← Back to Glossary
StrategyC

Customer Segmentation

Definition

Customer segmentation is the practice of dividing your customer base (or potential market) into groups that share meaningful characteristics. The goal is to make better product, marketing, and sales decisions by recognizing that different customers have different needs, different willingness to pay, and different definitions of value.

There are four common segmentation approaches. Firmographic segmentation divides by company attributes: industry, size, revenue, geography. Behavioral segmentation groups by how customers actually use your product: power users vs. casual, weekly active vs. monthly, single-feature vs. multi-feature. Needs-based segmentation clusters by the problem customers are solving: Airtable's users segment into project managers (need timelines), marketers (need content calendars), and product teams (need backlogs). Value-based segmentation ranks by revenue contribution or expansion potential.

The best segmentation combines multiple dimensions. Amplitude segments their customers by both firmographic data (startup vs. enterprise) and behavioral data (self-serve analytics vs. embedded analytics) because the combination predicts retention and expansion better than either dimension alone.

Why It Matters for Product Managers

Without segmentation, you're making one-size-fits-all decisions for a diverse customer base. Mailchimp learned this the hard way -- they initially built for small businesses and solo creators, but as enterprises adopted the platform, they faced competing needs. Small businesses wanted simplicity; enterprises wanted segmentation rules, API access, and compliance features. Mailchimp's 2019 decision to segment their product into tiers (Free, Essentials, Standard, Premium) directly reflected their customer segmentation work.

Segmentation drives three critical PM decisions. First, roadmap prioritization: knowing that your enterprise segment generates 70% of revenue but your SMB segment generates 70% of support tickets tells you where to invest differently. Second, pricing and packaging: different segments have different price sensitivities and feature needs. Third, retention strategy: the reasons a startup churns are completely different from why an enterprise churns.

Spotify's Discover Weekly feature emerged from behavioral segmentation. Their data showed that users who discovered new music in their first month retained at 2x the rate of users who only played familiar tracks. This segment-specific insight drove the feature that now serves 100M+ users.

How It Works in Practice

  • Start with your data. Export customer attributes (company size, industry, plan type, signup source) and behavioral data (feature usage, login frequency, support tickets, NPS scores). Tools like Amplitude, Mixpanel, or even a well-structured SQL query can do this.
  • Identify candidate dimensions. Test which attributes correlate with the outcomes you care about (retention, expansion, NPS). Company size alone is a weak segmenter. Company size + feature usage pattern + engagement frequency is usually much stronger.
  • Define 3-5 actionable segments. Each segment should be large enough to matter (10%+ of your base), distinct enough to treat differently, and stable enough that customers don't shift between segments weekly. Give each segment a memorable name your team will actually use -- "Growth-stage SaaS teams" is better than "Segment 3."
  • Validate with qualitative research. Numbers define segments; interviews explain them. Talk to 5-8 customers in each segment to understand their jobs-to-be-done, pain points, and definition of success. The quantitative data tells you who; the qualitative data tells you why.
  • Operationalize across the org. Segment data should flow into your CRM (for sales), your product analytics (for PM decisions), your support tool (for tiered support), and your marketing automation (for targeted messaging). If segmentation only exists in a spreadsheet, it won't change decisions.

    • Common Pitfalls

  • Over-segmenting. Creating 12 segments sounds thorough but is operationally impossible for most teams. If you can't give each segment a distinct product experience, messaging, and success metric, you have too many.
  • Segmenting on demographics alone. Industry and company size are easy to measure but weak predictors of behavior. Two 500-person SaaS companies can use your product in completely different ways. Layer in behavioral and needs-based dimensions.
  • Static segmentation. Segments aren't permanent. As your product evolves and your market shifts, your segments should too. Netflix re-segments their audience continuously based on viewing patterns, not once-a-year based on demographics.
  • Ignoring the segment you're losing. Teams naturally focus on the segment that's growing. But analyzing the segment that's churning or never converting often surfaces your biggest product gaps.

    • Related Concepts

    Ideal customer profile is a specific output of segmentation -- it defines your best-fit segment at the company level. Persona adds individual-level detail within each segment. Cohort analysis layers time-based analysis onto your segments to track how their behavior evolves.

    Related Terms

    PersonaIdeal Customer Profile (ICP)Cohort Analysis

    Frequently Asked Questions

    What is the difference between customer segmentation and cohort analysis?+
    Segmentation groups customers by shared attributes (industry, company size, behavior patterns) and is relatively stable over time. Cohort analysis groups customers by a shared time-based event (sign-up date, first purchase month) and tracks how that group's behavior changes over time. You might segment by company size, then run a cohort analysis within each segment to see if enterprise customers who signed up in Q1 retain differently than those from Q3.
    How many segments should a product team work with?+
    Three to five segments is the sweet spot for most SaaS products. Fewer than three means you're probably not differentiating enough. More than five creates operational complexity that most teams can't support -- each segment ideally gets distinct messaging, feature priorities, and success criteria. Spotify manages this at scale with dozens of listener segments, but they have a 200+ person product org to support it.
    ← PreviousCustomer OnboardingNext →DAU / MAU (Daily Active Users / Monthly Active Users)

    Explore More PM Terms

    Browse our complete glossary of 100+ product management terms.

    Full GlossaryExplore Frameworks