information-architecture

SKILL.md

Information Architecture

Design findable, usable content structures using the four IA systems: Organization, Labeling, Navigation, and Search.

When to Use

  • Designing website or application structure
  • Creating taxonomies and categorization schemes
  • Planning navigation systems and sitemaps
  • Modeling content for headless CMS
  • Conducting card sorting or tree testing
  • Optimizing findability and information scent

When NOT to Use

  • Visual design (colors, typography, spacing)
  • Interaction design (animations, micro-interactions)
  • Frontend implementation (HTML/CSS/JS coding)
  • Database schema design (use data modeling instead)

Quick Start

  1. Audit: Inventory existing content and user needs
  2. Organize: Choose organization scheme (LATCH)
  3. Label: Define terminology and controlled vocabulary
  4. Navigate: Design navigation patterns
  5. Validate: Test with tree testing or card sorting

The Four IA Systems

flowchart TB
    subgraph IA["Information Architecture"]
        direction TB
        OS["Organization Systems<br/>How we categorize"]
        LS["Labeling Systems<br/>How we represent"]
        NS["Navigation Systems<br/>How we browse"]
        SS["Search Systems<br/>How we query"]
    end

    OS --> |"defines structure for"| NS
    LS --> |"provides vocabulary for"| SS
    OS --> |"creates categories for"| LS
    NS --> |"supplements"| SS

    Users((Users)) --> NS
    Users --> SS
    Content[(Content)] --> OS
    Content --> LS
System Purpose Key Deliverables
Organization How content is categorized Taxonomies, hierarchies, facets
Labeling How content is named Controlled vocabulary, naming conventions
Navigation How users browse Sitemaps, menus, breadcrumbs
Search How users query Filters, facets, autocomplete

Core Procedure

Phase 1: Research and Discovery

Goal: Understand users, content, and context.

  • Conduct stakeholder interviews (business goals, constraints)
  • Perform content audit (inventory, gap analysis)
  • Research user mental models (interviews, analytics)
  • Analyze competitors and analogous systems

Checkpoint: Document findings in research summary.

Phase 2: Organization Design

Goal: Define how content will be categorized.

  1. Choose organization scheme (LATCH):

    • Location: Geographic/spatial
    • Alphabet: A-Z ordering
    • Time: Chronological
    • Category: Topical grouping
    • Hierarchy: Ranked by importance
  2. Design taxonomy:

    • Flat (single-level terms)
    • Hierarchical (parent-child tree)
    • Faceted (multiple dimensions)
    • Polyhierarchical (multiple parents)
  3. Define metadata schema:

    • Required fields per content type
    • Controlled vocabulary terms
    • Relationship types

Checkpoint: Taxonomy documented with 3+ levels and example content assigned.

Phase 3: Labeling Design

Goal: Create consistent, user-friendly terminology.

  • Extract user vocabulary from research
  • Define controlled vocabulary (preferred terms)
  • Map synonyms and aliases
  • Document naming conventions
  • Create labeling style guide

Checkpoint: Labeling guide with 10+ key terms defined.

Phase 4: Navigation Design

Goal: Design how users will move through content.

flowchart LR
    subgraph Global["Global Navigation"]
        G1[Main Menu]
        G2[Header Links]
        G3[Footer Links]
    end

    subgraph Local["Local Navigation"]
        L1[Sidebar Menu]
        L2[Section Tabs]
    end

    subgraph Contextual["Contextual"]
        C1[Inline Links]
        C2[Related Content]
    end

    subgraph Supplemental["Supplemental"]
        S1[Breadcrumbs]
        S2[Site Map]
    end
Pattern Use When
Global Nav Site-wide access to main sections
Local Nav Section-specific navigation
Breadcrumbs Deep hierarchies (3+ levels)
Faceted Nav Product catalogs, search results
Mega Menu Large sites with many categories
  • Create sitemap diagram
  • Define navigation patterns per section
  • Specify breadcrumb rules
  • Plan search/filter integration

Checkpoint: Sitemap with 3-4 levels max, navigation patterns specified.

