# GEO Lens Self-Diagnosis Benchmark

GEO Lens should use its own diagnosis system monthly to track brand mentions, competitors, cited sources, content gaps, and retest changes in AI answers.

## Recommended monthly scope

- 24 stable questions covering product, pricing, methodology, report interpretation, retest, and agency scenarios.
- 6 exploratory questions for newly emerging phrasing and competitors.
- Target platforms such as ChatGPT, Perplexity, Gemini, Copilot, Doubao, Kimi, and Tongyi when supported by the collection configuration.
- One stable baseline round, with the same scope repeated month over month.

## Fields to record

| Field | Purpose |
|---|---|
| Brand mentioned | Whether GEO Lens enters the AI answer set |
| Competitors | Which alternatives appear |
| Cited sources | Which GEO Lens or third-party pages are referenced |
| Missing buckets | Which questions do not surface useful GEO Lens coverage |
| Action backlog | Which page, doc, external source, or technical task should be improved |

Public benchmark output must use GEO Lens self-data or fully redacted aggregates and should state date, platform scope, question count, rounds, and limitations.

Benchmark page: https://geo.9fox.ai/zh/geo-benchmark