One check a week on how AI assistants read, trust and cite your website. Each Lab takes about 15 minutes, uses only a browser and a worksheet, and ends with a number you write down and date. No login, no tool, no developer. The checks are the 30 that LLM Audit Pro scores, taken one at a time, so that your team can run them, understand them and fix what they find.
Subscribe to The Revenue Signal to get each new Lab in one short email a week: https://revenueexperts.beehiiv.com/?utm_source=academy&utm_medium=lab&utm_campaign=lab-index
.
How the Lab works
- Check of the week (every Thursday). One of the 30 checks: the everyday scene where it goes wrong, the number from our own published research, the exact steps, what a good result looks like, what to do if yours is not, and the honest limit of the check. A one-page worksheet to download and keep as evidence.
- Saved answers (once a month). One question from a published Revenue Experts study, with the saved answers from ChatGPT, Claude and Perplexity shown as they were recorded: which companies were named, whose pages were cited, and how often. You can repeat any of them yourself.
- This index. Every check in order, with the Lab that covers it. Bookmark this page; it fills in week by week.
.
Why numbers and not opinions
Every Lab produces a count, not a feeling. “5 of 7 crawlers allowed” can be checked by anyone on your team, argued with, and fixed. “Our site is fine” cannot. The same rule runs through everything Revenue Experts publishes: the method is public, the answers are saved, and every claim traces back to something you can open yourself.
.
The 30 checks
Points are out of 100, as LLM Audit Pro scores them. Links appear as each Lab is published.
Check | What it asks | Points | Lab |
1. Retrieval Eligibility and Access | Can AI search crawlers reach and read the page at all? | 30 | |
1.1 Indexability controls | Does any tag or header tell crawlers not to index the page? | 6 | Coming |
1.2 Snippet restrictions | Does any setting stop crawlers from quoting the page? | 5 | Coming |
1.3 Search-crawler access evidence | Does robots.txt let the seven AI search crawlers in? | 8 | |
1.4 Raw or server-readable content | Is the text in the page itself, or only drawn by scripts after loading? | 4 | Coming |
1.5 Challenge interference | Does a security challenge page stand between crawlers and the content? | 3 | Coming |
1.6 Internal discovery paths | Can a crawler reach the page by following links from the home page? | 2 | Coming |
1.7 Canonical identity consistency | Does every version of the address point to one canonical page? | 2 | Coming |
2. Buyer Information and Evidence Quality | Does the page answer what a buyer asks, with evidence? | 30 | |
2.1 Task coverage | Does the page cover the questions a buyer brings to it? | 10 | Coming |
2.2 Explicit facts buyers need | Are prices, sizes, dates, limits and specifications stated plainly? | 5 | Coming |
2.3 First-party evidence | Does the page show its own data, tests or documents? | 6 | Coming |
2.4 Claims tied to evidence | Does every claim sit next to the thing that supports it? | 4 | Coming |
2.5 Dates and versions | Can a reader tell when the page was written and last changed? | 2 | Coming |
2.6 Useful structure | Do headings, lists and tables match how the question is asked? | 3 | Coming |
3. Entity, Brand and Trust Consistency | Does the page say who you are, and does every page agree? | 20 | |
3.1 Who the business is | Is the company named, described and located in plain text? | 5 | Coming |
3.2 Facts agree across pages | Do names, prices and claims match from page to page? | 5 | Coming |
3.4 Expertise, methods, accountable people | Can a reader see who stands behind the content and how it was made? | 5 | Coming |
3.5 Contact details | Are address, phone and email present as text, not only in images? | 3 | Coming |
3.6 Named author or owner | Does the page carry a named person responsible for it? | 2 | Coming |
4. Machine Comprehension | Can a machine read the page the way a person does? | 10 | |
4.1 Structured data agrees with the visible page | Does the hidden markup say the same as the visible text? | 3 | Coming |
4.2 Document structure | Is there one main heading and a sensible order under it? | 2 | Coming |
4.3 Labels and values connected | Does each number sit next to the label that explains it? | 2 | Coming |
4.4 Stable URLs | Do addresses stay the same over time, without tracking noise? | 1 | Coming |
4.5 Title and description match the page | Do the title and description describe what is on the page? | 1 | Coming |
4.6 Units and definitions nearby | Are units and terms defined where they are used? | 1 | Coming |
5. Technical Delivery and Usability | Does the page arrive quickly, whole and readable? | 10 | |
5.1 Valid HTML, successful response | Does the page return a normal success response and valid code? | 2 | Coming |
5.2 Server response speed | How long does the server take to start sending the page? | 2 | Coming |
5.3 Content present after rendering | Is the content still there once the page has fully loaded? | 2 | Coming |
5.4 Readable on a phone | Does the page read on a phone without zooming? | 1 | Coming |
5.5 Clear links and buttons | Can a reader and a crawler tell where each link goes? | 2 | Coming |
5.6 Clear next step | Does the page say what to do next? | 1 | Coming |
Saved Answers
Published monthly, from Revenue Experts citation studies. Each entry is one question, the three AI assistants, nine answers each, and the counts. First entry: coming in November 2026. The method behind the studies is public: RE-METH-AIV-003, Case Study Methodology, Public Edition 1.0 (https://revenueexperts.ai/wp-content/uploads/2026/10/RE-METH-AIV-003-Case-Study-Methodology-Public-Edition-1.0.pdf).
.
Worksheets
Every Lab ends with a one-page worksheet. Keep them dated in one folder; together they are your site’s record, and the thing to show whoever asks “are we visible to AI?”.
- Lab 1 worksheet: the seven-crawler robots.txt check (at the end of Lab 1).
.
When you want it done, or want your team to own it
The $55 LLM Audit Pro report runs all 30 checks on one page of your site and shows what to fix: https://llmauditpro.com/?utm_source=academy&utm_medium=lab&utm_campaign=lab-index
The AI Search Visibility Mastery course (40 lessons, $97) teaches the team behind your website the whole method, check by check: https://onlinemarketingacademy.ai/courses/
Subscribe to The Revenue Signal for one practical email a week: https://revenueexperts.beehiiv.com/?utm_source=academy&utm_medium=lab&utm_campaign=lab-index
.
Questions people ask
Is the Lab free? Yes. Every Lab, worksheet and saved-answers entry is free to read and use.
Do I need to run the checks in order? No. Each Lab stands alone. Category 1 is the place to start, because a page crawlers cannot reach fails every other check.
Is this the same as the LLM Audit Pro report? The checks are the same 30. The report runs them on a page in seconds and scores them; the Lab teaches you to run one yourself in about 15 minutes.
Does the Lab show what ChatGPT says about my company? The weekly checks do not; they show whether and how AI assistants can read your site. The monthly saved answers show what the assistants said about the companies in our published studies, so you can see the measurement itself.
.
.
.
.
.
Lab 1: Can the seven AI search crawlers read your site? A 15-minute robots.txt check
By Elizabeta Kuzevska, Co-Founder, Revenue Experts. Reading time about 7 minutes. Doing it for about 15 minutes. Short answer. Seven