Phase 5: Validation

Goal: Test IA against real user behavior.

Card Sorting (for organization):

  1. Create cards for content items (30-60)
  2. Recruit participants (15-20)
  3. Run open or closed sort
  4. Analyze similarity matrix
  5. Refine taxonomy based on results

Tree Testing (for navigation):

  1. Create text-only tree structure
  2. Define findability tasks (10-15)
  3. Recruit participants (50+)
  4. Measure success rate and directness
  5. Identify problem areas

Checkpoint: Validation results documented, IA refined based on findings.


Definition of Done

  • Content audit completed with gap analysis
  • Taxonomy documented with controlled vocabulary
  • Sitemap diagram created (max 4 levels deep)
  • Navigation patterns specified for all sections
  • Labeling guide with naming conventions
  • Accessibility requirements addressed (WCAG 2.4)
  • Validation method executed (card sort or tree test)

Guardrails

Do

  • Keep hierarchies shallow (3-4 levels max)
  • Use user vocabulary in labels (not internal jargon)
  • Provide multiple paths to content
  • Design for both browsing and searching
  • Consider accessibility from the start

Do Not

  • Create deep nested structures (>5 levels)
  • Use ambiguous or overlapping categories
  • Ignore mental models revealed in research
  • Skip validation before implementation
  • Mix organization schemes inconsistently

Accessibility Requirements

Per WCAG 2.4 (Navigable):

  • Provide skip navigation links (2.4.1)
  • Use descriptive page titles (2.4.2)
  • Ensure logical focus order (2.4.3)
  • Write descriptive link text (2.4.4)
  • Offer multiple ways to find content (2.4.5)

See references/accessibility.md for full WCAG navigation checklist.


Deliverables

Artifact Format Purpose
Sitemap Mermaid/diagram Visual hierarchy of pages
Taxonomy YAML/table Controlled vocabulary structure
Navigation Spec Markdown Pattern definitions per section
Labeling Guide Markdown Naming conventions and vocabulary
Content Model YAML/JSON Fields and relationships
Validation Report Markdown Test results and recommendations

Content Modeling (Headless CMS)

For headless/omnichannel content:

flowchart TB
    subgraph ContentModel["Content Model"]
        CT[Content Types]
        AT[Attributes/Fields]
        REL[Relationships]
        META[Metadata]
    end

    subgraph Channels["Delivery Channels"]
        WEB[Website]
        MOBILE[Mobile App]
        VOICE[Voice Assistant]
        API[API/Integration]
    end

    ContentModel --> |"API-first"| Channels

Principles:

  1. Separate content from presentation
  2. Use modular, reusable components
  3. Leverage reference fields (don't duplicate)
  4. Plan for internationalization
  5. Include SEO metadata as fields

See references/content-modeling.md for detailed patterns.


Tools

Sitemaps & Diagramming: Figma, Miro, Mermaid, draw.io Card Sorting: Optimal Workshop, UXtweak, Maze Tree Testing: Treejack, UXtweak Taxonomy Management: PoolParty, Neo4j, Notion Headless CMS: Contentful, Strapi, Sanity, Hygraph


References


Failure Modes & Recovery

Failure Symptom Recovery
Users can't find content High bounce, poor task completion Validate with tree testing, simplify hierarchy
Inconsistent labels User confusion, support tickets Create labeling guide, audit existing labels
Too many categories Paralysis of choice Apply 7+/-2 rule, use faceted navigation
Deep nesting Lost users, poor breadcrumbs Flatten hierarchy, add cross-links
Duplicate content paths SEO issues, user confusion Canonical paths, redirect strategy
Weekly Installs
1
First Seen
6 days ago
Installed on
mcpjam1
github-copilot1
junie1
windsurf1
zencoder1
crush1