News content management systems face a critical architectural challenge that creates severe duplicate content issues: the same article becomes accessible through multiple distinct URLs. Unlike intentional content duplication, this problem stems from CMS default behavior where category-based URL structures combined with insufficient validation logic allow identical content to appear at numerous URL paths.

This article examines the technical mechanisms behind news CMS multiple URL generation, analyzes real-world implementation cases, explains the SEO impact of this architectural flaw, and provides comprehensive solutions based on actual news platform optimization experience. Understanding this issue is essential for news publishers experiencing crawl budget waste, ranking dilution, and indexing problems.

What Is the News CMS Multiple URL Problem?

The news CMS multiple URL problem occurs when content management systems generate and allow access to the same article through different category-based URL paths without enforcing canonical URL validation. This creates a situation where one unique article exists at multiple distinct URLs, each appearing to search engines as separate pages containing identical content.

Definition and Technical Context

Multiple URL generation refers to CMS behavior where a single article identified by a unique code becomes accessible through any category path structure, regardless of whether that category was actually assigned to the article. According to Google Search Central, this violates the principle of one canonical URL per unique content piece and creates significant technical SEO challenges.

The Unique Code Vulnerability

News CMS platforms typically assign each article a unique identifier (code) that becomes part of the URL structure. However, many systems fail to validate that the category path preceding this unique code matches the article's actual categorization. This validation gap allows any article code to work with any category path.

Example of the problem:

Article unique code: amdppietgid7p
Actual category: Feature > Campus
Correct URL: /feature/campus/amdppietgid7p
Problem: The same code works with wrong categories:
❌ /bangladesh/chattogram/amdppietgid7p (returns 200 OK)
❌ /politics/national/amdppietgid7p (returns 200 OK)
❌ /sports/cricket/amdppietgid7p (returns 200 OK)

All these URLs resolve successfully and display identical article content, creating multiple competing URLs for one article. This differs from keyword cannibalization where different pages target the same keywords—here, identical content exists at multiple URLs.

How News CMS Generates Multiple URLs

Understanding the technical mechanism behind multiple URL generation requires examining CMS architecture, routing logic, and database query patterns.

URL Routing Without Category Validation

Most news CMS platforms use pattern-based URL routing that extracts the unique article code from the URL and queries the database directly for that code, ignoring the category path components.

Typical flawed routing logic:

// Simplified example of problematic code
function displayArticle($url) {
    // URL format: /category/subcategory/unique-code
    $parts = explode('/', $url);
    $unique_code = end($parts); // Gets last segment
    
    // Problem: Only uses unique code, ignores category
    $article = database_query("SELECT * FROM articles WHERE code = '$unique_code'");
    
    if ($article) {
        display_article($article); // Shows article regardless of URL path
    }
}

This approach retrieves and displays articles based solely on the unique code, making category paths meaningless. The URL appears to have hierarchical structure (/category/subcategory/code) but the system only processes the final code segment.

Database Schema Contributing to the Problem

News CMS databases typically store articles with category relationships but don't enforce URL-to-category validation at the application layer.

Common database structure:

Table: articles
- id (primary key)
- unique_code (indexed)
- title
- content
- primary_category_id
- created_at
Table: article_categories (many-to-many relationship)
- article_id
- category_id
- is_primary (boolean)
Table: categories
- id
- name
- slug
- parent_id

The article's true categorization exists in the database through article_categories relationships, but the URL routing doesn't validate requested URL paths against these stored relationships. This architectural separation between URL routing and category validation creates the multiple URL problem.

Category-Based URL Generation vs. Validation

News platforms often implement robust URL generation during article publishing but fail to implement corresponding validation during article retrieval.

Publishing phase (URL generation):

// During article creation - WORKS CORRECTLY
function generateCanonicalURL($article) {
    $primary_category = get_primary_category($article);
    $subcategory = get_primary_subcategory($article);
    
    $url = "/{$primary_category}/{$subcategory}/{$article->unique_code}";
    
    // Store in database
    save_canonical_url($article->id, $url);
    
    return $url;
}

Retrieval phase (URL validation) - MISSING:

// What SHOULD happen but usually DOESN'T
function validateAndDisplayArticle($requested_url) {
    $unique_code = extract_code($requested_url);
    $article = get_article($unique_code);
    
    // CRITICAL: Validate requested URL matches stored canonical
    if ($requested_url !== $article->canonical_url) {
        redirect_301($article->canonical_url); // Redirect to correct URL
    }
    
    display_article($article);
}

The absence of this validation step allows any category path to successfully display the article, creating unlimited URL variations for single articles.

