What does a GEO audit cover?
Four workstreams, each ending in findings you can verify yourself:
- Entity audit. Is your organization one consistent thing across the web? We check NAP consistency, knowledge-panel and Wikidata-class presence, alternateName and sameAs coverage, hreflang integrity for multilingual sites, and contradictions between your site, profiles and directories.
- Content citability audit. Do your pages answer questions in extractable form? We score answer-block coverage against a fan-out question map, structured-data validity, freshness, and the presence of quotable facts (numbers, dates, named entities).
- Technical audit. Can the engines reach you? Indexation on Google and Bing (ChatGPT’s dependency), IndexNow status, crawler access for GPTBot, OAI-SearchBot and friends, rendering of key content in served HTML.
- Answer-layer measurement. The heart of the audit: your fixed question set run across Google AI Overviews, ChatGPT, Perplexity and Gemini, scored on the four-dimension open standard — Visibility, Citation Rate, Semantic Coverage, Brand Perception (0-100) — with every answer screenshot-archived.
Deliverables
- Baseline measurement report: four-dimension scores per engine, screenshot evidence appendix.
- Gap analysis: every gap mapped to its root cause (indexation, content, entity, perception) — not a generic checklist.
- 90-day roadmap: interventions sequenced by leverage, each with an owner-ready specification.
- Question-set dossier: the frozen questions, so future measurement is comparable to this baseline.
What does the audit typically find?
Without citing any client specifics, the recurring findings across audits are remarkably consistent: entities fragmented by rebrands, acquisitions or bilingual name drift; key commercial questions with no text answer anywhere on the site; Bing indexation gaps that make ChatGPT blind to the business; review corpora too thin or too old to quote; and structured data that is present but contradicts the visible page. None of these requires genius to fix — they require inventory, prioritization and steady execution.
How the audit is measured — honestly
Our contract language is deliberate: we commit to verifiable targets, not outcomes we do not control. Typical verifiables: Google indexing rate for your canonical page set, valid structured-data coverage, delivery of the agreed question-set measurement on schedule, and Citation Rate coverage measured against the agreed baseline. We do not promise specific AI rankings or traffic numbers — no honest firm can, because the engines are third parties whose answers are stochastic. What we promise is a rigorous baseline, correctly prioritized work, and re-measurement you can audit. The methodology mirrors our public 90-day POC walkthrough.
Who it is for
Businesses that suspect — or want to disprove — that AI answers are sending customers to competitors: local businesses losing near-me answers, B2B firms absent from category comparisons, overseas Chinese businesses whose Chinese-language reputation never made it into Google. The audit is the entry point to everything else we do, and it is designed to stand alone: you can execute the roadmap in-house, with us, or with anyone else.
Process and timeline
Week 1: question-set workshop and baseline runs. Weeks 2-3: entity, content and technical audits in parallel. Week 4: synthesis, roadmap and readout. ChinaGEO is the international line of 海鹦云 (Haipengyun Holding Group); our domestic line covers Doubao, DeepSeek, Yuanbao, Kimi, Tongyi and Wenxin, so bilingual audits span both engine families. Contact: GEO@kchr.com.cn.