Product Segmentation
Product segmentation is the practice of partitioning your user base into distinct groups based on differences that matter for product decisions — usage patterns, business context, role, plan tier, lifecycle stage, or workflow. KnowMBA POV: positioning is downstream of segmentation. You cannot decide what to call your product, who to sell it to, or what to build next until you know which groups exist inside your base and which one you actually serve best. Useful segments meet three tests: (1) members of a segment behave more like each other than like outsiders on the dimension you care about, (2) the segment is large and accessible enough to act on, and (3) you can detect membership cheaply (in CRM, in product analytics, in form fields). Bad segmentation produces 'enterprise vs. SMB' buckets that hide the real distinctions; good segmentation produces 'design agencies running 5-15 client projects in parallel' — actionable.
The Trap
The trap is segmenting by what's easy to measure (company size, geography) rather than what predicts behavior (workflow, urgency, role of the buyer). 'Enterprise vs. mid-market vs. SMB' is the most common bad segmentation — it groups together a 200-person law firm and a 200-person manufacturing company that have nothing in common. The other trap: too many segments. A team that defines 12 segments will treat all 12 the same in practice (because no one can act on 12). The right number is usually 2-4 active segments, with a clear ranking of which one drives the roadmap. The third trap: stale segmentation. Segments defined at $2M ARR are usually wrong at $20M ARR — the business changed, the segment definitions did not.
What to Do
Run a segmentation refresh annually: (1) Pull behavior data on your top 200 customers — features used, usage frequency, expansion rate, churn rate. (2) Cluster by behavior, not by firmographics. K-means on usage vectors, or simple manual grouping by 'workflow type.' (3) Test segments against three filters — do members behave alike? is the segment >10% of revenue? can you identify members at signup? (4) Pick 1-2 'priority segments' that get product investment AND 1-2 'maintenance segments' that get bug fixes only. (5) Write segment definitions into your ICP doc, your CRM filters, and your product analytics dashboards. Re-validate quarterly: check that segment members still behave the way the definition predicts.
Formula
In Practice
April Dunford repeatedly emphasizes that the most common positioning failure is downstream of segmentation failure: companies cannot position because they have not made the segment cut. In her case work, she finds companies grouping users by company size when the actual behavioral split is by workflow (e.g., 'agencies billing time' vs. 'product teams shipping features'). The resulting positioning then fails because it tries to speak to both. Segment first, then position. Pendo's 2023 product analytics data across thousands of customer instances showed the same pattern internally: the strongest expansion-revenue segments were rarely the ones companies had originally labeled as 'enterprise' — they were behavioral clusters (e.g., 'workspaces with 3+ active integrations and >50 weekly active users') that cut across firmographic lines.
Pro Tips
- 01
Behavioral segmentation beats firmographic segmentation on every dimension that matters for product. Firmographic data is easy to collect and easy to act on at the SDR/marketing layer, but product decisions should be driven by what users DO, not who they ARE on paper.
- 02
Your 'priority segment' should drive at least 60% of net new ARR within 12 months of being declared. If it doesn't, either the segment is wrong or you didn't actually commit to it (you kept building for everyone).
- 03
Segment definitions should be testable in SQL. If you can't write a query that identifies all members of segment X in your data warehouse, the segment is too vague to act on.
Myth vs Reality
Myth
“More segments = more personalization = better outcomes”
Reality
Past 4-5 segments, marginal segments add coordination cost without changing decisions. Most companies operate effectively with 2-3 priority segments and treat the long tail uniformly. Personalization at the product layer is more valuable than segment proliferation at the strategy layer.
Myth
“SMB / Mid-Market / Enterprise is a real segmentation”
Reality
Company size is a coarse proxy for buying power, not a behavioral grouping. A 50-person digital agency and a 50-person dental practice share almost no product behavior despite identical headcount. Useful segmentation cuts across these labels.
Try it
Run the numbers.
Pressure-test the concept against your own knowledge — answer the challenge or try the live scenario.
Knowledge Check
Your B2B SaaS has segmented customers as 'SMB,' 'Mid-Market,' and 'Enterprise.' Within Mid-Market, you find churn is 8% for some accounts and 35% for others. What's the most likely root cause and the right diagnostic?
Industry benchmarks
Is your number good?
Calibrate against real-world tiers. Use these ranges as targets — not absolutes.
NRR Spread Between High-Fit and Low-Fit Segments
B2B SaaS — NRR difference between top behavioral quartile and bottom quartileTightly Targeted (small spread)
< 15 pts
Healthy
15-30 pts
Wide (segmentation needed)
30-50 pts
Very Wide (refocus required)
50-75 pts
Critical (you're serving wrong customers)
> 75 pts
Source: OpenView 2024 SaaS Benchmarks; Pendo Customer Cohort Data
Real-world cases
Companies that lived this.
Verified narratives with the numbers that prove (or break) the concept.
April Dunford Positioning Engagements (Pattern)
Multiple engagements 2018-2024
Across dozens of positioning engagements documented in 'Obviously Awesome,' April Dunford finds the same pattern: companies that struggle to position have failed to segment behaviorally. They've grouped customers by company size, geography, or industry — none of which predict product behavior. The repositioning work always begins with re-segmentation: pulling usage data, finding the behavioral cluster where unique attributes deliver disproportionate value, and rewriting the ICP around that cluster. Only then does positioning work follow. Companies that resist the segment cut continue to position generically and lose to focused competitors.
Companies Engaged
Dozens documented
Common Starting State
Firmographic segmentation, generic positioning
Common Intervention
Behavioral re-segmentation → narrow positioning
Typical Outcome
Win rate +20-40 pts in priority segment within 12 months
Positioning is downstream of segmentation. You cannot fix a positioning problem without first fixing the segmentation underneath it.
Related concepts
Keep connecting.
The concepts that orbit this one — each one sharpens the others.
Beyond the concept
Turn Product Segmentation into a live operating decision.
Use this concept as the framing layer, then move into a diagnostic if it maps directly to a current bottleneck.
Typical response time: 24h · No retainer required
Turn Product Segmentation into a live operating decision.
Use Product Segmentation as the framing layer, then move into diagnostics or advisory if this maps directly to a current business bottleneck.