The Brilad AI Search Audit measures how your website performs in Google search, in answer boxes on the results page, and in AI assistants such as ChatGPT, Perplexity and Gemini. You get a written report and a fix list ordered by likely impact, so your team knows which change to make first. The audit is Phase 1 of the Brilad HEO Framework⢠and takes about two weeks.
What we check
Crawler access and server logs
If you can share server access logs, we filter them by user agent to see which search and AI crawlers actually visit, how often, and which pages they request. User agent strings are easy to fake, so we verify the requests against published IP ranges or reverse DNS before trusting the numbers. We also read your robots.txt line by line against the crawler list for each AI company, because a single broad rule can block an AI search crawler by accident.
Rendering
For each key template (home page, service page, article, product) we compare the raw HTML the server sends with the page after JavaScript has run. Content, links or structured data that only exist in the second version are invisible to crawlers that do not execute JavaScript.
Indexing and technical health
We use Search Console’s page indexing and Core Web Vitals reports, crawl the site ourselves, and check that canonicals, redirects, sitemaps and internal links agree with each other.
Structured data
We extract every JSON-LD block on your key templates and check it for validity, duplicate Organization or WebSite entities (common when a theme and an SEO plugin both output schema), and whether the entities link together through @id references.
Entity consistency
We compare how your company is described on your own site, your Google Business Profile, LinkedIn, directories and any Wikidata or Wikipedia entry. Differences in name, address, founding details or service list make it harder for AI systems to connect those sources to one company.
Content
For your most important pages we check whether each section answers a specific question near its start, whether the page adds information that competing pages lack, and which search queries it should be matched to.
AI visibility baseline
We agree a list of prompts with you that reflect how your buyers ask about your category. Each prompt is run several times on each major assistant, because answers vary between runs. We record whether your brand appears, which URLs are cited, which competitors are named and whether the description of your company is correct. This baseline is what later work is measured against.
What you receive
| Deliverable | Contents |
|---|---|
| Findings report | Each issue explained in plain language, with the evidence (screenshots, log extracts, URLs) |
| Fix list | Every recommendation ranked by expected impact and effort, marked as content, technical or development work |
| AI visibility baseline | The prompt set, raw results per assistant and a summary of mentions and citations |
What we need from you
- Access to Google Search Console (a restricted user is enough)
- Access to analytics, if you use it
- Server access logs for the last 30 days or more, if your hosting provides them
The audit can run without logs, but the crawler findings are then based on testing and robots.txt rather than observed visits.
Background reading
The checks in this audit are explained in more depth in the AI Search Academy:
- AI crawlers and robots.txt
- JSON-LD schema for AI search
- Entity resolution, sameAs and Wikidata
- Retrieval-augmented generation
- AI hallucination and grounding
- Machine readability and clean HTML
- NAP consistency
Frequently asked questions
How long does the audit take?
About two weeks from the day we receive access.
Is this the same as an automated SEO tool report?
No. Automated tools are part of the process, but most of the findings come from manual checks of logs, rendering, schema and AI answers that those tools do not cover.
Can our own team carry out the fixes?
Yes. Each item in the fix list says what needs to change and why, so your developers or content team can work from it directly.