igny8/v2/Igny8 V2 New Final Plans/SAG-Master-Document.md

# Semantic Authority Grid (SAG) — Master Strategy Document

**Version:** 2.0  
**Status:** Definitive Reference  
**Scope:** Complete SAG methodology — strategy, clustering, architecture, content types, interlinking, and long-term authority building  
**Usage:** Strategic reference for SEO site architecture, content planning, and development implementation

---

## Table of Contents

1. [What is SAG](#1-what-is-sag)
2. [Core Principles](#2-core-principles)
3. [The Dimensional Framework](#3-the-dimensional-framework)
4. [Cluster Formation Methodology](#4-cluster-formation-methodology)
5. [Cluster Types & Content Mapping](#5-cluster-types--content-mapping)
6. [Taxonomy Derivation](#6-taxonomy-derivation)
7. [Site Architecture from Clusters](#7-site-architecture-from-clusters)
8. [Content Structures per Type](#8-content-structures-per-type)
9. [Internal Linking Architecture](#9-internal-linking-architecture)
10. [URL Structure Strategy](#10-url-structure-strategy)
11. [SEO Schema & Metadata](#11-seo-schema--metadata)
12. [Authority Compounding Model](#12-authority-compounding-model)
13. [SAG Application Workflow](#13-sag-application-workflow)
14. [Niche Adaptation Framework](#14-niche-adaptation-framework)
15. [Validation & Quality Criteria](#15-validation--quality-criteria)
16. [Anti-Patterns](#16-anti-patterns)
17. [Real-World Reference: MassagerSmart](#17-real-world-reference-massagersmart)
18. [Appendix: Clustering Prompt Specification](#18-appendix-clustering-prompt-specification)

---

## 1. What is SAG

The Semantic Authority Grid is a structured SEO and content architecture framework that organizes keywords, content, and commercial pages into interconnected clusters. Each cluster functions as a self-contained mini-ecosystem with its own hub page, supporting blogs, taxonomy term pages, and commercial content (products or services) — all bound together by systematic internal linking.

SAG is not a clustering algorithm. It is a site architecture philosophy where **keywords dictate structure**, not the other way around. The output of SAG is not a list of keyword groups — it is a complete site blueprint that defines every page type, every taxonomy, every internal link relationship, and every content template needed to build a topically authoritative website.

### What SAG Produces

Given a set of keywords (or a product/service catalog), SAG outputs:

- **Dimensional axes** — the natural facets that organize the niche (e.g., device type, body area, condition, technique, brand)
- **Clusters** — groups of keywords at meaningful dimensional intersections, each forming a self-contained topical ecosystem
- **Cluster types** — classification of each cluster's purpose (hub, blog, category, product, comparison, informational)
- **Taxonomy structure** — categories, tags, and custom taxonomies derived from dimensional axes
- **Page templates** — content structure specifications per cluster type
- **URL hierarchy** — logical URL paths reflecting the dimensional structure
- **Internal linking map** — directional authority flow between all pages
- **Content plan** — what to write, in what order, targeting which keywords

### Why SAG Matters

Search engines reward **depth + structure**, not isolated keyword pages. SAG prevents keyword fragmentation, thin content, random tagging, and chaotic navigation. Instead, it delivers depth, clarity, and authority signals — the foundation of long-term rankings.

Sites built on SAG architecture achieve:
- Higher ranking effectiveness per unit of domain authority (DR/DA)
- Dramatically more pages ranking without direct backlinks (25:1 ratio demonstrated)
- Faster crawling and indexation of new content (sub-10-minute crawl times demonstrated)
- Lower DR threshold for passive ranking of new content
- Compounding authority that increases over years, not just months

---

## 2. Core Principles

### 2.1 Keywords First, Structure Follows

Site categories, tags, taxonomies, and page hierarchy must be built AFTER keywords are analyzed and grouped into clusters. Never force keywords into a pre-existing site structure. The dimensional reality of the keyword set determines the architecture.

### 2.2 Clusters are Mini-Ecosystems

Each cluster is a self-contained topical authority unit. A complete cluster contains:
- 1 hub/anchor page (the authority center)
- 3-10+ supporting content pieces (blogs, guides, comparisons)
- Taxonomy term pages (derived from cluster dimensions)
- Commercial pages (products or services) where applicable
- Internal links connecting all elements

A cluster should feel like a complete resource on its topic. A user entering any page within the cluster should be able to naturally navigate to every other page through logical internal links.

### 2.3 Multi-Dimensional Analysis

Keywords are not grouped by single-word matching or simple semantic similarity. They are analyzed across multiple dimensions simultaneously. A keyword like "best heated shiatsu foot massager for neuropathy" sits at the intersection of: Product Type (foot massager) + Feature (heated) + Technique (shiatsu) + Condition (neuropathy). The cluster it belongs to is determined by which dimensional combination creates the strongest ecosystem.

### 2.4 Internal Linking as Authority Architecture

Internal links are not decorative. They are the mechanism that distributes authority from backlinked pages to un-linked pages. SAG defines directional linking rules:
- Blogs → their cluster hub
- Hubs → attribute term pages + products within the cluster
- Attribute term pages → hub + related products
- Products → hub + comparison/review blogs
- Cross-cluster links only between related hubs

This structured flow ensures every backlink built to any page in the system benefits every other page through predictable authority distribution.

### 2.5 Scalability Without Chaos

New keywords slot into existing clusters or form new ones without breaking the structure. New products inherit the dimensional framework automatically. The architecture grows organically because the dimensional axes are stable — only the values within them expand.

### 2.6 Exclusion Over Forced Inclusion

It is better to leave keywords unclustered than to force weak groupings. Only cluster keywords with strong multi-dimensional bonds. Outliers that don't fit any ecosystem are excluded from the structure entirely. Quality of clusters always trumps quantity.

---

## 3. The Dimensional Framework

### 3.1 What are Dimensions?

Dimensions are the natural facets or axes along which keywords in a niche can be classified. Every niche has its own dimensional structure, but dimensions generally fall into universal categories that manifest differently per industry.

