Prompts Operation Guide

Prompts are the monitored questions GEOly runs against AI platforms. They define what user intent, product category, competitor comparison, or brand scenario you want to measure.

---

I. Configure Monitoring Prompts

1. Add Monitoring Prompts

• Add Manually: Click + Add Prompt, then enter brand-related questions that real users might ask AI.
• Batch Import: If you already have many prompts, use import to upload supported files such as CSV or JSON when the option is available in your workspace.
• Prompt Quality: Write prompts as natural user questions. Include buying intent, comparison intent, use cases, product categories, and problem statements.

2. AI Prompt Research

AI Prompt Research helps expand a seed domain, topic, or keyword set into prompt ideas.

• Enter the brand domain and seed keywords.
• Select prompt language and target country/region where available.
• Review generated prompts before adding them to monitoring.
• Keep prompts grouped by topic so later analysis can be filtered cleanly.

---

II. AI Query Fan-Out

AI Query Fan-Out expands a core intent into related query angles. Use it to cover the ways users may phrase the same buying, comparison, or research need.

• Start with a specific seed intent, not a generic keyword.
• Keep useful generated queries and remove duplicates or off-topic suggestions.
• Group approved prompts into the correct topic before running monitoring.

---

III. Run Monitoring

After prompts are added, you can run all prompts or select specific prompts to run.

• Running prompts may consume monitoring credits depending on your workspace settings.
• Monitoring results are stored as prompt records by platform and date.
• Not every platform will return the same answer shape, citations, or product-card data.

---

IV. View Prompt Execution Results

Click a prompt to open its detailed analysis.

Prompt detail may include:

• Prompt-level AIGVR, mention rate, citation rate, and average position.
• Platform-level performance across ChatGPT, Gemini, Perplexity, Grok, Google AI, or other supported platforms.
• Latest monitoring records by platform, including mention status, position, sentiment, and answer details.
• Citation source links and cited pages where available.
• Competitor mentions and Share of Model context when competitor data is configured.

When reviewing prompt-level metrics, use the platform and date filters consistently. Prompt citation data shows source details; headline brand citation rate should still come from the brand overview when reporting overall performance.