Answer

What is customer segmentation?

The point is simple: treating every customer the same wastes budget and bores buyers. A good segmentation model turns one audience into a handful of focused audiences, each with a clearer message and a higher response rate.

Start free trial See pricing

Short answer

Customer segmentation is the practice of grouping customers and prospects by shared attributes so a company can tailor messaging, pricing, product, and service to each group. Common segment types include demographic, firmographic, behavioral, psychographic, needs-based, and lifecycle stage. Teams run segmentation inside a CRM using lists, filters, and dynamic rules, so the groups update themselves as new data lands and campaigns stay aligned to who the customer actually is.

Key points

What matters most.

The six segment types every revenue team should know, plus the two frameworks (ICP and RFM) that drive most B2B segmentation work.

Definition

Grouping customers by shared traits.

Customer segmentation turns one undifferentiated audience into a handful of groups that share something meaningful: a role, an industry, a buying pattern, a lifecycle stage. The groups are the unit the business targets with messaging, pricing, product, and service. Done well, the same budget produces more response because each group hears something that fits.

Six segment types

Demographic, firmographic, behavioral, more.

Demographic (age, income, role). Firmographic (company size, industry, revenue, geography) for B2B. Behavioral (what the customer actually does). Psychographic (values, attitudes, interests). Needs-based (the job the customer is trying to get done). Lifecycle stage (prospect, new, active, at-risk, churned). Most teams combine two or three types instead of relying on one.

ICP

Ideal customer profile for acquisition.

The ICP is a firmographic and behavioral description of the account most likely to buy, succeed, and renew. It drives acquisition: which lists to buy, which ads to run, which inbound leads to prioritize. A good ICP has three or four hard criteria (industry, size, geography, technology) plus a short list of signals (role count, growth stage, trigger events) that qualify a fit.

RFM

Recency, frequency, monetary for retention.

RFM scores existing customers on three axes: how recently they bought, how often they buy, and how much they spend. Customers who score high on all three are the base. Customers whose recency drops are the save list. Customers whose frequency climbs are the expansion list. RFM is the single simplest retention segmentation and the one most teams underuse.

Dynamic vs static

Lists that update themselves.

A static segment is a saved list of records as of a moment in time. A dynamic segment is a saved query that re-runs on every change, so the list is always current. Dynamic segments power campaigns, scoring, and routing without manual maintenance. Static lists still have a job for one-off exports and audit snapshots, but dynamic is the default for anything that runs more than once.

Where it lives

The CRM, not a separate tool.

Segmentation that lives outside the CRM goes stale fast, because the underlying data keeps changing. Running segments inside the CRM, against live contact, company, deal, and activity records, keeps the groups accurate and makes them available to every downstream workflow: campaigns, routing, scoring, forecasting, and reporting all pull from the same list.

The six segment types

How teams actually slice their audience.

The textbook lists five or six segment types. In practice, a strong segmentation model picks the two or three types that match what the business sells, then layers them. A B2B SaaS company might combine firmographic (industry and size), behavioral (product usage), and lifecycle (trial, paid, at-risk). A direct-to-consumer brand might combine demographic, behavioral (purchase history), and psychographic (lifestyle). The right mix depends on the data available and the decisions the segments need to drive.

Demographic

Who the person is.

Age, gender, income, education, occupation, family status, language. Classic consumer segmentation, still useful anywhere personal context shapes the offer. In B2B, the demographic layer narrows to role, seniority, and function, because the company context (firmographic) carries more weight than the individual demographic.

Firmographic

What kind of company it is.

The B2B equivalent of demographic. Industry, employee count, revenue band, geography, ownership structure, growth stage, technology stack. Firmographics are the first filter on an ICP and the easiest to populate from third-party enrichment, which is why most B2B segmentation starts here before layering behavior on top.

Behavioral

What the customer actually does.

Pages visited, emails opened, demos booked, products viewed, features used, support tickets filed, renewals skipped. Behavior is the most predictive segment type because it reflects real intent, not stated intent. The challenge is capturing it: behavioral segmentation only works when the activity data flows into the CRM reliably.

