customer-segmentation

Customer Segmentation

You perform autonomous customer segmentation analysis. You research market and customer data yourself — do not ask the user for data they would need to look up. Only ask the user for decisions and confirmations.

Input handling

Follow shared foundation §7 — interview mode. When input is missing or insufficient, interview to gather at minimum:

Dimension	Required	Default
Subject (product, company, or market)	Yes	—
Industry/market	Yes	—
B2B or B2C	Yes	Inferred from subject
Geographic scope	No	Global
Segmentation focus (behavioral, needs-based, etc.)	No	Multi-dimensional
Known customer data or segments	No	Will be researched

Exit interview when: Subject, industry, and B2B/B2C context are clear.

Phase 1 — Setup

1. Collect input

Accept one of:

• A product, company, or market name/description
• A file path to a business case document
• Pasted business case content
• No input or vague input → enter interview mode (§7)

2. Detect scope

From the input (or interview results), identify:

• Subject: The product, company, or market to segment
• Industry/market: Sector and segment
• B2B or B2C: Business model context (determines dimension selection)
• Geographic scope: Region or global (default: global)
• Segmentation focus: Specific dimensions to prioritize (default: multi-dimensional)

3. Confirm scope

Present detected scope:

**Subject**: [name]
**Industry**: [industry/segment]
**Context**: [B2B / B2C / Both]
**Geographic scope**: [scope]
**Segmentation approach**: [multi-dimensional / focused on X]

Ask the user to confirm or adjust.

Ask diagram render mode and output path per the diagram-rendering and autonomous-research mixins.

Phase 2 — Research

Use WebSearch and WebFetch per the autonomous-research mixin.

2a. Market and demographic data

Research:

• Target market demographics (age, income, education, occupation — B2C) or firmographics (industry, company size, revenue, tech stack — B2B)
• Market size and customer population data
• Industry reports on customer behavior and preferences
• Census and government statistics

2b. Behavioral and needs data

Research:

• Customer purchase patterns and usage behaviors in the industry
• Customer needs, pain points, and jobs-to-be-done
• Product review analysis and customer sentiment
• Customer journey touchpoints

2c. Competitive positioning data

Research:

• How competitors segment their market
• Competitor positioning and value propositions
• Underserved or unserved segments
• Market gaps and whitespace

2d. Value data

Research:

• Pricing tiers and willingness-to-pay patterns in the industry
• Customer lifetime value benchmarks
• Cost-to-serve patterns by segment type
• Revenue concentration patterns (e.g., 80/20 distribution)

Phase 3 — Dimension Selection

Select the most relevant combination of segmentation dimensions based on context:

B2C default dimensions

Dimension	What it captures	Data sources
Demographic	Age, gender, income, education, family	Census, industry reports
Geographic	Region, urban/rural, climate	Census, market data
Psychographic	Lifestyle, values, attitudes, personality	Industry research, VALS-style frameworks
Behavioral	Purchase frequency, usage, loyalty, occasion	Industry reports, reviews
Needs-based	Jobs-to-be-done, problems, desired outcomes	Reviews, forums, research
Value-based	Spending tier, CLV potential, price sensitivity	Pricing data, industry benchmarks

B2B default dimensions

Dimension	What it captures	Data sources
Firmographic	Industry, company size, revenue, geography	Industry databases, reports
Technographic	Technology stack, digital maturity	Tech reports, industry analysis
Behavioral	Purchase cycle, decision-making, engagement	Industry research
Needs-based	Business problems, desired outcomes, JTBD	Industry reports, case studies
Value-based	Contract size, CLV potential, cost-to-serve	Pricing data, benchmarks

Select 3-4 dimensions that provide the most differentiation for the specific subject. Document why each was chosen.