Multi-Category Assignment Complexity

News articles often legitimately belong to multiple categories for organizational purposes. An article about "political corruption in Chattogram district" might reasonably be assigned to:

• Geographic category: All Districts > Chattogram
• Topic category: Politics > Corruption
• Theme category: Investigative Journalism

However, without proper URL canonicalization, this creates three potential URL paths, each resolving successfully:

/all-districts/chattogram/article-code
/politics/corruption/article-code
/investigative/journalism/article-code

The CMS needs rules determining which category becomes the canonical URL while allowing the article to appear in other category listings without generating additional accessible URLs. Most systems lack this priority logic, resulting in all category combinations creating working URLs.

Real-World Case Study: Daily Amardesh Implementation

A comprehensive examination of a real Bangladeshi news website reveals specific technical implementation problems and their resolution.

Discovery of the Problem

Issue identification: During a technical SEO audit, investigation revealed that the unique article code amdppietgid7p (a feature article about a book fair at Dhaka University) was accessible through multiple incorrect category paths:

Correct URL structure:

Category: ফিচার (Feature)
Subcategory: ক্যাম্পাস (Campus)
URL: dailyamardesh.com/feature/campus/amdppietgid7p

Discovered duplicate URLs:

❌ dailyamardesh.com/bangladesh/chattogram/amdppietgid7p
❌ dailyamardesh.com/politics/national/amdppietgid7p
❌ dailyamardesh.com/sports/cricket/amdppietgid7p

Testing confirmed that the unique code worked with any arbitrary category path. This indicated complete absence of URL validation logic in the CMS routing system.

Impact Analysis

SEO consequences observed:

Crawl inefficiency: Google Search Console data showed Googlebot discovering multiple URLs for single articles, wasting crawl budget on duplicates.

Index bloat: Coverage reports indicated hundreds of duplicate URLs indexed, many showing "Duplicate, Google chose different canonical than user" warnings.

Ranking fragmentation: Articles showed inconsistent rankings as backlinks distributed across multiple URL variations rather than consolidating on one authoritative version.

Analytics fragmentation: GA4 data split pageview metrics across URL variations, making true article performance difficult to measure.

Category Priority Solution

To resolve the issue while maintaining editorial flexibility, a category priority system was implemented:

Priority rule hierarchy:

Level 1 - Geographic Priority (Highest): Articles from district/city correspondents receive geographic category as primary, regardless of topic.

Rule: IF article has correspondent location
      THEN primary_category = "All Districts"
           subcategory = correspondent_location
           
Example: Political news from Chattogram correspondent
         Primary URL: /all-districts/chattogram/article-code
         Secondary categories: Politics (for listing only)

Level 2 - Topic Priority (Secondary): Articles without geographic source use main topic category as primary.

Rule: IF article has no correspondent location
      THEN primary_category = main_topic_category
      
Example: National political analysis from Dhaka office
         Primary URL: /politics/national/article-code

This creates predictable, single canonical URLs while maintaining multi-category classification for editorial organization and site navigation.

Technical Implementation

Database schema updates:

-- Add canonical URL storage
ALTER TABLE articles ADD COLUMN canonical_url VARCHAR(255);
ALTER TABLE articles ADD INDEX idx_canonical_url (canonical_url);
-- Add category priority
ALTER TABLE article_categories ADD COLUMN priority_level INT DEFAULT 0;
-- priority_level: 1 = primary, 2 = secondary, 3 = tertiary

URL validation function:

function validateArticleURL($requested_url) {
    // Extract unique code from requested URL
    $unique_code = basename($requested_url);
    
    // Get article from database
    $article = Article::where('unique_code', $unique_code)->first();
    
    if (!$article) {
        return response()->view('errors.404', [], 404);
    }
    
    // Get stored canonical URL
    $canonical_url = $article->canonical_url;
    
    // Compare requested URL with canonical
    if ($requested_url !== $canonical_url) {
        // Wrong category path - redirect to canonical
        return redirect($canonical_url, 301);
    }
    
    // Correct URL - display article
    return view('article', ['article' => $article]);
}

Category display logic:

// Primary category page shows only primary articles
function getPrimaryCategoryArticles($category_id) {
    return Article::whereHas('categories', function($query) use ($category_id) {
        $query->where('category_id', $category_id)
              ->where('priority_level', 1); // Primary only
    })->get();
}
// Secondary category pages show all articles
function getSecondaryCategoryArticles($category_id) {
    return Article::whereHas('categories', function($query) use ($category_id) {
        $query->where('category_id', $category_id); // Any priority
    })->get();
}

This ensures articles appear in multiple category listings but link only to their canonical URL.