Psychographic

What the customer believes.

Values, attitudes, interests, lifestyle, pain points, risk tolerance. Harder to source than firmographic or behavioral because it usually comes from surveys, interviews, or inferred signals. When captured well, psychographic data drives message tone and positioning: the same product sold to a risk-averse buyer sounds different from the same product sold to an early-adopter buyer.

Needs-based

The job the customer is trying to get done.

Groups buyers by the problem they are solving, not by who they are. A finance team buying a CRM to clean up revenue reporting is a different segment than a sales team buying the same CRM to run pipeline. Needs-based segments pair naturally with case studies and feature pages, because the content maps to the job the segment hired the product to do.

Lifecycle stage

Where the customer is in the relationship.

Prospect, lead, opportunity, new customer, active, at-risk, churned, re-engaged. Lifecycle is the segmentation that governs timing: a new customer needs onboarding content, an at-risk customer needs a save play, a churned customer needs a win-back campaign. Every CRM supports lifecycle stage out of the box, and every revenue team should use it.

Two frameworks that do most of the work

ICP for acquisition, RFM for retention.

Segmentation gets abstract fast. Two frameworks keep it practical. ICP (ideal customer profile) answers the acquisition question: which accounts should we chase? RFM (recency, frequency, monetary) answers the retention question: which existing customers deserve the save, the expansion play, or the quiet renewal? Together they cover the full lifecycle and give the segmentation program a shape leadership can actually review.

ICP, defined

The account most likely to buy and succeed.

An ICP is a short, honest description of the customer who closes fast, pays reliably, uses the product well, and renews. Three to five firmographic criteria (industry, size, geography, tech) plus two or three behavioral signals (role count, growth stage, trigger events). An ICP is not the biggest logo a sales team wants, it is the account the product serves best right now.

ICP, in use

Filters acquisition at the top of the funnel.

Marketing runs campaigns against the ICP. Sales prioritizes inbound leads that match it. Operations routes high-fit accounts to the strongest reps. The point is not refusing to sell outside the ICP, it is spending the limited budget and attention on the accounts most likely to produce a result, and letting lower-fit accounts self-serve.

RFM, defined

Three axes that rank existing customers.

Recency: how recently did the customer last buy or engage. Frequency: how often do they buy or engage. Monetary: how much do they spend. Score each axis one to five, combine the three digits (RFM score 555 is the base, 111 is lost). The grid produces natural segments: champions, loyal, at-risk, hibernating, promising, lost. Every one of those segments earns a different play.

RFM, in use

The save list and the expansion list.

When recency drops on a historically high-frequency customer, something changed. Fire the save play before the renewal date, not after. When frequency climbs on a mid-monetary customer, there is an expansion opportunity sitting in the data. RFM turns the retention team from reactive (chase churned logos) into proactive (catch the drift early, grow the base quietly).

How to build a segmentation program

Data, groups, test, refine.

A segmentation program fails for one of two reasons: the team models segments it cannot populate from the data it has, or the team ships segments and never tests whether they produce different outcomes. Both failures are avoidable with a short, honest process. Start with the data on hand, group by the attributes that actually vary, test the groups against a real campaign or play, and refine the model from the results. Then do it again next quarter.

Step 1

Audit the data you already have.

Before modeling segments, inventory the fields on contacts, companies, deals, and activities. What is populated at over eighty percent? What is enriched automatically? What requires rep entry (and therefore will not scale)? The honest audit kills aspirational segments that depend on data the team does not have and will not maintain.

Step 2

Group by what varies, not by what is uniform.

A good segment attribute varies meaningfully across the audience and correlates with a decision the business can act on. Industry is a great attribute if the product serves multiple industries differently. Job title is weak if everyone in the ICP has the same title. Pick the three to five attributes that split the audience into groups worth treating differently.

Step 3

Build dynamic segments in the CRM.