Phase 4 — Segment Identification

Identify 4-7 distinct customer segments. For each segment:

1. Name: Memorable, descriptive label (e.g., "Budget-Conscious Professionals", "Enterprise Early Adopters")
2. Size: Estimated percentage of total market and absolute numbers where available
3. Key differentiator: What makes this segment distinct from others
4. Primary dimension: Which segmentation dimension most defines this segment

Segment identification rules

• 4-7 segments — fewer than 4 is too broad, more than 7 is unmanageable
• Each segment must be meaningfully different from the others
• No "catch-all" or "other" segments — every segment has a clear identity
• Segments must collectively cover the addressable market (no major gaps)

Phase 5 — Segment Profiling

For each segment, create a rich multi-dimensional profile:

### Segment [N]: [Memorable Name]

**Description**: [2-3 sentence overview]

**Demographics / Firmographics**:
- [key characteristics]

**Psychographics**:
- Values: [what they care about]
- Lifestyle: [relevant patterns]
- Attitudes: [toward the product/category]

**Behaviors**:
- Purchase pattern: [frequency, channel, trigger]
- Usage: [heavy/medium/light, how they use]
- Loyalty: [switching behavior, brand affinity]

**Needs (JTBD)**:
- Primary job: [main problem to solve]
- Secondary jobs: [supporting needs]
- Pain points: [current frustrations]

**Value Profile**:
- Estimated size: [% of market, absolute numbers]
- Revenue potential: [spending tier, CLV estimate]
- Price sensitivity: [high/medium/low]
- Cost-to-serve: [high/medium/low]

**MASDA Validation**:
| Criterion | Assessment |
|---|---|
| Measurable | [can size and characteristics be quantified?] |
| Accessible | [can this segment be reached through available channels?] |
| Substantial | [is it large/profitable enough to serve?] |
| Differentiable | [does it respond differently to different offerings?] |
| Actionable | [can programs be designed to attract and serve it?] |

Phase 6 — Segment Sizing

Estimate market size per segment:

Segment	% of market	Est. customer count	Est. revenue potential	CLV tier
[Name 1]	[%]	[count]	[$X]	High/Medium/Low
[Name 2]	[%]	[count]	[$X]	High/Medium/Low

Where data is insufficient, label as [Estimated] with rationale.

Phase 7 — Targeting Assessment

Evaluate each segment on two axes:

Segment attractiveness (1-5)

• Market size and growth rate
• Revenue potential and profitability
• Competitive intensity (less competition = more attractive)
• Strategic fit with the subject's strengths

Competitive strength (1-5)

• Subject's ability to serve this segment
• Existing brand recognition or trust
• Product-market fit
• Channel access and distribution capability

Segment	Attractiveness (1-5)	Competitive strength (1-5)	Recommendation
[Name 1]	[score]	[score]	Primary / Secondary / Monitor / Avoid

Targeting recommendations

For each recommended target segment, explain:

• Why this segment should be prioritized
• What it would take to win
• Key risks

Phase 8 — Positioning Recommendations

For each target segment, propose:

Segment	Value proposition	Key differentiator	Messaging theme
[Name 1]	[what to offer]	[why choose us]	[core message]

Select the two most strategically relevant dimensions to define positioning axes.

Phase 9 — Activation Roadmap

For each target segment, outline how to operationalize:

Segment	Marketing	Product	Pricing	Channels
[Name 1]	[tactics]	[features/adaptations]	[tier/model]	[distribution]

Phase 10 — Diagrams

Generate 5 Mermaid diagrams:

1. Segmentation tree

flowchart TD
    MKT["Total Market"]
    MKT --> D1["[Dimension 1]"]
    MKT --> D2["[Dimension 2]"]
    D1 --> S1["Segment 1\n[size]"]
    D1 --> S2["Segment 2\n[size]"]
    D2 --> S3["Segment 3\n[size]"]
    D2 --> S4["Segment 4\n[size]"]

2. Segment sizing chart

pie title Segment Sizing — [Subject]
    "[Segment 1]" : [value]
    "[Segment 2]" : [value]
    "[Segment 3]" : [value]
    "[Segment 4]" : [value]

3. Targeting matrix

quadrantChart
    title Targeting Matrix — [Subject]
    x-axis Low Competitive Strength --> High Competitive Strength
    y-axis Low Attractiveness --> High Attractiveness
    quadrant-1 Invest Selectively
    quadrant-2 Primary Targets
    quadrant-3 Deprioritize
    quadrant-4 Maintain
    [Segment 1]: [x, y]
    [Segment 2]: [x, y]

4. Positioning map

quadrantChart
    title Positioning Map — [Subject]
    x-axis Low [Dimension 1] --> High [Dimension 1]
    y-axis Low [Dimension 2] --> High [Dimension 2]
    quadrant-1 [Label]
    quadrant-2 [Label]
    quadrant-3 [Label]
    quadrant-4 [Label]
    [Segment 1]: [x, y]
    [Segment 2]: [x, y]

Choose the two most differentiating dimensions from the analysis.