### 3.2 Universal Dimension Categories

| # | Dimension Category | Description | Examples |
|---|-------------------|-------------|----------|
| 1 | **Topic/Subject** | The core thing being discussed | Foot massager, vitamin C serum, auto repair |
| 2 | **Problem/Need** | The pain point or desire driving the search | Neuropathy relief, skin brightening, brake failure |
| 3 | **Solution/Method** | How something is done or achieved | Shiatsu technique, LED therapy, OBD diagnosis |
| 4 | **Feature/Attribute** | Specific characteristics | Heated, waterproof, organic, portable, cordless |
| 5 | **Persona/Audience** | Who needs this | Beginners, professionals, pregnant women, diabetics |
| 6 | **Use-Case/Application** | Where/when/why it's used | Travel, home office, summer, pregnancy |
| 7 | **Product Type/Format** | Category or form factor | EMS pad, roller, pillow, serum, cream |
| 8 | **Brand/Manufacturer** | Specific brands | Renpho, TheraFace, Nekteck, Mario Badescu |
| 9 | **Comparison/Alternative** | Explicit or implied comparisons | Shiatsu vs percussion, vitamin C vs retinol |
| 10 | **Context/Modifier** | Geographic, temporal, or situational qualifiers | UK, near me, 2024, under $50 |

### 3.3 Niche-Specific Dimension Manifestation

The same universal categories manifest differently per industry:

**E-Commerce (Massagers):**
- Topic → Body area (neck, foot, face, back)
- Product Type → Device type (shiatsu, EMS, percussion, roller)
- Solution → Therapy technique (electrical stimulation, deep tissue, vibration)
- Problem → Health condition (neuropathy, plantar fasciitis, circulation)
- Feature → Device features (heated, cordless, portable)
- Brand → Manufacturer (Renpho, Nekteck, TheraFace)

**SaaS Company:**
- Topic → Feature category (analytics, automation, reporting)
- Persona → Company size (startup, SMB, enterprise)
- Use-Case → Industry vertical (fintech, healthcare, e-commerce)
- Product Type → Integration type (API, plugin, native)
- Problem → Business challenge (scaling, compliance, efficiency)
- Comparison → Competitor alternatives (vs Competitor A, vs Competitor B)

**Services Business (Law Firm):**
- Topic → Practice area (personal injury, family law, corporate)
- Persona → Client type (individual, business, nonprofit)
- Use-Case → Case type (car accident, divorce, contract dispute)
- Context → Location (city, county, state)
- Problem → Legal need (lawsuit, defense, compliance)
- Solution → Service delivery (consultation, representation, mediation)

**Content/Media Site:**
- Topic → Subject area (technology, health, finance)
- Persona → Reader level (beginner, intermediate, expert)
- Use-Case → Content purpose (tutorial, news, opinion, review)
- Context → Timeliness (evergreen, trending, seasonal)
- Format → Content type (guide, listicle, comparison, case study)

### 3.4 Discovering Dimensions

Dimensions are discovered, not assumed. The process:

1. **Collect all keywords** — gather the complete keyword universe for the niche
2. **Identify recurring modifiers** — find words/phrases that appear across multiple keywords as qualifying descriptors
3. **Group modifiers into axes** — modifiers that represent the same type of qualification form a dimension
4. **Validate dimensional independence** — each dimension should represent a genuinely different facet (not a synonym of another dimension)
5. **Enumerate dimension values** — list all unique values within each dimension

**Example Discovery Process:**

Keywords: "heated foot massager," "shiatsu foot massager," "foot massager for neuropathy," "best foot massager 2024," "heated shiatsu back massager," "neck massager for pain," "EMS foot massager mat"

Recurring modifiers:
- Body area: foot, back, neck → **Dimension: Massage Area**
- Technique: heated, shiatsu, EMS → **Dimension: Device Type/Technique**
- Condition: neuropathy, pain → **Dimension: Relief Focus**
- Intent: best, 2024 → **Dimension: Search Context** (may or may not form a structural dimension)
- Format: mat, massager → **Dimension: Product Format**

### 3.5 Dimensional Intersection = Cluster Potential

Clusters form at meaningful intersections of 2+ dimensions. The more dimensions that intersect, the stronger the cluster:

- **2-dimensional intersection:** Foot (area) × Shiatsu (type) → "Shiatsu Foot Massagers" — viable cluster
- **3-dimensional intersection:** Foot (area) × EMS (type) × Neuropathy (condition) → "EMS Foot Massagers for Neuropathy" — strong cluster
- **4-dimensional intersection:** Foot (area) × Heated (feature) × Shiatsu (type) × Circulation (condition) → Very specific, may be too narrow for a standalone cluster but excellent as supporting content within a broader cluster

The optimal cluster sits at a 2-3 dimensional intersection that is:
- Broad enough to support 5-20 keywords
- Specific enough to have a clear hub page topic
- Deep enough to support 3-10+ supporting content pieces

---

## 4. Cluster Formation Methodology

### 4.1 Step 1: Identify Natural Semantic Overlaps

Keywords cluster together when they share 2+ dimensional intersections. Shared words alone don't make a cluster. "Back massager" and "back support pillow" share the word "back" but represent completely different product ecosystems.

**Dimensional intersection test:**
- "best foot massagers for plantar fasciitis" → Product Type (foot massager) + Condition (plantar fasciitis) + Intent (best/commercial)
- "plantar fasciitis foot massager reviews" → Product Type (foot massager) + Condition (plantar fasciitis) + Format (reviews)
- "do foot massagers help plantar fasciitis" → Product Type (foot massager) + Condition (plantar fasciitis) + Intent (informational)

These share 2 strong dimensions (product type + condition) and form a natural cluster.

### 4.2 Step 2: Map User Journey Patterns

Each cluster should support natural user exploration. Valid journey patterns:

**Problem → Information → Solution → Product:**
"what causes plantar fasciitis" → "best treatments for plantar fasciitis" → "foot massagers for plantar fasciitis" → specific product pages

**General → Specific → Variant:**
"foot massagers" → "shiatsu foot massagers" → "heated shiatsu foot massagers with remote"

**Question → Explanation → Options → Decision:**
"do vitamin C serums work" → "how vitamin C brightens skin" → "best vitamin C serums" → "vitamin C vs retinol"

**Feature Discovery → Feature Comparison → Feature + Use-Case:**
"heated massagers" → "heated vs unheated massagers" → "heated massagers for arthritis"

**Cluster journey test:** Can a user naturally move from any keyword in the cluster to any other keyword through a logical research or buying journey?

### 4.3 Step 3: Identify Cluster Anchors (Hub Potential)

Every cluster needs one clear root keyword that serves as the hub page anchor.

