SpyderBot Documentation
Getting Started
Prepare brand context carefully before you start. Accurate context improves entity resolution and helps the analysis use prompts that reflect the intended business scope.
Before starting an analysis
- Brand display name and legal name where relevant
- Primary domain
- Product or service scope
- Category
- Market and geography
- Language
- Known aliases
- Important competitors
- Brand context that should guide prompt selection
Starting an analysis
- Enter a brand or domain.
- Confirm the resolved brand entity.
- Review and edit brand context.
- Review the category and market.
- Select the competitor comparison set.
- Review the planned Prompt Set and AI-system scope.
- Review expected executions, credits, and duration.
- Confirm and start the analysis.
If the brand or domain could refer to more than one entity, SpyderBot asks for confirmation before the analysis begins.
Competitor selection
Competitors define the comparison scope for metrics such as AI Share of Voice and Category Rank. Choose organizations that compete for similar customer intents, products, categories, or markets.
The selected competitor set is saved with the report run. Later changes to a live competitor list do not alter a completed report.
Prompt Set preview
Before starting, review Prompt Set size, Stable and Adaptive counts, included AI systems, prompt-family coverage, planned execution budget, expected prompt-system coverage, estimated credits, and estimated duration.
Prompt Set size depends on plan capacity, qualified prompt availability, required coverage, and evidence depth. SpyderBot does not add filler prompts merely to reach a package maximum.
Plan scope
Agent count represents operational execution capacity. It is not the number of unique prompts and is not a statistical sample size. Rendered plan values follow the versioned plan-policy configuration shown in the product.
| Plan | LLM Agents | AI systems | Prompt Set range | Default Prompt Set | Planned execution budget |
|---|---|---|---|---|---|
| Free | 30 | ChatGPT, Gemini, Copilot | 19â21 | 20 | 150 |
| Go | 50 | ChatGPT, Grok, Gemini, Copilot, Perplexity | 21â23 | 22 | 250 |
| Pro | 100 | ChatGPT, Grok, Gemini, Copilot, Perplexity, Llama, Claude | 29â31 | 30 | 500 |
| Business | 300 | ChatGPT, Grok, Gemini, Copilot, Perplexity, Llama, Claude | 50â54 | 52 | 1,500 |
During analysis
- Building Prompt Set
- Assigning LLM Agents
- Collecting AI-system responses
- Validating responses
- Extracting observations
- Computing metrics
- Reconciling the report
- Completed or Needs Attention
A run is not ready for reporting until required validation and reconciliation checks pass.
Managing analyses
Common states include Draft, Queued, Running, Validating, Completed, Needs Attention, Failed, and Archived.
Use View Report only for a completed, accessible report run. Each completed row must open the exact immutable Report Run ID.