Claim Levels
Define what authority claims Rankfender tests for your brand across AI systems.
Overview
Claim Levels represent a five-tier hierarchy of authority claims that Rankfender tests for your brand across AI systems. Each level represents an increasingly strong assertion about your brand, from basic existence to market leadership.
The Five-Tier System
| Level | Name | Example Claim | Evidence Required |
|---|---|---|---|
| Level 0 | Existence | "Brand X exists" | AI systems acknowledge your brand exists when asked about it directly. |
| Level 1 | Usage | "Brand X is used by customers" | AI systems reference your brand as an active product or service with real users. |
| Level 2 | Comparative | "Brand X compares favorably" | AI systems mention your brand positively when comparing options in your category. |
| Level 3 | Evaluative | "Brand X is evaluated positively by experts" | AI systems cite expert opinions, reviews, or authoritative evaluations that favor your brand. |
| Level 4 | Superlative | "Brand X is the best/leader" | AI systems position your brand as the top choice, market leader, or best-in-class solution. |
Per-Claim Tracking
| Field | Description |
|---|---|
| claim_text | The specific statement being tested. |
| evidence_required | The type of evidence needed to validate this claim in AI responses. |
| evidence_current | The evidence currently found in AI responses supporting or contradicting this claim. |
| approved | Whether the claim has been reviewed and approved for active tracking. |
| performance_score | A score from 0 to 100 indicating how strongly AI responses support this claim overall. |
| mention_rate | A score from 0 to 100 representing the percentage of relevant prompts where this claim is supported. |
| refusal_rate | A score from 0 to 100 representing the percentage of prompts where AI systems refuse to make or support this claim. |
Auto-Detection of Claims
During prompt generation, Rankfender automatically detects potential claims based on your brand book, competitor data, and industry context. These auto-detected claims are added to the appropriate level and marked as unapproved until you review them.
Manual Claim Management
- Create — Add a new claim at any level by specifying the claim text and the evidence required to validate it.
- Edit — Modify the claim text, evidence requirements, or level assignment of an existing claim.
- Approve — Mark an auto-detected or newly created claim as approved for active tracking and inclusion in reports.
- Delete — Remove a claim entirely from your project. This also removes all associated tracking data.
Best Practices
Start by ensuring your Level 0 (Existence) and Level 1 (Usage) claims are well-supported before focusing on higher levels. Monitor the refusal rate closely: a high refusal rate indicates that AI systems consider the claim unsupported or controversial, which may require strengthening your online evidence through content and citations.