**Hub characteristics:**
- Broad enough to encompass all cluster keywords
- Specific enough to be commercially or informationally valuable
- Natural landing point for the user journey
- Can support 3-10+ supporting content pieces

**Good hubs:**
- ✅ "Foot Massagers for Plantar Fasciitis" — supports safety, features, comparisons, use-cases
- ✅ "Vitamin C Serums for Skin Brightening" — supports skin types, vs alternatives, application guides
- ✅ "Best Heated Neck Massagers" — supports brands, features, conditions, buying guide

**Bad hubs:**
- ❌ "Massagers" — too broad, no natural ecosystem boundary
- ❌ "Renpho foot massager cord length" — too narrow, can't support an ecosystem
- ❌ "Products" — meaningless, not topical

### 4.4 Step 4: Extract Attribute Dimensions

For each cluster, identify taxonomy-worthy dimensions — recurring modifiers that could become product filters, content tags, URL parameters, or faceted navigation.

**Extraction method:** Look for modifiers that repeat across cluster keywords:
- Material: cotton, linen, silk, organic
- Feature: heated, waterproof, portable, wireless
- Size: king, queen, twin, compact
- Type: shiatsu, percussion, vibration, EMS
- Audience: men, women, seniors, athletes
- Context: travel, home, office, car

These become the custom taxonomies in the site architecture.

### 4.5 Step 5: Resolve Keyword Conflicts

Before finalizing:
1. Identify keywords that could fit multiple clusters
2. Assign to the cluster with the MOST dimensional intersections
3. If tied, assign to the MORE SPECIFIC cluster (narrower topic)
4. If still tied, assign to the cluster representing the PRIMARY user intent
5. Last resort: create a new cluster if the keyword represents a unique ecosystem
6. Each keyword appears in exactly one cluster — no duplication

### 4.6 Formation Rules

| Rule | Requirement |
|------|-------------|
| **Semantic Coherence** | Keywords must share meaningful semantic relationships, not just word overlap |
| **Dimensional Intersection** | Minimum 2 shared dimensions required; 3+ = strong cluster member |
| **User Journey Viability** | Natural navigation paths must exist within the cluster |
| **Ecosystem Completeness** | Must support 1 hub + 3-10 supporting articles |
| **Minimum Size** | 5 keywords minimum (below this, it's not an ecosystem) |
| **Maximum Size** | 20 keywords maximum (above this, consider splitting) |
| **Optimal Size** | 7-15 keywords (hub + supporting + variants) |
| **No Duplication** | Each keyword in exactly one cluster |
| **Quality Over Quantity** | 5 strong clusters > 15 weak clusters |

---

## 5. Cluster Types & Content Mapping

Not all clusters are the same. Each cluster has a **type** that determines what kind of content it produces, what page template it uses, and how it fits into the site architecture.

### 5.1 Cluster Type Classification

| Cluster Type | Purpose | Hub Page Type | Content Focus | Example |
|-------------|---------|---------------|---------------|---------|
| **Product Cluster** | Group commercial products by dimensional intersection | Product category page | Feature comparison, buying guidance, product listings | "Shiatsu Foot Massagers" |
| **Condition/Problem Cluster** | Address specific problems or use-cases | Cluster hub / landing page | Problem education, solution overview, product recommendations | "Foot Massagers for Neuropathy" |
| **Feature Cluster** | Organize by specific product/service features | Feature landing page | Feature explanation, comparison with alternatives, best-of lists | "Heated Neck Massagers" |
| **Brand Cluster** | Group by manufacturer/brand | Brand comparison page | Brand review, model comparison, brand vs brand | "Top Neck Massager Brands Compared" |
| **Informational Cluster** | Pure educational content | Blog/guide hub | How-to, tips, techniques, safety information | "How to Massage Your Neck" |
| **Comparison Cluster** | Head-to-head evaluations | Comparison/roundup page | A vs B, buying guide, machine comparison | "Neck Massage Machine Buying Guide" |
| **Attribute Cluster** | Organize by taxonomy dimension value | Attribute term landing page | Attribute-specific content with related products/services | "EMS Therapy Devices" |

### 5.2 How to Determine Cluster Type

Analyze the dominant intent and dimensional composition:

1. **If the cluster is anchored by a product category keyword** → Product Cluster
2. **If the cluster is anchored by a problem/condition keyword** → Condition/Problem Cluster
3. **If the cluster is anchored by a specific feature** → Feature Cluster
4. **If the cluster groups multiple brands** → Brand Cluster
5. **If the cluster is purely informational** → Informational Cluster
6. **If the cluster centers on comparisons** → Comparison Cluster
7. **If the cluster represents a single dimension value** → Attribute Cluster

### 5.3 Cluster Type → Content Type Mapping

| Cluster Type | Primary Content Type | WordPress Post Type | Content Structure |
|-------------|---------------------|--------------------|--------------------|
| Product Cluster | Product category page | `product_cat` term or custom page | `category_archive` |
| Condition/Problem Cluster | Cluster hub page | `page` or custom post type | `cluster_hub` |
| Feature Cluster | Feature landing page | `page` | `landing_page` |
| Brand Cluster | Brand comparison post | `post` | Blog article |
| Informational Cluster | Blog post | `post` | Blog article |
| Comparison Cluster | Comparison/roundup post | `post` | Blog article |
| Attribute Cluster | Attribute term page | Taxonomy term (e.g., `pa_type`) | `attribute_archive` |

---

## 6. Taxonomy Derivation

### 6.1 Dimensions Become Taxonomies

The dimensional axes discovered during keyword analysis translate directly into site taxonomies. This is not metaphorical — each dimension literally becomes a WordPress taxonomy (or equivalent in other CMS platforms).

**Massager Niche Example:**

| Dimension | WordPress Taxonomy | Term Examples |
|-----------|-------------------|---------------|
| Body Area | `massage_area` | Neck, Foot, Face, Back, Eye, Scalp |
| Device Type | `massager_type` | Shiatsu, EMS, Percussion, Vibration, Roller |
| Technique | `massage_technique` | Deep Tissue, Electrical Stimulation, Kneading |
| Condition | `relief_focus` | Neuropathy, Plantar Fasciitis, Circulation, Pain Relief |
| Feature | `product_feature` (tag) | Heated, Cordless, Portable, Waterproof |
| Brand | `brand` | Renpho, Nekteck, TheraFace, Homedics |

### 6.2 Taxonomy Hierarchy

Not all taxonomies are equal. They form a natural hierarchy:

**Primary Taxonomies (Categories):** The major organizational axes that define site navigation. Usually 2-3 per site.
- Example: Massage Area (Neck, Foot, Face) + Device Type (Shiatsu, EMS, Heated)

**Secondary Taxonomies (Custom Taxonomies):** Important classification axes that provide filtering and additional landing pages.
- Example: Relief Focus (Neuropathy, Circulation), Technique (Deep Tissue, Kneading)

**Tertiary Taxonomies (Tags):** Descriptive attributes that don't warrant their own archive pages but aid in cross-referencing.
- Example: Features (Cordless, Portable), Price Range (Under $50, $50-100)

### 6.3 Every Taxonomy Term is a Potential SEO Landing Page

In SAG, taxonomy terms are not just organizational labels. Each term can (and should, for primary/secondary taxonomies) have its own rich content — an SEO-optimized landing page that:
- Defines and explains the term
- Lists related products/services
- Links to relevant blog content
- Includes FAQs
- Targets term-specific keywords
- Uses appropriate schema markup

A "Shiatsu" taxonomy term page is not just a list of shiatsu products. It's a comprehensive landing page about shiatsu massage technology, with product recommendations, technique explanations, comparison content, and internal links to the shiatsu-related cluster hub and supporting blog posts.