SEO Impact of Multiple URLs

The multiple URL problem creates measurable negative effects across all aspects of search engine optimization.

Crawl Budget Waste

Google crawl budget represents the number of pages Googlebot will crawl within a given timeframe. Multiple URLs for identical content directly waste this limited resource.

Impact calculation example:

News site metrics:
- 200 articles published daily
- Average 3 category paths per article
- 600 URLs generated daily for 200 unique articles
- 400 URLs (67%) are duplicates wasting crawl budget
Monthly impact:
- 12,000 duplicate URLs crawled
- Time spent crawling duplicates instead of new content
- Reduced discovery speed for genuinely new articles

For high-volume news publishers, this multiplication factor significantly impairs Googlebot's ability to discover and index new content promptly, affecting time-sensitive news ranking.

Link Equity Dilution

When external websites link to different URL variations of the same article, the ranking value distributes across multiple URLs instead of consolidating.

Link distribution scenario:

Article: "Major Policy Announcement"
Total backlinks: 50
Link distribution:
- /politics/national/article-code: 22 backlinks
- /all-districts/dhaka/article-code: 18 backlinks
- /government/policy/article-code: 10 backlinks
Result: Three URLs each with partial authority instead of one URL with full 50-backlink authority

This fragmentation directly impacts ranking potential, as search engines view each URL separately rather than consolidating all signals onto one authoritative version.

Index Coverage Problems

Google Search Console coverage reports show multiple issues stemming from duplicate URLs:

Common coverage report messages:

• "Duplicate, Google chose different canonical than user" - indicates Google selecting a different URL than the one marked canonical
• "Duplicate, submitted URL not selected as canonical" - appears when sitemaps include non-canonical variations
• "Discovered - currently not indexed" - may occur when index quota is exhausted by duplicates

These issues prevent efficient indexing and can lead to deindexing problems where important content fails to appear in search results.

User Experience Confusion

Multiple URLs for identical content create user-facing problems:

Social sharing fragmentation: Different users share different URL variations, splitting social signals and engagement metrics:

Facebook shares: /politics/article-code (150 shares)
Twitter shares: /all-districts/dhaka/article-code (200 shares)
LinkedIn shares: /government/article-code (80 shares)
Total engagement: 430 shares fragmented across 3 URLs

Broken bookmark consistency: Users bookmarking different URL variations encounter inconsistent experiences if one variation later redirects or becomes inaccessible.

Analytics complexity: Traffic analysis requires URL consolidation to understand true article performance, complicating reporting and decision-making.

Technical Solutions and Implementation

Addressing multiple URL generation requires implementing validation logic, canonical tags, and 301 redirects.

Solution 1: URL Validation and 301 Redirects

The most effective solution enforces canonical URLs through validation and automatic redirection.

Implementation approach:

Step 1: Store canonical URL during publishing

// During article creation
function publishArticle($article_data) {
    // Apply category priority rules
    $primary_category = determinePrimaryCategory($article_data);
    
    // Generate canonical URL
    $canonical_url = generateCanonicalURL($article_data, $primary_category);
    
    // Store in database
    $article = Article::create([
        'unique_code' => $article_data['code'],
        'title' => $article_data['title'],
        'content' => $article_data['content'],
        'canonical_url' => $canonical_url,
        // ... other fields
    ]);
    
    // Save category relationships
    saveCategoryRelationships($article, $article_data['categories']);
    
    return $article;
}

Step 2: Validate on every request

// Article display controller
public function show($full_url_path) {
    // Extract unique code
    $unique_code = basename($full_url_path);
    
    // Get article
    $article = Article::where('unique_code', $unique_code)->firstOrFail();
    
