Free audit
Is your site ready for AI assistants?
More and more people no longer work through ten blue links. They ask ChatGPT, Gemini or Perplexity directly. Type your address and we will look at 20 things that decide whether an assistant can find your site, understand it and recommend it. You see the result straight away, with no sign-up.
Free, with no limits. We do not ask for your email address until the result is on the screen.
Can an AI reach the page
AI crawler rules in robots.txt, HTTPS, the sitemap, and whether the content is already in the server response or only appears in the browser.
Can it tell who you are
The schema.org company record, how complete it is, and whether the name and contact match the business register.
Is there an answer on the page
llms.txt, the heading structure, how much content there is, a question and answer block, and how fresh it all looks.
Is there a way to recommend you
Contact details as text, the page language, the description, and whether the site is more than one page.
An honest note: this audit measures readiness, not citations. It looks at whether your page is readable and understandable to a machine - not at what ChatGPT says about you today. robots.txt is a request, not a lock: some crawlers do not honour it. And a good result is necessary but not sufficient, because part of this depends on how much other people talk about you.
Why this is a separate question, not just SEO
A search engine hands over a list and the visitor chooses. An assistant hands over one answer, and the choosing is already done. To be on the list it is enough that the page is good. To be inside the answer, a machine also has to be able to read your page, identify your company beyond doubt, and find the answer on it. Those are three different things, and none of them happens by itself.
What the audit actually looks at
AI crawler rules
We read your robots.txt and check whether GPTBot, ClaudeBot, PerplexityBot, Google-Extended and the rest are allowed or turned away. Plenty of sites have shut them out by accident, after copying rules from somebody else's blog.
Structured data and llms.txt
We look at whether the page carries a schema.org record about your company, and whether it is actually filled in. We also check for /llms.txt - a short file that tells an assistant, in your own words, what you do.
Details that agree
We compare the name, email and phone on your page with what the Estonian business register holds. When sources disagree a machine picks one, and you do not get to know which.
Whether the content is there at all
We check whether the text arrives with the server response or is only loaded by JavaScript. Most AI crawlers do not run JavaScript, so in the second case they see a blank page.
The shape of an answer
The heading structure, how much content there is, and whether there is a question and answer block. What gets quoted are passages that stand on their own, and a subheading is what makes a passage stand on its own.
Freshness
We look at the latest dates in your sitemap and the year in your footer. Freshness is one of the strongest signals for whether an answer is worth taking from you.
If you would rather somebody did the work
The audit tells you what is out of place. From there you have three routes, and you do not have to choose right away - start with whatever matters most for the state your site is in.
One-off work
Fix-it package
The whole list worked through and done, not just handed over as a report.
- schema.org structured data for the company, the services and the questions people ask
- llms.txt and the AI crawler rules in robots.txt, set as a deliberate choice
- product feeds, if you run a shop
- a clarity pass over the content: does the first paragraph answer the question
- facts brought into line - the same name, address and prices everywhere
Ongoing service
AI visibility monitoring
A regular check of what the big assistants say about your company, and who they recommend instead of you.
- a fixed list of the questions your customers actually ask
- the answers written down, so change can be compared
- who else in your field turns up in the answers, and for what
- your site's agent readiness measured again
When the groundwork is missing
Content and technical groundwork
If the audit says there is too little content, or that the page does not render without JavaScript, the fix is bigger than markup.
- content written as answers, and question and answer blocks
- server rendering, so a crawler sees what a visitor sees
- the classic SEO groundwork, without which AEO does not work either
It does not ask ChatGPT what it thinks of you - that number would swing from one day to the next, and for Estonian-language prompts the answer would usually be empty. It measures readiness: whether a machine can reach your page, tell who you are, and find an answer on it. robots.txt is a request, not a lock. llms.txt is cheap insurance, not a magic wand. And even a flawless page does not guarantee a place in an answer, because part of it depends on how much other people talk about you. If somebody promises you a guaranteed place in an AI answer, be careful.
Frequently asked questions
How ready your site is for a reader who is not a person but an AI assistant. The assistant has to reach the page, work out whose page it is, and find on it an answer to the question somebody put to it. Agent readiness is the sum of those three.
Yes. You see the score and the two most important findings straight away, without entering anything. The rest of the findings, with instructions for fixing each one, we send by email - that is the only place we ask for an address, and you can see what you are giving it for before you do.
We send you the result of the audit, and we get in touch if it looks like we can help. We do not sell addresses on and we do not add you to a newsletter. Every email can be unsubscribed from in one click, and you can ask us to delete your data at any time.
We read your public homepage, your robots.txt and your llms.txt - exactly what any search engine does. We store the result of the audit so that we can send you the report and you can pass the link on. The content of the page itself we do not keep.
Technically yes, and that is more than most sites manage. But a good result means a machine CAN use your page - not that it will. The other half is decided by how much you are talked about elsewhere: reviews, directories, press coverage and links from other sites.
Usually not as bad as the number looks. Most findings are a matter of hours to fix, not weeks - a line in robots.txt, a structured data record, a couple of subheadings. The one finding that means real work is "content only appears in the browser": that is a technical change to how the site is built.
No. Classic Google search still brings a large share of the traffic. Treat agent readiness as an extra layer: first you need proper SEO groundwork and clear content, and then comes the work that makes the same information unambiguous to a machine. Much of that work is shared.
Yes, the audit only reads public information. Running it on a competitor's site is a perfectly ordinary way to see where you stand next to them.
Also free
Classic search still brings a large share of the traffic. Check whether Google finds your site too, and whether it can tell what the site is about.