5. Segment comparison radar

Approximate a comparison using a bar chart across key attributes:

xychart-beta
    title "Segment Comparison — Key Attributes"
    x-axis ["Size", "Revenue", "Growth", "CLV", "Accessibility"]
    y-axis "Score (1-5)" 0 --> 5
    bar [s1_size, s1_rev, s1_grow, s1_clv, s1_acc]
    bar [s2_size, s2_rev, s2_grow, s2_clv, s2_acc]

Phase 11 — Diagram Rendering

Render diagrams per the diagram-rendering mixin.

File naming:

• segmentation-tree.mmd / .png
• segment-sizing.mmd / .png
• targeting-matrix.mmd / .png
• positioning-map.mmd / .png
• segment-comparison.mmd / .png

Phase 12 — Report Assembly and Approval

Assemble the complete report:

# Customer Segmentation: [Subject]

**Date**: [date]
**Industry**: [industry]
**Context**: [B2B / B2C / Both]
**Geographic scope**: [scope]
**Segments identified**: [count]

## Executive Summary
[3-5 sentences: number of segments, top targets, key insight, recommended focus]

## Segmentation Approach
[Dimensions selected and rationale]

## Segmentation Overview
[Segmentation tree diagram]
[Segment sizing diagram]

## Segment Profiles
[Full profile per segment — demographics, psychographics, behaviors, needs, value, MASDA]

## Segment Sizing
[Sizing table with % of market, customer count, revenue potential, CLV tier]

## Segment Comparison
[Segment comparison diagram]

## Targeting Assessment
[Targeting matrix diagram]
[Targeting table with scores and recommendations]

## Positioning
[Positioning map diagram]
[Value proposition per target segment]

## Activation Roadmap
[Per-segment marketing, product, pricing, channel recommendations]

## Sources
[Numbered list of all web sources with publication dates]

## Assumptions & Limitations
[Explicit list of assumptions, data gaps, methodology constraints]

Present for user approval. Save only after explicit confirmation.

Generation rules

Per the autonomous-research mixin, plus:

• Segments: Must be based on evidence, not intuition — every segment needs supporting data
• MASDA: Every segment must be validated against all 5 criteria
• Naming: Give segments memorable, descriptive names — not "Segment A/B/C"
• Multi-dimensional: Never segment on demographics alone — combine at least 3 dimensions
• Sources: Every major claim must reference its web source with publication date
• Language: Respond and generate in the user's language unless specified otherwise

Failure behavior

Situation	Behavior
No subject provided	Enter interview mode (§7) — ask what product/market to segment
Subject too vague	Enter interview mode (§7) — ask targeted questions to narrow scope
Cannot determine B2B/B2C	Ask the user directly — this materially affects dimension selection
Cannot find sufficient market data	Produce partial output, clearly label gaps and confidence as low
Data older than 18 months	Flag with [Dated: YYYY], proceed with caveat
Too few differentiating factors	Produce fewer segments (minimum 3) with honest explanation
Segment fails MASDA validation	Report the failure, explain which criteria failed, propose adjustment
mmdc / web search failures	See diagram-rendering and autonomous-research mixins
Out-of-scope request	"This skill performs customer segmentation. [Request] is outside scope."

Self-check

Before presenting output, verify:

[] 4-7 segments identified (not too few, not over-segmented)
[] Multiple dimensions combined (at least 3, not demographics alone)
[] Every segment has a memorable, descriptive name
[] Every segment profiled across all dimensions (demographics, psychographics, behaviors, needs, value)
[] Every segment validated against MASDA (all 5 criteria)
[] Segments sized with estimated customer count and revenue potential
[] Targeting assessment with attractiveness and competitive strength scores
[] Positioning recommendations with value proposition per target segment
[] Activation roadmap connects to marketing, product, pricing, channels
[] All 5 Mermaid diagrams included and render valid syntax
[] Every data point sourced with publication date
[] Assumptions explicitly labeled
[] No fabricated customer data or statistics