LLM Visibility

Track when AI assistants like ChatGPT, Perplexity, and Claude mention your brand. Monitor competitors, log citations, and optimize your content to appear in AI-generated answers.

How it works

RankFlo tracks LLM visibility in two ways:

1. AI referrer tracking — automatically detects when visitors arrive from AI platforms (built into analytics)
2. Mention tracking — tests prompts against AI models and logs when your brand is mentioned in responses

Step 1: Connect your AI provider

LLM visibility tracking requires API keys to query AI models. Go to Settings → AI and connect at least one provider:

How to connect

1. Go to Settings → AI in your dashboard
2. Select your preferred provider (OpenAI, Anthropic, or Google)
3. Paste your API key
4. Click Save settings

Once connected, your key is used for both AI writing features AND LLM visibility tracking.

Step 2: Enable automated tracking

RankFlo includes a cron endpoint that automatically tests prompts against AI models and logs mentions. Set up a daily cron job:

# Call this daily (e.g., via cron-job.org, GitHub Actions, or crontab)
curl -X POST https://app.rankflo.io/api/cron/llm-track \
  -H "Authorization: Bearer YOUR_CRON_SECRET" \
  -H "Content-Type: application/json"

What it does on each run

1. Picks 2 random prompts from a built-in library of SEO-related queries
2. Sends each prompt to ChatGPT, Claude, and Perplexity (if keys are configured)
3. Scans responses for your brand name and competitor names
4. Logs mentions with sentiment analysis (positive, neutral, negative)
5. Extracts the relevant snippet for context

Built-in test prompts

The tracker tests prompts like:

• "What is the best headless CMS for startups?"
• "Recommend an open source blog platform with SEO tools"
• "What are the best WordPress alternatives for developers?"
• "How to optimize content for AI search engines"
• "Best CMS with built-in analytics"

Setting up the cron job

Option A: cron-job.org (free)

1. Sign up at cron-job.org
2. Create a new job pointing to https://your-domain.com/api/cron/llm-track
3. Set method to POST
4. Add header: Authorization: Bearer YOUR_CRON_SECRET
5. Schedule: once daily (e.g., 6:00 AM)

Option B: GitHub Actions

# .github/workflows/llm-track.yml
name: LLM Visibility Tracker
on:
  schedule:
    - cron: '0 6 * * *'  # Daily at 6 AM UTC
jobs:
  track:
    runs-on: ubuntu-latest
    steps:
      - run: |
          curl -X POST https://your-domain.com/api/cron/llm-track \
            -H "Authorization: Bearer ${{ secrets.CRON_SECRET }}"

Option C: Server crontab

# Add to crontab -e
0 6 * * * curl -s -X POST https://your-domain.com/api/cron/llm-track -H "Authorization: Bearer YOUR_SECRET"

Step 3: View your LLM dashboard

Switch to LLM Search mode in the top nav bar. The dashboard shows:

Mention stats

• Total mentions — how many times AI assistants referenced your brand
• Trend — period-over-period change
• By platform — breakdown across ChatGPT, Perplexity, Claude, Gemini
• Sentiment — positive, neutral, and negative mention ratio

Visibility score

A 0-100 score calculated from five components:

Share of Voice

See how your brand compares to competitors in AI recommendations. The tracker automatically logs competitor mentions (Contentful, Strapi, Ghost, etc.) alongside your own.

Mention list

Every logged mention shows: platform, query prompt, response snippet, sentiment, and date. Filter by platform or sentiment.

Manual mention logging

You can also log mentions manually when you spot them. In the LLM Search dashboard, click Log mention and fill in:

• Platform — which AI assistant (ChatGPT, Perplexity, Claude, etc.)
• Query — what the user asked
• Snippet — the relevant part of the AI response
• URL — if the AI linked to your content
• Sentiment — positive, neutral, or negative

How to improve your LLM visibility

1. Add an llms.txt file — RankFlo generates this automatically at /llms.txt for every project. It tells AI crawlers about your content.
2. Write clear, factual content — AI models prefer content that states facts directly with specific numbers.
3. Use structured data — JSON-LD schema markup helps AI understand your content. RankFlo adds this automatically.
4. Build topical authority — Publish 10-20 posts around each core topic. Depth signals expertise to AI models.
5. Get cited by third parties — Guest posts, press mentions, and backlinks from authoritative sites increase AI citation likelihood.
6. Keep content fresh — Update key articles quarterly. AI with web access prefers recent sources.

See the complete LLM visibility guide for more strategies.

Prompt testing library

The LLM dashboard includes a prompt library where you can save prompts to test regularly:

1. Go to LLM Search → Prompts
2. Add prompts your target customers might ask AI assistants
3. Mark prompts as "tested" after checking the results
4. Track which prompts return mentions over time