---

## 7. Site Architecture from Clusters

### 7.1 Architecture Hierarchy

SAG produces a clear site hierarchy:

```
Homepage
├── [Primary Category 1] ← Taxonomy dimension 1
│   ├── [Cluster Hub A] ← Condition/Problem cluster
│   │   ├── Blog Post 1 ← Supporting content
│   │   ├── Blog Post 2 ← Supporting content
│   │   ├── Product 1 ← Commercial page
│   │   └── Product 2 ← Commercial page
│   ├── [Cluster Hub B] ← Feature cluster
│   │   ├── Blog Post 3
│   │   └── Product 3
│   └── [Attribute Term Page] ← Taxonomy term landing page
├── [Primary Category 2]
│   ├── [Cluster Hub C]
│   │   ├── Blog Post 4
│   │   └── Product 4
│   └── [Attribute Term Page]
├── Blog Archive
│   └── All blog posts (also accessible through cluster hubs)
└── Brand Pages
    ├── Brand A Comparison
    └── Brand B Comparison
```

### 7.2 MassagerSmart Architecture Example

```
massagersmart.com
├── /foot-massager/ ← Product Category (Primary Taxonomy: Massage Area)
│   ├── /best-foot-massagers-for-neuropathy-.../ ← Cluster Hub (Condition Cluster)
│   │   ├── /blog/ankle-heel-foot-massagers-.../ ← Supporting Blog
│   │   ├── /blog/foot-message-vs-massage-.../ ← Supporting Blog
│   │   └── /shop/foot-massager/portable-ems-foot-massager-.../ ← Product
│   ├── /best-foot-massagers-for-plantar-fasciitis-.../ ← Cluster Hub
│   ├── /best-shiatsu-foot-massagers-.../ ← Cluster Hub (Product Type Cluster)
│   ├── /top-heated-electric-ems-foot-massagers-.../ ← Cluster Hub (Feature Cluster)
│   └── /top-foot-massager-brands-compared-.../ ← Brand Cluster
├── /neck-massager/
│   ├── /portable-cordless-handheld-neck-massagers-.../ ← Cluster Hub
│   ├── /heated-electric-neck-massagers-.../ ← Cluster Hub
│   ├── /shiatsu-neck-massagers-.../ ← Cluster Hub
│   ├── /neck-shoulder-massagers-.../ ← Cluster Hub
│   └── /top-neck-massager-brands-compared-.../ ← Brand Cluster
├── /face-massager/
│   ├── [Electric/LED/Microcurrent cluster hubs]
│   ├── [Manual tools cluster hub]
│   ├── [Skin healing cluster hub]
│   └── [Brand cluster hub]
├── Taxonomy Term Pages (each with rich SEO content)
│   ├── /massager-type/shiatsu/ ← Attribute term: Shiatsu
│   ├── /massager-type/ems/ ← Attribute term: EMS
│   ├── /relief-focus/neuropathy/ ← Attribute term: Neuropathy
│   ├── /massage-technique/electrical-stimulation/ ← Attribute term
│   └── ...
└── /blog/
    ├── /blog/how-to-massage-your-neck/ ← Info cluster blog
    ├── /blog/neck-shoulder-massage-techniques-.../ ← Info cluster blog
    └── /blog/neck-massage-machine-buying-guide/ ← Comparison cluster blog
```

### 7.3 Architecture Rules

1. **Every page belongs to at least one cluster** — no orphan pages
2. **Every cluster has exactly one hub page** — the authority anchor
3. **Every product/service is assigned to a primary cluster** — and optionally tagged with secondary clusters
4. **Every taxonomy term with 3+ associated products/pages gets its own landing page** — below that threshold, it exists only as a tag
5. **Blog posts always support a specific cluster** — random blog posts without cluster assignment violate SAG
6. **Cross-cluster relationships are defined explicitly** — not every hub links to every other hub; only dimensionally related hubs cross-link

---

## 8. Content Structures per Type

### 8.1 Cluster Hub Page

The authority anchor for the cluster. Comprehensive, long-form content that positions the site as the definitive resource on the cluster's topic.

**Structure:**
1. Introduction + topic definition (what it is, why it matters)
2. Key considerations / buyer's guide / selection criteria
3. Top recommendations (product/service listings with brief reviews)
4. Comparison table (if applicable)
5. Detailed breakdowns by sub-category or feature
6. Use-case specific guidance
7. FAQs (targeting long-tail question keywords)
8. Related resources (internal links to supporting blogs and attribute pages)

**Word count:** 2,000-5,000+ words
**Schema:** `CollectionPage` or `WebPage`, `FAQPage` if FAQs present, `ItemList` for product listings

### 8.2 Blog / Supporting Content

Depth pieces that expand semantic coverage for the cluster. Each targets specific supporting keywords.

**Structures by sub-type:**

**How-To Guide:**
1. Introduction + what you'll learn
2. Step-by-step process
3. Tips and common mistakes
4. When to seek professional help
5. Related product recommendations

**Comparison Post:**
1. Introduction + comparison criteria
2. Side-by-side analysis
3. Comparison table
4. Verdict / recommendation
5. FAQs

**Informational Article:**
1. Topic introduction
2. Key concepts explained
3. Benefits / risks / considerations
4. Expert insights or data
5. Actionable takeaways
6. Related resources

**Word count:** 1,000-2,500 words
**Schema:** `Article`, `BlogPosting`, `HowTo` where applicable

### 8.3 Product Page

Individual product or service offering with conversion-focused content.

