EmailContact
Skip to main content

LLM-SEO Troubleshooting Guide

🎯 Quick Summary​

  • Diagnose and fix common LLM-SEO issues
  • Citation rate not improving - root causes and solutions
  • Technical problems (schema errors, crawler blocks, indexing)
  • Strategic issues (wrong content type, competition, positioning)
  • Platform-specific troubleshooting for ChatGPT, Claude, Gemini, Perplexity

📋 Table of Contents​

  1. Diagnostic Framework
  2. Citation Rate Issues
  3. Technical Problems
  4. Content Quality Issues
  5. Platform-Specific Troubleshooting
  6. Advanced Diagnostics

🔑 Key Concepts at a Glance​

  • Citation Rate Stuck: Most common issue (solutions vary by cause)
  • Schema Errors: Invalid markup blocks AI understanding
  • Crawler Blocks: robots.txt or server blocks prevent access
  • Content Gaps: Missing E-E-A-T signals or wrong format
  • Platform Differences: What works on ChatGPT may not work on Claude

🏷️ Metadata​

Tags: troubleshooting, debugging, issues, fixes, diagnosis Status: %%ACTIVE%% Complexity: %%ADVANCED%% Max Lines: 450 (this file: 445 lines) Reading Time: 12 minutes Last Updated: 2025-01-18

Diagnostic Framework​

Systematic Troubleshooting Process​

5-Step Diagnostic Method:

Step 1: Identify the Symptom
□ Citation rate is 0%
□ Citation rate stuck (not improving)
□ Lost citations (was cited, now not)
□ Platform-specific issue (works on A, not B)
□ Competitor outranking you

Step 2: Gather Data
□ Current citation status (all platforms)
□ Historical trends (last 90 days)
□ Competitor comparison
□ Technical audit results
□ Content quality assessment

Step 3: Hypothesize Root Cause
□ Technical block (robots.txt, server, schema)
□ Content quality (E-E-A-T, format, depth)
□ Competition (superior content exists)
□ Timing (not enough time passed)
□ Platform algorithm change

Step 4: Test Hypothesis
□ Run technical tests
□ Compare against working examples
□ Check platform documentation
□ Analyze top-cited competitors

Step 5: Implement Fix
□ Apply solution
□ Document change
□ Monitor for 14-30 days
□ Measure impact

If no improvement → Return to Step 3 (new hypothesis)

Quick Diagnostic Checklist​

Run this checklist first (5 minutes):

Technical Basics:
□ robots.txt allows AI crawlers (check at yoursite.com/robots.txt)
□ Content is publicly accessible (no login required)
□ HTTPS enabled (SSL certificate valid)
□ Page loads in <3 seconds
□ No meta noindex tag present
□ Schema markup validates (schema.org validator)

Content Basics:
□ Content is >500 words
□ Has answer-first structure (answer in first 60 words)
□ Includes schema markup (FAQPage, Article, or HowTo)
□ Has author attribution
□ Updated in last 6 months

Platform Status:
□ UnrealSEO tracking active (if using)
□ No recent platform algorithm changes
□ Competitor content exists (not a dead query)

If all checked ✓ and still no citations → Continue to specific diagnostics

Citation Rate Issues​

Issue 1: Citation Rate is 0%​

Problem: No citations on any platform after 30+ days

Diagnostic Steps:

1. Check Crawler Access
   Test: View yoursite.com/robots.txt

   ❌ Bad:
   User-agent: GPTBot
   Disallow: /

   ✓ Good:
   User-agent: GPTBot
   Allow: /

   Fix: Update robots.txt to allow AI crawlers

2. Check Content Accessibility
   Test: Open page in incognito/private mode

   If you see: Login wall, paywall, 404 error
   → Content not publicly accessible
   Fix: Make content public or use preview/snippet

3. Check Indexing
   Test: site:yoursite.com/your-page in Google

   If not indexed by Google → Won't be indexed by AI
   Fix: Submit to Google Search Console, check for noindex tag

4. Check Schema Validity
   Test: https://validator.schema.org

   Errors found: Invalid JSON-LD, missing required fields
   Fix: Correct schema syntax

5. Check Content Quality
   Minimum bar:
   • 500+ words ✓
   • Answer-first structure ✓
   • Schema markup ✓
   • Author attribution ✓

   If missing → Add

6. Wait Time
   Have you waited 14+ days since publishing?
   AI platforms don't crawl immediately
   Typical timeline: 7-21 days for first citation chance

Solution Priority:

1. Fix robots.txt (immediate impact)
2. Make content public (immediate)
3. Add/fix schema markup (1-2 weeks)
4. Improve content quality (2-4 weeks)
5. Build authority (1-3 months)

Issue 2: Citation Rate Stuck (Not Improving)​

Problem: Cited on 1-2 platforms, not improving beyond 25-50%

Diagnostic:

Current: 25% citation rate (1/4 platforms)
Goal: 75%+ (3+/4 platforms)
Timeline: Stuck for 60+ days

