How to run a GEO audit.

Can the intended page be crawled?

Check the actual URL, not just the homepage.

Fetch the public page and record the final status, redirect destination and response headers. Read robots.txt for the search crawlers you intend to admit. A homepage allowance does not mean the guide, research page or its stylesheet is allowed. Check the relevant path and its rendering assets.

If the request times out or returns an access challenge, record the check as unavailable. You have not established that content or markup is missing. Verify CDN and hosting rules before changing the page.

If logs are available, record and verify the crawler request before inferring why a page was not cited.

Is the page eligible for indexing?

Crawl permission and indexing permission are separate.

Inspect the robots meta tag and X-Robots-Tag header. Confirm that the canonical URL identifies the public page you want search engines to use. Put that URL in the sitemap and link to it from a relevant public section. Keep test environments, private reports and unfinished studies outside the indexing set.

Google’s AI search features use the same foundational search requirements. The page must be indexed and eligible for a snippet; there is no special AI schema or required AI text file.

Does the page answer a real question?

Start with the visitor’s task.

Choose a question that the page can answer with direct experience or evidence. Put the answer, scope and important conditions in visible text. Use descriptive headings and links so readers can follow definitions, methods and supporting material. Check the server HTML as well as the rendered page.

For Gist’s CRM study, the question is whether being cited means being recommended. The page shows the precise prompts, cited publishers, suggested products and number of saved answers. It does not turn selected examples into a category-wide claim.

Can a reader check the evidence?

Put the source beside the finding.

Name the author or publishing organization, explain how an observation was collected and distinguish collection dates from publication dates. Link to primary sources or a methods appendix. For a chart, show its denominator, counting rule and limitations. Make supporting text readable without requiring an interaction.

Add structured data that describes what is actually on the page. Article markup can identify an article; FAQ markup must mirror visible questions and answers. Do not invent reviews, authors, outcomes or claims to fill a schema field.

Does the site describe the entity consistently?

Make it clear who and what the page is about.

Check your organization name, product name, official URL and description across the homepage, About page and factual reference page. Explain ambiguous names. Link the evidence back to a stable entity page and keep the visible facts consistent with its structured data.

A factual entity page and a question-focused guide serve different purposes. Gist’s facts page identifies the product; the research and guides address specific questions. Publishing more copies of the same description does not add useful evidence.

Did the change improve the outcome you care about?

Keep technical verification separate from answer measurement.

After publishing, verify the served HTML, canonical URL, assets and internal links. Submit the production sitemap in Search Console and request indexing for a few priority URLs where appropriate. Inspect indexing status over time. Keep the release date and exact changes in your record.

For AI visibility, compare saved answers to a consistent question set. Track citations, mentions and product recommendations separately. Record the engine, collection method and repeated-answer counts. A fixed source list or repeated answer text should not be mistaken for independent confirmation.