FAQ
Questions about Meridian
These answers use published product facts only. For visibility issues engines actually show, see AI search visibility problems.
- What does Meridian measure?
- Meridian measures how a brand is discovered, cited, and recommended by AI answer engines. The unit of measurement is an AI buyer prompt — a commercially relevant question a prospect would ask — not a keyword.
- What is Qualified AI Recommendation Share?
- It is the percentage of commercially relevant buyer prompts for which the company is meaningfully recommended. The secondary metric is competitive share of voice against a named set of alternatives.
- Can Meridian make ChatGPT recommend my company?
- No. Meridian does not sell the idea that GEO can force ChatGPT or any other engine to recommend you. We measure, diagnose, and test. Causal language is withheld unless the evidence supports it.
- What is a GEO scan?
- A scan is a new GEO audit or a remeasurement of an existing audit. Confirming a company profile and rediscovery do not use another scan.
- What does a GEO audit include?
- The workflow is domain → company discovery → competitor discovery → prompt suite → benchmark → website audit → source analysis → gap analysis → GEO score → prioritized action plan.
- Which AI engines can be benchmarked live?
- Live connectors exist for OpenAI, Anthropic, Gemini, and Perplexity when you provide API credentials. Unsupported surfaces such as Copilot or Google AI Mode are never fabricated. Simulated engines are labeled and never mixed into live scores silently.
- How is company information collected?
- You start with a company name and URL. Optional fields improve discovery. OpenAI then runs a forensic investigation of category, buyers, problems, features, competitors, and buyer questions before you confirm and the prompt suite is generated. OPENAI_API_KEY is required for that process.
- How are recommendations produced?
- Every recommendation traces to a prompt, an engine response, a citation, or a page on the site. Findings are labeled observed fact, inference, or recommendation.
- What does Starter cost?
- Starter is $29.90 per month and includes 5 GEO scans. Unlimited is $129 per month with no monthly scan cap. Both include the prompt suite, scorecard, and evidence-linked action plan. You can cancel anytime.
- Why do AI answer engines miss a brand?
- Common, observable causes include a thin or JavaScript-heavy site, missing entity schema, no XML sitemap, an unclear category, an incomplete competitive set, and content that does not answer the buyer prompts engines actually receive. Meridian reports the evidence; it does not claim a single fix will change an engine.