Root Causes:

Cause 1: Platform-Specific Issues
Symptom: Works on ChatGPT, not on Claude/Gemini
Diagnosis:
  • Check Claude-specific requirements (longer content preferred)
  • Check Gemini-specific signals (schema critical)
  • Platform may have different quality thresholds

Cause 2: Competitive Content Superior
Symptom: Competitor always cited instead
Diagnosis:
  • Compare your content vs competitor
  • Check E-E-A-T signals (author credentials, links)
  • Check content depth (word count, examples, data)

Cause 3: Content Format Mismatch
Symptom: Good content, but wrong format for AI
Diagnosis:
  • Cited content has FAQ format → yours doesn't
  • Cited content has comparison table → yours doesn't
  • Cited content has step-by-step → yours doesn't

Cause 4: Insufficient Authority
Symptom: Everything technically correct, still not cited
Diagnosis:
  • New website ( under 6 months old)
  • Few inbound links ( under 10)
  • No author recognition
  • No brand mentions elsewhere

Solutions by Cause:

Platform-Specific:
→ Optimize specifically for missing platforms
→ Claude: Add longer explanations, cite sources
→ Gemini: Ensure perfect schema markup
→ Perplexity: Emphasize recency, update dates

Competitive:
→ Analyze top-cited competitor (view in AI answer)
→ Match or exceed their content quality
→ Add unique data/research they lack
→ Build superior E-E-A-T signals

Format:
→ Add FAQ section (10-15 questions)
→ Add comparison table (your solution vs alternatives)
→ Add step-by-step guide (numbered lists)
→ Add visual content (charts, infographics - describe in alt text)

Authority:
→ Get featured on industry publications (links + mentions)
→ Guest post on authoritative sites
→ Build author profiles (LinkedIn, Twitter)
→ Collect testimonials and reviews
→ Time (authority builds over 3-6 months)

Issue 3: Lost Citations​

Problem: Was cited, now not cited (regression)

Diagnostic:

Timeline:
Jan 1-15: Cited on ChatGPT (4/4 platforms)
Jan 16: Lost ChatGPT citation
Jan 20: Lost Claude citation
Current: 2/4 (50% → regression)

Root Causes:

Cause 1: Content Changed
Check: Did you update the page?
  • Removed key content AI was citing
  • Changed answer-first structure
  • Removed schema markup
  • Made content less clear

Fix: Revert changes or improve new version

Cause 2: Competitor Published Better Content
Check: Search same query in AI platforms
  • New competitor content appeared
  • Competitor improved their existing content
  • Competitor now has superior E-E-A-T

Fix: Out-compete with better content

Cause 3: Platform Algorithm Change
Check: UnrealSEO community, Twitter, industry news
  • ChatGPT updated citation algorithm (Jan 2025)
  • New quality thresholds announced
  • Platform preferences shifted

Fix: Adapt to new algorithm requirements

Cause 4: Technical Issue
Check: robots.txt, site access, errors
  • robots.txt accidentally blocked crawler
  • Server returned errors during AI crawl
  • Page became inaccessible

Fix: Restore technical access

Cause 5: Content Became Stale
Check: Last updated date
  • Content published in 2022, not updated
  • Competitors have 2025 content
  • AI platforms prefer fresh content

Fix: Update content, change publish date

Technical Problems​

Schema Markup Errors​

Common Schema Errors:

Error 1: Invalid JSON-LD Syntax

<!-- ❌ Wrong: Missing closing brace -->
<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [{
    "@type": "Question",
    "name": "What is CRM?",
    "acceptedAnswer": {
      "@type": "Answer",
      "text": "CRM is..."
  }]  <!-- Missing closing brace here -->
</script>

<!-- ✓ Fixed: Proper JSON -->
<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "FAQPage",
  "mainEntity": [{
    "@type": "Question",
    "name": "What is CRM?",
    "acceptedAnswer": {
      "@type": "Answer",
      "text": "CRM is..."
    }
  }]
}
</script>

Validation: https://validator.schema.org

Error 2: Missing Required Fields

// ❌ Wrong: Missing "acceptedAnswer"
{
  "@type": "Question",
  "name": "What is CRM?"
  // Missing acceptedAnswer field
}

// ✓ Fixed: All required fields
{
  "@type": "Question",
  "name": "What is CRM?",
  "acceptedAnswer": {
    "@type": "Answer",
    "text": "CRM (Customer Relationship Management)..."
  }
}

Error 3: Duplicate Schema

<!-- ❌ Problem: Same schema twice -->
<script type="application/ld+json">
{ "@type": "Article", ... }
</script>

<!-- Plugin also adds Article schema -->
<script type="application/ld+json">
{ "@type": "Article", ... }
</script>

Result: Conflicting data, AI confusion

Fix: Remove duplicate, use one source only

Crawler Blocking Issues​

Diagnosis: Are AI crawlers blocked?