**Structure:**
1. Product name + tagline
2. Hero image + gallery
3. Key specifications table
4. Feature/benefit sections (organized by dimensional attributes)
5. Use-case scenarios
6. Comparison with alternatives (links to comparison blog)
7. Customer reviews / testimonials
8. FAQs specific to product
9. Related products (same cluster)

**Additional Information Section (taxonomy-driven):**
- Device Type & Target Area (Massager Category, Massage Area, Massager Type)
- Therapy Details (Massage Technique, Relief Focus, Therapy Method)
- Each attribute value links to its taxonomy term page

**Word count:** 800-2,000 words
**Schema:** `Product`, `Review`, `AggregateRating`, `FAQPage`

### 8.4 Taxonomy Term Landing Page

SEO-optimized landing page for each taxonomy term value.

**Structure:**
1. Introduction + term definition
2. Key subtopics (H2/H3 outline covering the term's scope)
3. Top related products/services (filtered by this term)
4. Related blog posts
5. Comparison with alternative terms (e.g., "Shiatsu vs Percussion")
6. FAQs
7. Internal links to related cluster hubs

**Word count:** 1,000-2,000 words
**Schema:** `CollectionPage` or `WebPage`, `FAQPage`

### 8.5 Product/Service Category Page

Top-level category page organizing products/services by primary taxonomy.

**Structure:**
1. Category introduction + overview
2. Sub-category navigation (links to cluster hubs within this category)
3. Best-of / featured items
4. Buying guide summary
5. All products/services grid with filters
6. FAQs
7. Related categories

**Word count:** 1,500-3,000 words
**Schema:** `CollectionPage`, `ItemList`, `BreadcrumbList`

### 8.6 Brand Page

Comparison and review content for a specific brand or across multiple brands.

**Structure:**
1. Brand overview / introduction
2. Product lineup analysis
3. Feature comparison table (brand's products or brand vs competitors)
4. Pros and cons
5. Best use-cases for this brand
6. Customer sentiment summary
7. Alternatives to consider

**Word count:** 1,500-3,000 words
**Schema:** `Article`, `Brand` (where applicable)

### 8.7 Service Page

For service-based businesses rather than product-based.

**Structure:**
1. Service overview + value proposition
2. Process steps (how the service works)
3. Outcomes / deliverables
4. Who it's for (persona targeting)
5. Pricing / packages (if applicable)
6. Case studies / results
7. FAQs
8. Local SEO sections (if location-based)
9. Contact / CTA

**Word count:** 1,000-2,500 words
**Schema:** `Service`, `LocalBusiness` (if applicable), `FAQPage`

---

## 9. Internal Linking Architecture

### 9.1 The Authority Flow Model

Internal links in SAG are not random or decorative. They form a deliberate authority distribution system. The goal: every backlink built to any page in the site benefits every other page through predictable internal link equity flow.

### 9.2 Linking Rules by Page Type

| Source Page | Links TO | Purpose |
|------------|----------|---------|
| **Blog Post** | Its cluster hub (mandatory) | Funnels authority to hub |
| **Blog Post** | Related blogs in same cluster (1-3) | Cross-pollination within ecosystem |
| **Blog Post** | Relevant products (1-2) | User journey: info → commercial |
| **Cluster Hub** | All blogs in its cluster | Distributes hub authority to supporting content |
| **Cluster Hub** | Attribute term pages (relevant to cluster) | Connects to taxonomy layer |
| **Cluster Hub** | Products/services in cluster | Hub → commercial funnel |
| **Cluster Hub** | Related cluster hubs (1-2 max) | Cross-cluster authority for shared dimensions |
| **Product Page** | Its primary cluster hub (mandatory) | Funnels product authority to hub |
| **Product Page** | Comparison/review blog (if exists) | User journey: product → evaluation |
| **Product Page** | Related products (same cluster) | Cross-selling within ecosystem |
| **Attribute Term Page** | Its associated cluster hub(s) | Taxonomy → hub authority flow |
| **Attribute Term Page** | Products with this attribute | Term → commercial pages |
| **Attribute Term Page** | Relevant blogs | Term → informational depth |
| **Category Page** | All cluster hubs in category | Top-level → hub distribution |
| **Category Page** | Featured products | Category → conversion |
| **Homepage** | Category pages / top cluster hubs | Root authority distribution |

### 9.3 Cross-Cluster Linking Rules

Cross-cluster links connect related ecosystems but must be controlled to avoid authority dilution:

1. **Only link between hubs that share at least 1 dimension** — "Shiatsu Foot Massagers" hub can link to "Shiatsu Neck Massagers" hub (shared: technique dimension)
2. **Maximum 2 cross-cluster links per hub** — more dilutes the ecosystem's authority concentration
3. **Blogs should NOT cross-link to other clusters' hubs** — they link only to their own hub
4. **Category pages link to all their child cluster hubs** — this is hierarchical, not cross-cluster

### 9.4 Link Density Guidelines

| Page Type | Outbound Internal Links | Inbound Internal Links |
|-----------|------------------------|------------------------|
| Homepage | 5-15 | Maximum (all backlinked) |
| Category Page | 8-20 | From homepage + child hubs |
| Cluster Hub | 10-25 | From all cluster members + category |
| Blog Post | 3-8 | From hub + related blogs |
| Product Page | 3-8 | From hub + attribute terms |
| Attribute Term Page | 5-15 | From products + hub + blogs |

### 9.5 Anchor Text Strategy (Internal)

- Use descriptive, keyword-rich anchor text for internal links
- Vary anchor text naturally — don't use the exact same anchor for every link to a page
- Hub links should use the hub's primary keyword as anchor
- Product links should use product name or primary keyword
- Blog links should use the blog title or a natural description of the content

---

## 10. URL Structure Strategy

### 10.1 URL Hierarchy Reflects Site Architecture

URLs should mirror the dimensional and cluster structure:

```
/                                           ← Homepage
/[category]/                                ← Primary category (dimension 1)
/[category]/[cluster-hub-slug]/             ← Cluster hub page
/shop/[category]/[product-slug]/            ← Product page
/blog/[blog-post-slug]/                     ← Blog post
/[taxonomy-name]/[term-slug]/               ← Taxonomy term page
```

### 10.2 URL Examples (MassagerSmart)

```
/foot-massager/                             ← Category
/best-foot-massagers-for-neuropathy.../     ← Cluster hub
/shop/foot-massager/portable-ems-foot.../   ← Product
/blog/ankle-heel-foot-massagers.../         ← Blog
/massager-type/shiatsu/                     ← Attribute term
/relief-focus/neuropathy/                   ← Attribute term
```

### 10.3 URL Rules

1. **Keep URLs descriptive and keyword-rich** — they should read naturally
2. **Use hyphens, not underscores** — standard SEO practice
3. **Reflect the hierarchy** — a product URL should indicate its category
4. **Blog posts can be flat** (/blog/slug/) — they don't need category in URL since they link to hubs via internal links
5. **Taxonomy terms use their taxonomy name as prefix** — /massager-type/shiatsu/ not just /shiatsu/
6. **Avoid URL depth beyond 3 levels** — deeper URLs get less crawl priority

