Prompt Explorer

Overview

Prompt Explorer measures your "AI visibility" — how often AI engines like ChatGPT and Perplexity mention your site when users ask relevant questions in your industry.

As AI search becomes more popular, traditional SERP rankings matter less. Prompt Explorer helps you understand if AI engines are recommending your content.

How it works

1. GrowDeck generates industry-relevant prompts based on your site's content
2. Sends these prompts to ChatGPT (GPT-4o) and Perplexity
3. Checks if AI responses mention your site's domain
4. Calculates visibility score: (mentions / total_probes) × 100%

Running a visibility check

1. Navigate to your site → Prompt Explorer tab
2. Click "Run now"
3. Wait 30-60 seconds while AI engines are probed
4. View results: per-model mention rates and example excerpts

Probes can also be triggered:

• On a schedule (daily/weekly)PRO
• Via API: POST /api/v1/sites/:siteId/prompt-explorer

What gets tracked

For each probe run:

• Total prompts sent (typically 20-30 per run)
• Mentions per AI model (ChatGPT vs. Perplexity)
• Overall visibility score (%)
• Example excerpts where your site was mentioned
• Prompts that didn't mention you (opportunity gaps)
• Historical trend over time

Interpreting results

• 0-10%: Low visibility — AI engines rarely mention you
• 10-30%: Moderate visibility — mentioned occasionally
• 30-50%: Good visibility — frequently cited
• 50%+: Excellent visibility — trusted authority in your niche

Improving AI visibility

To increase your visibility score:

• Create comprehensive guides — AI engines prefer depth over breadth
• Answer specific questions — target "how to" and "what is" queries
• Use structured data — schema markup helps AI understand your content
• Build backlinks — authority signals matter to AI training
• Optimize for GEO — see GEO Engine docs

Per-model breakdown

Prompt Explorer tests both major AI engines:

• ChatGPT (GPT-4o): Most popular AI assistant (450M+ users)
• Perplexity: AI search engine with real-time web access

Scores often differ between models due to different training data and web access patterns.

Example use cases

• Developer tools: Check if ChatGPT recommends your library/framework
• SaaS products: See if AI suggests your tool for common problems
• Content sites: Measure if AI cites your articles as sources
• E-commerce: Track product recommendation frequency

API access

Requires API keysPRO

• OpenAI API key: Required for ChatGPT probes (uses GPT-4o)
• Perplexity API key: Optional but recommended for full coverage

Without Perplexity API key, only ChatGPT visibility is measured (still valuable).

Tracking trends

Visibility scores are tracked over time. The dashboard shows:

• 7-day, 30-day, and all-time trends
• Score improvements after content updates
• Comparison between AI models
• Best-performing content types