# Check robots.txt
curl https://yoursite.com/robots.txt

# Check for blocks:
User-agent: GPTBot
Disallow: /  ← BLOCKED

User-agent: Claude-Web
Disallow: /  ← BLOCKED

# Check server response to crawler
curl -A "GPTBot/1.0" https://yoursite.com/page

# If returns: 403 Forbidden or 503 → Blocked at server level

Common Blocking Scenarios:

Scenario 1: Security Plugin Blocking
Problem: Wordfence, Cloudflare blocks AI user-agents
Fix: Whitelist AI crawler user-agents

Scenario 2: Rate Limiting
Problem: AI crawler makes many requests, triggers rate limit
Fix: Increase rate limits for known AI crawlers

Scenario 3: Geo-Blocking
Problem: AI platforms crawl from specific IPs, geo-blocked
Fix: Whitelist AI platform IP ranges

Scenario 4: WAF Rules
Problem: Web Application Firewall blocks "bot" patterns
Fix: Create exception for legitimate AI crawlers

Content Quality Issues​

E-E-A-T Signal Deficiencies​

Diagnostic: E-E-A-T Checklist

Experience (Have you done this?):
□ First-person language ("In my 10 years...")
□ Specific examples from your work
□ Screenshots/photos of your process
□ Case studies with real data

If missing → Add experiential content

Expertise (Do you know this?):
□ Author credentials visible (degrees, certs)
□ Technical depth beyond surface-level
□ Unique insights not found elsewhere
□ Data, research, or original analysis

If missing → Add expertise signals

Authoritativeness (Do others recognize you?):
□ Inbound links from industry sites
□ Brand mentions on authoritative sites
□ Media coverage, awards, recognition
□ Social proof (followers, engagement)

If missing → Build authority over time

Trustworthiness (Can you be trusted?):
□ Contact information visible
□ HTTPS enabled
□ Privacy policy, terms of service
□ Transparent about affiliations/sponsors
□ Correction policy (updates noted)

If missing → Add trust signals immediately

Platform-Specific Troubleshooting​

ChatGPT Not Citing​

Specific diagnostics for ChatGPT:

Issue: Works on Claude/Gemini, not ChatGPT

Potential Causes:

1. Content Too Long
   ChatGPT prefers: 1,500-2,500 words
   Your content: 5,000+ words
   Fix: Break into multiple focused pages

2. Answer Not Clear Enough
   ChatGPT needs: Direct answer in first paragraph
   Your content: Answer buried on line 50
   Fix: Move answer to top (answer-first)

3. Too Technical
   ChatGPT audience: General, broad
   Your content: PhD-level technical jargon
   Fix: Simplify language, add glossary

4. No FAQ Section
   ChatGPT loves: Question-answer format
   Your content: Long prose, no Q&A
   Fix: Add FAQ section with schema

Claude Not Citing​

Issue: Works on ChatGPT, not Claude

Claude Preferences:

1. Longer, Detailed Content
   Claude prefers: 2,500-4,000 words
   (Longer than ChatGPT preference)
   Fix: Expand content with more depth

2. Citations to Sources
   Claude values: Links to research, data sources
   Your content: No external citations
   Fix: Add 5-10 citations to authoritative sources

3. Nuanced, Balanced Perspective
   Claude likes: "It depends..." explanations
   Your content: Absolute statements only
   Fix: Add nuance, caveats, multiple perspectives

4. Recent Content
   Claude emphasizes: Freshness signals
   Your content: Published 2022, never updated
   Fix: Update content, change published date

Gemini Not Citing​

Issue: Works elsewhere, not Gemini

Gemini Requirements:

1. Perfect Schema Markup
   Gemini strict: Schema must validate 100%
   Your schema: Has warnings/errors
   Fix: Fix all schema validation errors

2. Structured Data Emphasis
   Gemini loves: Tables, lists, comparisons
   Your content: Long paragraphs only
   Fix: Add comparison table, bullet lists

3. Mobile-First
   Gemini prioritizes: Mobile experience
   Your content: Desktop-only, slow mobile load
   Fix: Optimize for mobile, improve speed

4. Visual Content Descriptions
   Gemini uses: Image analysis (alt text critical)
   Your images: Missing or poor alt text
   Fix: Add descriptive alt text to all images

Advanced Diagnostics​

Citation Rate Benchmarking​

Is your rate actually low?

Your Citation Rate: 18%
Your Assessment: "Too low, something wrong"

Industry Benchmarks:
• SaaS (your industry): 22% median, 38% top 10%
• Your content type (guides): 19% median
• Your domain age (2 years): 16% expected

Reality Check:
18% is ABOVE average for 2-year-old site
Not a problem, just needs time + optimization

When to Worry:
• <10% after 12 months of optimization
• 0% after 6 months
• Declining trend over 3+ months

📚 Related Topics​

Technical:

Content:

🆘 Need Help?​

Still Stuck?

Report Platform Issues:

Last updated: 2025-01-18 | Edit this page

Last updated on