---

## 11. SEO Schema & Metadata

### 11.1 Schema by Page Type

| Page Type | Primary Schema | Additional Schema |
|-----------|---------------|-------------------|
| Homepage | `WebSite`, `Organization` | `SiteNavigationElement` |
| Category Page | `CollectionPage` | `BreadcrumbList`, `ItemList` |
| Cluster Hub | `WebPage` or `CollectionPage` | `FAQPage`, `ItemList`, `BreadcrumbList` |
| Blog Post | `BlogPosting` or `Article` | `FAQPage`, `HowTo`, `BreadcrumbList` |
| Product Page | `Product` | `Review`, `AggregateRating`, `FAQPage`, `BreadcrumbList` |
| Service Page | `Service` | `LocalBusiness`, `FAQPage`, `BreadcrumbList` |
| Attribute Term Page | `CollectionPage` | `FAQPage`, `BreadcrumbList` |
| Brand Page | `Article` | `BreadcrumbList` |

### 11.2 Meta Title & Description Patterns

**Cluster Hub:** "[Primary Keyword]: [Value Proposition] | [Brand]"
- Example: "Best Foot Massagers for Neuropathy: Doctor-Recommended Picks | MassagerSmart"

**Blog Post:** "[Topic]: [Angle or Promise] | [Brand]"
- Example: "How to Massage Your Neck (And When You Shouldn't) | MassagerSmart"

**Product Page:** "[Product Name] — [Key Benefit] | [Brand]"
- Example: "Portable EMS Foot Massager — Neuropathy & Nerve Relief | MassagerSmart"

**Taxonomy Term Page:** "[Term]: [Category Context] & [Value Proposition] | [Brand]"
- Example: "Shiatsu: Massage Devices & Deep Tissue Relief | MassagerSmart"

---

## 12. Authority Compounding Model

### 12.1 How SAG Compounds Authority Over Time

SAG creates a compounding authority effect that increases over time, not linearly but exponentially. This is the key strategic advantage.

**Phase 1: Foundation (Months 0-6)**
- Site structure is built according to SAG blueprint
- Cluster hubs and initial supporting content published
- Taxonomies created with basic content
- Internal linking architecture implemented
- External backlink campaign begins targeting T1-T2 pages
- **SAG impact:** 30-40% efficiency gain in backlink effectiveness vs flat site

**Phase 2: Early Authority (Months 6-12)**
- DR/DA reaches tipping point range (DR 25-35 depending on market)
- First un-linked pages begin ranking for low-KD keywords
- Google crawl frequency increases as site structure proves reliable
- New content published into existing cluster structure ranks faster
- **SAG impact:** Un-linked pages ranking 2-4 DR points earlier than expected

**Phase 3: Authority Establishment (Months 12-24)**
- 50%+ of new content ranks without direct backlinks
- Crawl times for new content drop to hours/minutes
- Cross-cluster authority flow becomes measurable
- Taxonomy term pages begin ranking independently
- **SAG impact:** 50%+ improvement in pages ranking without backlinks vs flat site at same DR

**Phase 4: Compounding Dominance (Year 2+)**
- Site can rank for competitive keywords with DR well below competitors
- New content published into SAG structure achieves page 1-3 within weeks
- Authority ratio reaches 25:1+ (un-linked pages ranking per backlinked page)
- Site becomes the topical authority for its niche in search engines
- **SAG impact:** Site operates in a fundamentally different competitive league

### 12.2 Why This Compounds

1. **Every new page strengthens every existing page.** When a new blog post is added to a cluster, it creates new internal links to the hub and other cluster content. This increases the hub's internal link count, which increases its ranking power, which increases the authority it passes back to all other cluster pages. It's a positive feedback loop.

2. **Google learns the site's authority pattern.** Over time, Google's crawler learns that this site consistently produces relevant, well-structured content in its niche. This builds a trust signal that goes beyond raw DR — it's topical trust. New pages inherit this trust immediately upon publication.

3. **Taxonomy term pages accumulate authority passively.** As more products and blog posts link to taxonomy terms, those term pages gain ranking power without any direct SEO effort. A "Neuropathy Relief" term page gets stronger every time a new neuropathy-related blog post or product is published.

4. **Cross-cluster links compound.** As individual clusters strengthen, the controlled cross-cluster links become more valuable, strengthening adjacent clusters. This creates a network effect where the whole site becomes stronger than the sum of its parts.

### 12.3 Real-World Evidence

An SAG-structured site in a competitive niche, after 8 years of consistent implementation:
- DR 29 (modest by competitive standards)
- Dominates country-level rankings for primary category keyword
- 1000+ pages on page 1 top 5 positions without any direct backlinks
- Only 30-40 pages have external backlinks
- New blog posts crawled within 10 minutes of publication
- New content ranks for target keywords within days

This represents a 25:1+ ratio of un-linked pages ranking per backlinked page — a ratio that flat-structure sites at DR 50+ cannot achieve.

---

## 13. SAG Application Workflow

### 13.1 Complete Process (9 Steps)

**Step 1: Collect Keywords**
Gather all relevant search queries from Ahrefs, SEMrush, Google Search Console, competitor analysis, and internal product/service data. Build a comprehensive keyword pool across all intents.

**Step 2: Discover Dimensions**
Analyze the keyword set to identify the natural dimensional axes. What are the recurring facets that differentiate keywords from each other? Group modifiers into dimensions. Validate dimensional independence.

**Step 3: Form Clusters**
Group keywords at meaningful 2-3 dimensional intersections. Each cluster must pass the ecosystem test: hub potential, user journey viability, minimum 5 keywords, semantic coherence.

**Step 4: Classify Cluster Types**
Determine each cluster's type (product, condition, feature, brand, informational, comparison, attribute). This determines the content type, page template, and position in the site hierarchy.

**Step 5: Derive Taxonomies**
Convert dimensional axes into site taxonomies. Define primary (categories), secondary (custom taxonomies), and tertiary (tags) taxonomy levels. Enumerate all term values.

**Step 6: Map Site Architecture**
Build the complete page hierarchy: homepage → categories → cluster hubs → supporting content → products/services. Assign every page to its cluster. Define URL structure.

**Step 7: Define Internal Linking Map**
Specify all internal link relationships following SAG linking rules. Hub → blogs, blogs → hub, hub → products, products → hub, hub ↔ attribute terms, controlled cross-cluster links.

