pricing, offer or retention decision, interview a sample of its members before scaling the campaign. To hear how the moderator follows up on your answers, try an interview.

Related concepts

Behavioral segmentation connects to several other research and segmentation topics:

• Psychographic segmentation and how to research values and attitudes
• The types of market segmentation and how to choose between them
• What AI-moderated interviews are and how they work
• What 50 participants said about talking to an AI moderator

FAQ

Which variables count as behavioral segmentation variables?

Any variable that records an observed action toward a product or brand counts. The standard list covers purchase behavior, occasion and timing, benefits sought, usage rate, user status, loyalty status and engagement. Benefits sought sits at the edge of the category, since teams often infer it from behavior but sometimes ask for it directly.

What are the four main types of market segmentation?

The four main types are demographic, geographic, psychographic and behavioral. B2B marketers add firmographic segmentation, which the AMA describes as grouping companies by factors such as revenue, industry, location or number of employees.

How does behavioral segmentation differ from demographic, geographic and psychographic segmentation?

The other three types of segmentation group people by who they are, where they live and what they value, so their values change slowly, while behavioral values change with every purchase or session. Behavioral segments therefore suit retention, offers and timing, and trait-based segments suit reaching people who have not bought yet.

What data is needed for behavioral segmentation?

The minimum is a customer ID, a timestamp and a record of each action, usually transactions. Most programs add site and app events, CRM records and loyalty program data, joined on the same customer ID.

Does behavioral segmentation replace customer research?

Behavioral segmentation identifies which customers act alike, and customer research explains why they act that way. Teams get the most from segments when they interview members of a segment before deciding what to change for it.

Have a segment whose behavior needs explaining? Bring your research question to us and we will talk it through. See Strella in action