    // Validate URL matches canonical
    $canonical_path = parse_url($article->canonical_url, PHP_URL_PATH);
    $requested_path = '/' . trim($full_url_path, '/');
    
    if ($requested_path !== $canonical_path) {
        // Redirect to canonical with 301
        return redirect($article->canonical_url, 301);
    }
    
    // Display article
    return view('article.show', compact('article'));
}

Step 3: Update robots.txt and sitemaps

<!-- Sitemap includes only canonical URLs -->
<?xml version="1.0" encoding="UTF-8"?>
<urlset xmlns="http://www.sitemaps.org/schemas/sitemap/0.9">
  <url>
    <loc>https://newssite.com/politics/national/article-code-1</loc>
    <lastmod>2026-01-12</lastmod>
    <priority>0.8</priority>
  </url>
  <!-- Only canonical URLs included -->
</urlset>

Solution 2: Canonical Tag Implementation

For situations where redirects aren't feasible, canonical tags inform search engines of the preferred URL.

Implementation in article template:

<!DOCTYPE html>
<html>
<head>
    <meta charset="UTF-8">
    <title><?php echo $article->title; ?></title>
    
    <!-- Self-referencing canonical on canonical URL -->
    <!-- Cross-domain canonical on non-canonical URLs -->
    <link rel="canonical" href="<?php echo $article->canonical_url; ?>" />
    
    <!-- Additional meta tags -->
    <meta name="description" content="<?php echo $article->meta_description; ?>">
</head>
<body>
    <!-- Article content -->
</body>
</html>

Important considerations:

Canonical tags are hints, not directives. Google may ignore them if:

• Multiple pages claim different canonicals for the same content
• Canonical points to significantly different content
• Redirect chains involve the canonical URL
• Technical implementation errors exist

301 redirects are stronger signals and should be preferred when possible.

Solution 3: CMS Configuration Changes

Prevent multiple URL generation at the CMS level through architectural modifications.

Option A: Flat URL structure

Eliminate category paths from article URLs entirely:

Instead of: /category/subcategory/article-code
Use: /article-code or /article/article-code

Benefits:

• Eliminates category-based duplication entirely
• Simplifies URL management
• Reduces technical complexity

Trade-offs:

• Loses URL-based category context
• May reduce topical relevance signals
• Requires site-wide URL structure change

Option B: Primary category enforcement

Restrict articles to single primary category at the editorial workflow level:

// In article creation form
class ArticleForm {
    public function validateCategorySelection($categories) {
        if (count($categories) > 1) {
            throw new ValidationException(
                'Article must have exactly ONE primary category. ' .
                'Use tags for additional classification.'
            );
        }
    }
}

This prevents multi-category assignment that creates multiple URL possibilities, though it limits editorial flexibility.

Option C: Virtual categories for navigation

Use categories for site navigation but not URL generation:

// Articles have single URL based on unique code
// Categories are pure database relationships for filtering
function getCategoryArticles($category_id) {
    return Article::whereHas('categories', function($query) use ($category_id) {
        $query->where('category_id', $category_id);
    })->get();
}
// All articles link to: /article/unique-code
// No category in URL

This maintains multi-category classification for editorial purposes while enforcing single URL per article.

Prevention Strategies for New Implementations

News organizations building or migrating CMS platforms can prevent multiple URL problems from inception.

URL Architecture Design Principles

Principle 1: Single canonical URL per article

Every article must have exactly one canonical URL, determined during publishing and enforced during retrieval.

Principle 2: URL-to-content validation

Every URL request must validate that the requested path matches the article's stored canonical URL before displaying content.

Principle 3: Category-independent article identification

Article unique codes should be globally unique identifiers not dependent on category structure for resolution.