**Step 8: Generate Content Plan**
For each cluster, define: hub page content brief, supporting blog topics, taxonomy term page content needs, product/service page requirements. Prioritize by cluster importance and commercial value.

**Step 9: Execute & Scale**
Publish hub pages first (highest priority). Add supporting blogs for depth. Create taxonomy term pages. Add products/services. Implement internal links. As new keywords emerge, slot them into existing clusters or form new ones. Launch external backlink campaign targeting hub pages and homepage.

### 13.2 Execution Priority Order

1. **Homepage** — foundation of authority
2. **Category pages** — primary navigation structure
3. **Cluster hub pages** — authority anchors (prioritize by commercial value)
4. **Supporting blog posts** — 2-3 per cluster initially
5. **Taxonomy term pages** — for primary/secondary taxonomy terms
6. **Product/service pages** — commercial conversion pages
7. **Additional blog posts** — expand depth per cluster
8. **Brand comparison pages** — if brand cluster exists
9. **Cross-cluster content** — content that bridges related clusters

---

## 14. Niche Adaptation Framework

### 14.1 E-Commerce Site

**Typical Dimensions:** Product type, feature, condition/use-case, brand, price range, audience
**Primary Cluster Types:** Product clusters, condition clusters, feature clusters, brand clusters
**Key Architecture:** Categories → cluster hubs → product pages + supporting blogs
**Special Consideration:** Product attribute taxonomies are critical — they become filter/faceted navigation AND SEO landing pages

### 14.2 SaaS / Technology Company

**Typical Dimensions:** Feature category, use case, industry vertical, company size, integration type
**Primary Cluster Types:** Feature clusters, use-case clusters, comparison clusters, informational clusters
**Key Architecture:** Solution pages → feature hubs → case studies + blog posts
**Special Consideration:** Comparison clusters (vs competitors) are often the highest-value commercial content

### 14.3 Services Business

**Typical Dimensions:** Service type, industry served, location, methodology, outcome
**Primary Cluster Types:** Service clusters, industry clusters, informational clusters
**Key Architecture:** Service pages → industry-specific hubs → case studies + educational content
**Special Consideration:** Location dimension creates geo-specific cluster variants (same service, different cities)

### 14.4 Content / Media Site

**Typical Dimensions:** Subject area, reader level, content format, timeliness
**Primary Cluster Types:** Topic clusters, informational clusters, comparison clusters
**Key Architecture:** Topic categories → pillar content hubs → supporting articles
**Special Consideration:** Evergreen vs trending content balance — SAG structure should prioritize evergreen clusters with trending content as supporting pieces

### 14.5 Local Business

**Typical Dimensions:** Service type, location, audience, urgency
**Primary Cluster Types:** Service clusters, location clusters, problem clusters
**Key Architecture:** Service pages → location-specific variants → local educational content
**Special Consideration:** Location × service dimensional intersections create highly targeted local landing pages

---

## 15. Validation & Quality Criteria

### 15.1 Cluster Validation Checklist

Before finalizing any cluster, verify:

- [ ] **Hub Potential:** Cluster has 1 clear anchor keyword suitable for a hub page
- [ ] **Dimensional Overlap:** Keywords share 2+ semantic dimensions
- [ ] **User Journey:** Natural navigation paths exist within the cluster
- [ ] **Attribute Dimensions:** Recurring modifiers can become filters/taxonomies
- [ ] **Ecosystem Completeness:** Supports 1 hub + 3-10 supporting content pieces
- [ ] **Semantic Coherence:** Keywords genuinely belong together (not just word overlap)
- [ ] **No Duplication:** Each keyword appears in exactly one cluster
- [ ] **Minimum Size:** Cluster has 5+ keywords
- [ ] **Maximum Size:** Cluster has ≤20 keywords (split if larger)
- [ ] **Hub Naming:** Cluster name is natural, SEO-relevant, and specific

### 15.2 Site Architecture Validation

- [ ] **No orphan pages:** Every page belongs to at least one cluster
- [ ] **Complete cluster ecosystems:** Every cluster has hub + supporting content + commercial pages
- [ ] **Taxonomy coverage:** All primary/secondary dimensions have taxonomy term pages
- [ ] **Internal linking implemented:** All SAG linking rules are active
- [ ] **URL structure reflects hierarchy:** URLs follow the defined pattern
- [ ] **Schema markup applied:** Every page type has appropriate structured data
- [ ] **Cross-cluster links are controlled:** Maximum 2 per hub, dimension-justified

### 15.3 Content Quality Criteria

- [ ] **Hub pages are comprehensive** (2,000+ words, multiple sections, FAQs)
- [ ] **Blog posts provide genuine depth** (1,000+ words, unique angle)
- [ ] **Product pages include dimensional attributes** (taxonomy-driven specification sections)
- [ ] **Taxonomy term pages have rich content** (not just product lists)
- [ ] **All content targets specific keywords from the cluster**
- [ ] **Internal links use descriptive, varied anchor text**

---

## 16. Anti-Patterns

### 16.1 Clustering Anti-Patterns

| Anti-Pattern | Why It's Wrong | Correct Approach |
|-------------|---------------|-----------------|
| **Single-word grouping** ("all 'heated' keywords together") | Feature alone isn't an ecosystem | Cluster by dimensional intersection (heated + product type + use-case) |
| **Forced categorization** ("must fit everything") | Creates weak clusters | Exclude keywords that don't fit strong ecosystems |
| **Shallow connections** (shared word, no semantic bond) | "Back massager" ≠ "back support pillow" | Verify 2+ dimensional intersections |
| **Traditional intent labels as clusters** ("informational cluster") | Intent is a property, not a dimension | Cluster by topic ecosystem, classify intent within |
| **Category-first thinking** ("put all foot stuff together") | Categories are outputs, not inputs | Let keyword dimensions determine categories |
| **Word-matching only** ("all 'massage' keywords") | Ignores semantic reality | Analyze dimensional intersections |
| **Duplicate keyword assignment** (keyword in 2 clusters) | Breaks SAG structure | Assign each keyword to exactly one cluster |

### 16.2 Architecture Anti-Patterns

| Anti-Pattern | Why It's Wrong | Correct Approach |
|-------------|---------------|-----------------|
| **Random blog posts** (not assigned to clusters) | Orphan content wastes authority | Every blog supports a specific cluster |
| **Flat taxonomy** (one level of categories, no custom taxonomies) | Misses dimensional depth | Derive multi-level taxonomies from dimensions |
| **Unlimited cross-cluster linking** | Dilutes authority concentration | Maximum 2 cross-cluster links per hub |
| **Empty taxonomy term pages** (just product lists) | Wastes SEO opportunity | Rich content on every term page |
| **No internal linking plan** | Authority distribution is random | Implement SAG linking rules systematically |
| **Products not assigned to clusters** | Products float outside ecosystems | Every product belongs to a primary cluster |