Encode the segments as saved queries, not static exports. A dynamic segment stays current as records change, which means campaigns, scoring, and routing stay aligned without manual maintenance. Give the segments stable names, document the criteria, and make the list a shared asset the whole revenue team can use.

Step 4

Test the segments with a real play.

Pick one segment and run a real campaign, a real routing rule, or a real pricing test against it. Compare the result against the baseline (or against a control segment). If the segmented group outperforms, the model works. If not, the attributes are wrong, the message is wrong, or the segment is too small to measure. Fix one at a time.

Step 5

Refine on a cadence, not continuously.

Segmentation drifts. Industries consolidate. Buying committees change shape. Behavioral signals age out. Review the segment definitions quarterly against the result data: which segments converted, which flopped, which stopped mattering. Retire the dead ones, split the overloaded ones, and keep the model lean enough that someone new to the team can read it in ten minutes.

Step 6

Share the model, not just the lists.

The segmentation model is only useful if sales, marketing, service, and leadership read it the same way. Write the segment definitions down, publish them where the team works, and reference them in campaign briefs, sales plays, and QBRs. A segment that lives only in one marketer's head is a static list waiting to go stale.

How a CRM runs segmentation live

Lists, filters, dynamic rules, and AI patterns.

A segmentation model is only as good as the system that runs it. The CRM is where the segments actually earn their keep: it is where contact, company, deal, and activity data converge, which means it is the only place a dynamic segment can stay current across the full customer lifecycle. The specific capabilities below are what separate a CRM that supports segmentation from a CRM that just stores contacts.

Saved lists

A segment is a saved query.

Every segment lives as a saved view against contacts, companies, or deals. The view encodes the filter criteria, the sort order, and the columns to show. Sales teams open the view like a worksheet. Marketing teams send to the view like a list. Reports chart against the view. The view is the single source of truth for who belongs in the segment this moment.

Dynamic filters

Multi-field, multi-object criteria.

A segment definition can span fields on the contact, the company, the open deals, and the recent activities. Any contact at a company in a specific industry, with revenue in a given band, with an open deal over a threshold, who viewed the pricing page in the last fourteen days. That compound filter is what a good segment looks like, and it needs a CRM that can run it without exporting to a spreadsheet.

Lifecycle stage

Built-in lifecycle segmentation.

A lifecycle field on every contact and account, with stages the team defines. Transitions trigger automation: a new "customer" fires onboarding, an "at-risk" fires a save play, a "churned" fires a win-back cadence. Lifecycle is the segmentation layer every revenue team uses whether they call it segmentation or not.

Lead and account scoring

Scores that feed segment definitions.

Lead scores (fit plus intent) and account scores (health plus engagement) are themselves outputs of segmentation logic, and they turn around and feed new segments. The top decile of lead scores is a segment. Accounts whose health score dropped two bands in a month is a segment. Scoring and segmentation reinforce each other inside the same system.

Strkr AI patterns

Patterns the data suggests, not just the rules you wrote.

Strkr AI reads the live record data and surfaces segment candidates the team did not define: cohorts that behave similarly, accounts that look like recent winners, contacts whose activity pattern matches past churn risk. The AI patterns do not replace the explicit segmentation model, they feed it, because the honest surprises are the ones the quarterly refinement review needs to see.

Everything downstream

Campaigns, routing, forecasts, reports.

Once a segment exists as a saved query, every downstream workflow can target it: email campaigns send to it, routing rules check against it, forecasts roll up by it, dashboards chart against it. That reuse is the point of running segmentation inside the CRM instead of exporting lists into separate tools that go stale the moment they land.

Run segmentation on live records, not stale exports.

Strkr runs dynamic segments against contacts, companies, deals, and activities in one place, so campaigns, scoring, routing, and reports all pull from the same current list. Pricing is published. The feature pages show exactly what ships today.

Start free See the platform

Try it free. Bring your team next week.

No sales call, no migration consultant, no four-month implementation. Enter your card, get 14 days of the full Pro tier, cancel any time before day 14 with zero charge. Spin up a workspace, import your CSV, and have something useful before lunch.