Category Management Best Practices

Primary category designation:

Implement clear rules for determining primary categories:

IF article from geographic correspondent:
    primary_category = geographic_location
    URL format: /location-news/district-name/article-code
ELSE IF article is breaking news:
    primary_category = breaking-news
    URL format: /breaking/article-code
ELSE IF article has dominant topic:
    primary_category = main_topic
    URL format: /topic-name/article-code
ELSE:
    primary_category = general-news
    URL format: /news/article-code

Secondary category handling:

Secondary categories affect navigation and filtering but never generate URLs:

// Secondary categories stored in database
$article->categories()->attach($secondary_categories, [
    'is_primary' => false,
    'for_navigation_only' => true
]);
// Category pages show articles regardless of primary/secondary status
// But all link to article's canonical URL

Editorial Workflow Integration

Publishing checklist implementation:

□ Article title and content complete
□ Primary category selected (required)
□ Secondary categories selected (optional)
□ Canonical URL generated and verified
□ Preview URL matches expected canonical format
□ No duplicate URL exists in system
□ Publish button enabled only when checklist complete

URL preview in CMS:

Display generated canonical URL prominently during article creation:

┌─────────────────────────────────────────┐
│ Article Preview                          │
├─────────────────────────────────────────┤
│ Title: Major Policy Announcement         │
│                                          │
│ Canonical URL:                           │
│ https://newssite.com/politics/article-1  │
│                                          │
│ This is the ONLY URL for this article    │
└─────────────────────────────────────────┘

Database Integrity Constraints

Implement database-level enforcement where possible:

-- Ensure every article has canonical URL
ALTER TABLE articles
  MODIFY canonical_url VARCHAR(255) NOT NULL;
-- Unique constraint on canonical URLs
CREATE UNIQUE INDEX idx_unique_canonical 
  ON articles(canonical_url);
-- Check constraint for category priority
ALTER TABLE article_categories
  ADD CONSTRAINT chk_primary_category
  CHECK (
    -- Each article must have exactly one primary category
    (SELECT COUNT(*) FROM article_categories 
     WHERE article_id = article_categories.article_id 
     AND is_primary = 1) = 1
  );

These constraints prevent data-level inconsistencies that could lead to multiple URL scenarios.

Monitoring and Maintenance

After implementing solutions, ongoing monitoring ensures continued effectiveness and identifies new issues.

Google Search Console Monitoring

Coverage report analysis:

Monitor for duplicate-related issues:

• "Duplicate without user-selected canonical" - indicates missing or incorrect canonical tags
• "Duplicate, Google chose different canonical" - Google selecting different URL than specified
• "Submitted URL not selected as canonical" - sitemap contains non-canonical URLs

URL inspection tool usage:

Regularly test article URLs to verify:

Test URL: /politics/national/article-code
Google-selected canonical: /politics/national/article-code ✓
Status: Indexed
Test URL: /all-districts/dhaka/article-code (non-canonical)
Google-selected canonical: /politics/national/article-code ✓
Status: 301 redirect (not indexed) ✓

Crawl Analysis

Use enterprise SEO tools to crawl your site regularly:

Screaming Frog audit:

Crawl settings:
- Follow internal links
- Check canonicals
- Identify redirects
- Find duplicates
Analysis:
- List all URLs by unique article code
- Identify articles with multiple accessible URLs
- Verify canonical tags point correctly
- Check 301 redirects function properly

Automated alerts:

Set up monitoring for:

• New duplicate URL discovery
• Canonical tag errors
• 404 errors from broken redirects
• Crawl budget spikes indicating duplicate crawling

Analytics Tracking

Implement GA4 tracking that consolidates metrics by article ID rather than URL:

// Add custom dimension for article ID
gtag('event', 'page_view', {
  'article_id': '<?php echo $article->unique_code; ?>',
  'article_category': '<?php echo $article->primary_category; ?>',
  'url_type': '<?php echo is_canonical_url() ? "canonical" : "redirect"; ?>'
});

This enables accurate performance measurement regardless of which URL variation users initially accessed.

Periodic Technical Audits

Schedule quarterly technical SEO audits focusing on:

URL integrity checks:

• Verify canonical URL enforcement
• Test redirect functionality
• Validate category priority rules still apply
• Check new article types don't bypass validation

Database consistency:

• Ensure all articles have canonical URLs stored
• Verify category relationships integrity
• Check for orphaned URLs
• Validate primary category assignments

Performance metrics:

• Compare crawl efficiency before/after implementation
• Measure ranking consolidation for key articles
• Track index coverage improvements
• Monitor organic traffic trends

Comparison: Solution Approaches

Solution	Implementation Complexity	SEO Effectiveness	Editorial Impact	Technical Maintenance
301 Redirects + Validation	High	Excellent (100%)	Minimal (transparent to users)	Low (once implemented)
Canonical Tags Only	Medium	Good (80-90%)	None	Medium (ongoing monitoring)
Flat URL Structure	Very High	Excellent	High (loses category context)	Low
Primary Category Enforcement	Medium	Excellent	High (limits flexibility)	Low
Virtual Categories	High	Excellent	Medium (workflow change)	Medium

Industry-Specific Considerations

Different news organization types face unique multiple URL challenges.

