SpyderBot Documentation
Prompt Intelligence Overview
Prompt Intelligence is SpyderBot's investigation platform for understanding AI behavior at the prompt level.
Overview
While Brand Insights provides a strategic view of overall AI Visibility, Prompt Intelligence investigates individual prompts to explain how and why AI systems generate specific responses.
It helps organizations move beyond summary metrics and examine the AI behaviors that influence recommendations, citations, entities, competitive positioning, and response structure.
Rather than asking:
"How visible is our brand?"
Prompt Intelligence asks:
"Why did AI generate this response?"
This investigative approach enables organizations to understand the mechanisms behind AI-generated responses and identify opportunities to improve AI Visibility at the prompt level.
Why Prompt Intelligence Matters
Every AI-generated response begins with a prompt.
Whether users ask AI to recommend software, compare competitors, explain a concept, or suggest products, each prompt represents a unique decision environment.
Organizations often discover that:
- Some prompts consistently recommend their brand.
- Others never mention them.
- Some prompts change significantly after AI model updates.
- Different AI models generate different responses for the same prompt.
Without prompt-level investigation, these behaviors remain difficult to explain.
Prompt Intelligence was designed to make them observable.
Business Decisions
Prompt Intelligence helps organizations answer strategic questions such as:
- Why does AI recommend one competitor instead of another?
- Which prompts matter most to our business?
- What changed after an AI model update?
- Which AI model behaves differently?
- Which prompts should we monitor continuously?
- Which prompts should we prioritize for optimization?
Two Complementary Products
Prompt Intelligence consists of two complementary products.
Together they provide both investigation and continuous observation.
Prompt Explorer
Prompt Explorer analyzes prompts on demand.
It helps organizations investigate individual prompts, compare AI responses, analyze recommendations, understand entities, examine citations, and evaluate response behavior across AI models.
Prompt Explorer is best suited for exploratory analysis and prompt-level investigation.
Prompt Explorer
Prompt Observatory
Prompt Observatory continuously monitors strategically important prompts.
It tracks AI response changes over time, detects meaningful shifts, identifies likely change drivers, and maintains historical observations.
Prompt Observatory is designed for continuous monitoring rather than one-time analysis.
Prompt Observatory
How Prompt Intelligence Fits Within SpyderBot
Prompt Intelligence builds on the strategic insights provided by Brand Insights.
A typical investigation workflow looks like this:
While Brand Insights identifies where opportunities exist, Prompt Intelligence explains why those opportunities exist.
Typical Investigation Workflow
Although Prompt Intelligence supports many use cases, a common workflow includes the following steps.
1. Identify an Important Prompt
Choose a prompt that is strategically important to your organization.
This may be a recommendation prompt, a comparison prompt, a purchasing question, or any prompt that influences customer decisions.
2. Investigate with Prompt Explorer
Analyze the prompt across supported AI models.
Review recommendations, citations, entities, response structure, and supporting evidence.
3. Explain AI Behavior
Identify the observations that best explain why AI generated the response.
Investigate competitive differences, entity relationships, recommendation patterns, and other supporting signals.
4. Monitor with Prompt Observatory
If the prompt is strategically important, create a Prompt Observatory.
Continuous monitoring helps detect meaningful AI changes without requiring repeated manual analyses.
5. Improve and Measure Again
Apply optimization efforts, continue monitoring, and investigate future changes as AI systems evolve.
Prompt Intelligence supports continuous investigation rather than one-time analysis.
Core Capabilities
Prompt Intelligence provides capabilities including:
- Prompt-level analysis
- Cross-model comparison
- Recommendation analysis
- Citation analysis
- Entity analysis
- Response structure analysis
- Historical prompt monitoring
- Prompt Change Intelligence™
- AI Change Attribution Engine™
- Evidence Layer
These capabilities work together to transform AI responses into explainable intelligence.
Related Concepts
To better understand Prompt Intelligence:
- Concepts → Prompt Sets
- Concepts → AI Models
- Concepts → Observation
- Concepts → Evidence Layer
- Concepts → AI Visibility
- Concepts → AI Perception
Related Products
Prompt Intelligence works closely with:
Together these products support the complete AI Visibility lifecycle.