---

## 17. Real-World Reference: MassagerSmart

### 17.1 Dimensional Axes

| Dimension | Taxonomy | Values |
|-----------|----------|--------|
| Body Area | `massage_area` | Neck, Foot, Face, Back, Eye, Scalp |
| Device Type | `massager_type` | Shiatsu, EMS Pad/Mat, Percussion, Vibration, Roller, LED, Microcurrent |
| Technique | `massage_technique` | Electrical Stimulation, Deep Tissue, Kneading, Vibration |
| Condition | `relief_focus` | Neuropathy Relief, Plantar Fasciitis, Circulation, Pain Relief, Skin Tightening, Lymphatic Drainage |
| Therapy | `therapy_method` | NMES Therapy, TENS Therapy, Shiatsu, Heat Therapy |
| Brand | `brand` | Renpho, Nekteck, TheraFace, Homedics, Naipo, SKG, Hugterra |

### 17.2 Cluster Map (Foot Massagers)

| Cluster Name | Type | Hub Page | Supporting Content |
|-------------|------|----------|-------------------|
| Neuropathy & Nerve Pain Massagers | Condition Cluster | /best-foot-massagers-for-neuropathy-.../ | Doctor recommendations, EMS for neuropathy blog, nerve pain relief guide |
| Plantar Fasciitis Foot Massagers | Condition Cluster | /best-foot-massagers-for-plantar-fasciitis-.../ | Exercise + massage blog, night splint comparison |
| Heated, Electric & EMS Massagers | Feature Cluster | /top-heated-electric-ems-foot-massagers-.../ | EMS vs TENS blog, heated massager safety guide |
| Shiatsu Foot Massagers | Product Type Cluster | /best-shiatsu-foot-massagers-.../ | Deep kneading technique blog, shiatsu at home guide |
| Leg & Calf Foot Massagers | Product Type Cluster | /best-leg-calf-foot-massagers-.../ | Calf pain relief blog, leg circulation guide |
| Foot Massager Brands Compared | Brand Cluster | /top-foot-massager-brands-compared-.../ | Individual brand reviews |
| General Product Category | Intent Cluster | /foot-massager/ (category page) | Best of year roundup, buying guide |

### 17.3 Blog Clusters (Foot Massagers)

| Blog Cluster | Type | URL Pattern | Purpose |
|-------------|------|-------------|---------|
| Heel & Ankle Relief | Use-Case Cluster | /blog/ankle-heel-foot-massagers-.../ | Supports neuropathy and general foot clusters |
| Foot Message vs Massage | Intent Cluster | /blog/foot-message-vs-massage-.../ | Captures misspelling/confusion traffic |
| Spa & Water Massagers | Function Cluster | /blog/best-foot-spa-massagers-.../ | Captures spa-related search intent |
| Foot Machine Types | Comparison Cluster | /blog/foot-massager-vs-massage-machine-.../ | Comparison/buying decision content |

### 17.4 Product Page Example: Dimensional Attributes

**Product:** Portable EMS Foot Massager for Neuropathy & Nerve Relief

**Specification Section (taxonomy-driven):**

| Attribute | Value (linked to term page) |
|-----------|---------------------------|
| Massager Category | EMS Foot Massagers |
| Massage Area | Feet |
| Massager Type | EMS Pad / Mat |
| Massage Technique | Electrical Stimulation |
| Relief Focus | Foot Circulation, Neuropathy Relief |
| Therapy Method | Foot Circulation, Neuropathy Relief, NMES Therapy, TENS Therapy |

Each value in the specification section is a link to its taxonomy term page, creating bidirectional authority flow between product pages and term pages.

---

## 18. Appendix: Clustering Prompt Specification

This section provides the AI prompt specification for automated keyword clustering following SAG methodology. This is used by AI systems (including IGNY8) to perform SAG-compliant clustering.

### 18.1 AI Clustering Prompt Role

The AI acts as a semantic strategist building topic ecosystems using SAG methodology. It analyzes keywords and groups them into mini-ecosystems where each cluster represents a complete, self-contained topic authority area.

### 18.2 Input Format

```json
{
  "keywords": ["keyword 1", "keyword 2", "keyword 3", "..."]
}
```

Keywords may include volume and difficulty metrics, but clustering decisions should be based on semantic analysis, not volume/difficulty data.

### 18.3 Output Format

```json
{
  "clusters": [
    {
      "name": "Natural, SEO-relevant cluster name representing the root topic",
      "description": "2-3 sentences explaining: (1) what semantic dimensions bind these keywords, (2) what user journey or problem space this cluster addresses",
      "keywords": ["keyword 1", "keyword 2", "keyword 3"]
    }
  ]
}
```

### 18.4 Clustering Algorithm

1. **Identify dimensional axes** in the keyword set
2. **Find 2+ dimensional intersections** between keywords
3. **Group keywords at meaningful intersections** forming ecosystems
4. **Identify hub anchor keyword** for each cluster
5. **Extract taxonomy-worthy attributes** from each cluster
6. **Resolve conflicts** (keywords fitting multiple clusters → assign to strongest fit)
7. **Validate** each cluster against formation rules
8. **Exclude** keywords that don't fit any strong ecosystem

### 18.5 Naming Guidelines

**Good names:** Descriptive, SEO-relevant, represent the root topic
- "Heated Shiatsu Back Massagers for Pain Relief"
- "Organic Cotton Bedding for Summer"
- "Vitamin C Serums for Skin Brightening"

**Bad names:** Too broad, meaningless, just a feature, or meta-descriptive
- "Massagers" / "Products" / "Heated" / "Queries about safety"

### 18.6 Description Template

"This cluster covers [TOPIC] focused on [PROBLEM/USE-CASE] for [AUDIENCE]. Keywords share dimensions of [DIMENSION 1] and [DIMENSION 2], forming a natural ecosystem for users researching [USER JOURNEY]."

### 18.7 Quality Rules

- Minimum 5 keywords per cluster
- Maximum 20 keywords per cluster (optimal: 7-15)
- No keyword duplication across clusters
- Each cluster must have clear hub potential
- Exclude keywords that don't fit strong ecosystems
- 5 strong clusters > 15 weak clusters

---

*This document is the definitive reference for the Semantic Authority Grid methodology. All SAG-related implementation, content planning, and site architecture decisions should reference this document as the source of truth.*