High-Volume National News Agencies

Organizations publishing 500+ articles daily face:

Scale challenges:

• Manual URL verification impossible
• Automated validation critical
• Crawl budget extremely limited
• Index bloat severely impacts performance

Recommended approach:

• Automated canonical URL generation with strict priority rules
• Comprehensive 301 redirect implementation
• Real-time URL validation on every request
• Dedicated monitoring systems

Regional News Networks

Local news publishers covering specific geographic areas encounter:

Geographic categorization dominance:

• Most articles inherently location-based
• Geographic category should nearly always be primary
• Clear category priority rules essential

Implementation focus:

• Geographic category as default primary
• Simple topic-based secondary categorization
• Location-first URL structure

Specialized News Publications

Topic-focused publications (business news, technology news, sports news) experience:

Simpler category structures:

• Fewer category hierarchy levels
• Less multi-category complexity
• More straightforward canonical determination

Simplified implementation:

• Straightforward primary category rules
• Less complex validation logic
• Lower risk of multiple URL scenarios

Frequently Asked Questions

What causes news CMS to create multiple URLs for the same article?

News CMS platforms create multiple URLs when the system lacks URL validation logic that matches requested category paths against articles' actual category assignments. The CMS uses only the unique article code to retrieve content, ignoring whether the category path in the URL is correct, allowing any article code to work with any category path combination.

How do I identify if my news site has multiple URL problems?

Test by taking an article's unique code and trying it with different category paths. If the same content loads successfully at multiple URL patterns, you have the problem. Check Google Search Console for "Duplicate" issues in coverage reports. Use crawl tools to identify articles accessible via multiple URLs. Monitor for inconsistent ranking performance across URL variations.

Should I use canonical tags or 301 redirects to fix multiple URLs?

301 redirects are preferred because they enforce single URLs, automatically correcting user and search engine access to non-canonical variations. Canonical tags work but are weaker signals that Google may ignore. For news sites, implement 301 redirects at the routing level to prevent non-canonical URLs from displaying content at all, redirecting immediately to canonical URLs instead.

How do I determine which URL should be canonical when articles fit multiple categories?

Establish category priority rules: geographic categories typically take priority for articles from location correspondents; topic categories become primary for articles without geographic association; breaking news may override other priorities. Define clear rules in your CMS and apply consistently. Store the determined canonical URL in your database during publishing and enforce it during retrieval.

Will fixing multiple URLs hurt my existing search rankings?

Initially, rankings may fluctuate as Google consolidates signals onto canonical URLs, but long-term performance improves significantly. Implement redirects carefully to preserve ranking signals. Monitor Search Console closely during transition. Most sites see ranking improvements within 4-8 weeks as authority consolidates and crawl efficiency improves.

Can I keep articles in multiple categories without creating duplicate URLs?

Yes. Articles can appear in multiple category listings on your site while having only one canonical URL. Store category relationships in your database but link to the canonical URL from all category pages. Users navigate through different categories but all links point to the same article URL. This maintains editorial flexibility without URL duplication.

How long does it take to implement URL validation in a news CMS?

Implementation timeline depends on CMS complexity: custom platforms require 2-4 weeks for routing changes, database modifications, and testing; managed CMS platforms may need 1-2 weeks for plugin development or configuration; enterprise systems might take 4-8 weeks including comprehensive testing. Plan adequate time for testing across all article types and category combinations.

What happens to external links pointing to non-canonical URLs?

After implementing 301 redirects, external links to non-canonical URLs automatically redirect to canonical versions, preserving link equity. Search engines update their index to reflect canonical URLs over time. Consider outreach to high-value linking sites requesting URL updates, though redirects handle this automatically for SEO purposes.

Should news archives use the same canonical URL approach?

Yes, archived articles should maintain their original canonical URLs permanently. Don't create archive-specific URL variations. Archived articles should continue redirecting non-canonical paths to canonical URLs. This prevents creating new duplicate URL problems as content ages. Maintain historical URLs consistently even as site structure evolves.