Next Steps
Continue with:
- Products → Prompt Intelligence → Prompt Explorer — Analyze individual prompts.
- Products → Prompt Intelligence → Prompt Observatory — Monitor important prompts over time.
Prompt Explorer Overview
Overview
Prompt Explorer is SpyderBot's investigation workspace for analyzing individual AI prompts.
It enables organizations to examine how AI systems respond to a specific prompt, compare behavior across AI models, investigate recommendations, analyze citations, evaluate entities, and understand the factors influencing AI-generated responses.
Unlike Brand Insights, which provides a strategic view across thousands of observations, Prompt Explorer focuses on one prompt at a time.
Its purpose is to help organizations understand not only what AI answered, but why AI answered that way.
Business Decision
Prompt Explorer helps answer one strategic decision:
Why did AI generate this response, and what can we learn from it?
Understanding an individual AI response often requires deeper investigation than a high-level report can provide.
Prompt Explorer provides that investigative capability.
Business Questions
Prompt Explorer helps answer questions such as:
- Why does AI recommend one brand instead of another?
- Which entities influence this response?
- Which citations support the generated answer?
- How do different AI models respond to the same prompt?
- Which competitors appear most frequently?
- Which observations explain the response?
What Prompt Explorer Provides
Prompt Explorer transforms a single prompt into a structured AI investigation.
Depending on the selected configuration, investigations may include:
- AI-generated responses
- Recommendation analysis
- Citation analysis
- Entity analysis
- Competitive analysis
- Cross-model comparison
- Response structure analysis
- Evidence Layer
Rather than displaying a single AI response, Prompt Explorer helps organizations understand the observations behind that response.
Investigation Workflow
A typical Prompt Explorer investigation follows five stages.
Each investigation is designed to answer progressively deeper questions rather than simply display AI outputs.
When to Use Prompt Explorer
Prompt Explorer is particularly useful when organizations need to investigate questions such as:
Why did AI recommend a competitor?
Analyze recommendation behavior, supporting entities, citations, and comparative responses.
Why did our AI Visibility change?
Compare observations across AI models or repeat investigations to identify meaningful differences.
How do different AI models behave?
Compare responses generated by multiple supported AI systems using the same prompt.
Which prompts influence our business most?
Investigate strategically important prompts before creating long-term monitoring within Prompt Observatory.
How should we optimize?
Identify the observations most likely to explain the current AI response before implementing optimization efforts.
Prompt Explorer vs Prompt Observatory
Although both products analyze prompts, they serve different purposes.
| Prompt Explorer | Prompt Observatory |
|---|---|
| Investigates individual prompts | Continuously monitors important prompts |
| On-demand analysis | Scheduled observation |
| Explains current AI behavior | Detects changes over time |
| Best for investigation | Best for monitoring |
Many organizations use Prompt Explorer first and Prompt Observatory afterward.
Core Capabilities
Prompt Explorer includes capabilities such as:
- Prompt configuration
- AI model selection
- Agent coordination
- Prompt-level investigation
- Recommendation analysis
- Citation analysis
- Entity analysis
- Evidence Layer
- Export
Each capability contributes to a deeper understanding of AI-generated responses.
Related Concepts
To better understand Prompt Explorer:
- Concepts → Prompt Sets
- Concepts → AI Models
- Concepts → Observation
- Concepts → Evidence Layer
- Concepts → AI Perception
Related Products
Prompt Explorer works closely with:
- Products → Prompt Intelligence → Prompt Observatory — Monitor strategically important prompts over time.
- Products → Brand Insights — Identify which prompts deserve deeper investigation.
Together they support the complete prompt investigation lifecycle.
Next Steps
Continue with:
- Products → Prompt Explorer → Creating Runs
- Products → Prompt Explorer → Prompt Sets
- Concepts → AI Models
- Products → Prompt Explorer → Agent Count
These pages explain how to configure and perform prompt investigations using Prompt Explorer.