AI visibility audit
Black Slate Roofing
Black Slate Roofing: what an assistant can read, and what it cannot check
Partly measured on 17 September 2026. 15 live pages crawled, plus a direct request as each of 6 AI crawlers.
What this comes down to
13 things make Black Slate Roofing's website harder for an assistant to find, read and check, 6 of them urgent. The first is: No sitemap. 3 of the things a buyer would go and check are missing, starting with: The site never says what anything costs.
Read
Can an AI read this website?
This is the foundation. Before an assistant can say anything about a business, it has to be able to fetch the pages and find real text in them. None of this is about being liked or ranked: it is about whether the door opens and whether there is anything readable behind it.
- Fails, priority A+. No sitemap No sitemap.xml and none declared in robots.txt, so pages are found only by following links that point at them.
What this means
A sitemap is the list of pages you are telling machines you have. Without one, discovery leans on links, yours and other people's, so a page nothing points at is the one most likely to be missed.
- Fails, priority A+. No structured data anywhere Not one of the 15 pages carries JSON-LD, so nothing on the site states outright what the business is, what it sells or who is behind it. Those facts are left to be inferred from the prose.
What this means
Structured data is the machine-readable version of the facts already on your pages: who you are, what you sell, where you work. Without it those facts have to be inferred from prose, and inference is where you get described as something you are not.
- Fails, priority A+. Every address returns the same page 8 of the 23 addresses requested came back as the same document as a made-up address: the same visible text, within 200 bytes of the same length. The site answers everything with one page. One fault, and it costs several things at once: no inventory of pages can be trusted, mistyped and stale links look like real content, and anything crawling the site is fed addresses nobody ever wrote.
What this means
One document answering every address means there is nothing page-shaped to read: an assistant asking for a service page, a location page or a typo gets the same few words back, and cannot tell any of them apart.
- Fails, priority A. robots.txt is there but says nothing a crawler can use The file returns 200 but the body is the site's HTML page, which is the error page wearing a filename. A does-it-exist check passes this and a read does not.
What this means
Without robots.txt a crawler has no instruction file to read. It is also where the sitemap is declared, so its absence usually costs both.
- Fails, priority A. Duplicate titles "Black Slate Roofing" on 2 pages
What this means
Two pages with the same title look like the same page. Anything choosing between them has nothing to choose on.
- Fails, priority A. Pages with no meta description 15 of 15: /, /?page_id=81, /?page_id=91, /?page_id=107, /?page_id=141, /?page_id=133 and 9 more. This is the sentence an engine quotes when it lists them.
What this means
The meta description is often the sentence quoted about the page. Without one, the quoting engine takes whatever text it finds first, which is frequently a menu.
- Fails, priority B. Pages with no canonical tag 1 of 15: /index.php?rest_route=
What this means
The canonical tag is where you say which address is the real one. Without it a search engine picks one itself, and the same page reached two ways can end up counted twice.
- Fails, priority B. Pages with no Open Graph tags 15 of 15. Nothing on the page sets the title, description and image used when one of its addresses is shared, so whatever is quoting it chooses for itself.
What this means
Open Graph tags are how you choose the title, description and image shown when one of your addresses is shared. Without them, whatever is quoting the page chooses for you.
- Fails, priority B. No clear primary heading 13 page(s) do not state their subject in one main heading: 13 where several compete for it. /, /?page_id=81, /?page_id=91, /?page_id=107, /?page_id=141, /?page_id=133 and 7 more
What this means
The H1 is the page's own statement of what it is. None leaves it unstated, several leave it ambiguous.
- Fails, priority B. Pages with almost no text 2 page(s) under 150 words of visible text: /?page_id=151, /index.php?rest_route=. Counted from rendered text, not markup.
What this means
A page with almost no words has little to answer with, so it is rarely the thing that gets quoted.
- Fails, priority C. llms.txt is there but there is nothing in it to quote The file returns 200 and the body is the site's HTML page, so anything reading it gets the website, not a description of the business.
What this means
llms.txt is a short plain-English summary of what the business is and what it sells, written for assistants. It is not yet a standard and nobody should promise results from it, but it costs one file and it is the one place you get to state the business in your own words.
- Fails, priority C. Title length off 3 page(s) outside 30 to 65 characters: /, /?page_id=561, /index.php?rest_route=
What this means
Very short titles carry no information, very long ones get cut mid-phrase.
- Fails, priority C. Heading levels skip 15 page(s), for example /: h2 to h4, h1 to h5
What this means
Headings are how a machine works out which parts of a page belong to which. Skipping levels leaves that hierarchy less clear than it could be.
- Not checked. Whether the text is in the HTML before JavaScript runs not checked Comparing what a fetch returns against what a browser ends up with needs the page opened in a real browser, and that was not done on this run.
What this means
Some sites send an almost empty page and then draw the content with code once it is open in a browser. People see a full page. An assistant that only reads what was sent sees the empty one.
8 checks pass
- Passes. Those crawlers actually get in Asked for the homepage as each of the 6 in turn, and every one was let in with a 200. This is a separate check from robots.txt: a rule at the CDN can block a crawler the file allows.
- Passes. Every live page has a title.
- Passes. No description is repeated across pages.
- Passes. Images carry alt text.
- Passes. Every structured-data block parses.
- Passes. No tracking tag is double-counted.
- Passes. Pages link to each other, so the site can be walked.
- Passes. Every live page has something linking to it.
Trust
Can a buyer check what this site says?
A site can be perfectly readable and still not be believed. This step compares what the site claims against what a stranger can verify: a number behind every claim, a person behind the business, and the same details wherever the business appears.
- Fails, priority A. The site never says what anything costs No price, rate, retainer or fee appears on any of the 15 pages crawled. Custom work is a real reason not to publish a fixed price, and a range or a starting number is still a number.
What this means
A buyer cannot judge an offer without knowing what it costs, and a site that does not say invites them to ask elsewhere.
- Fails, priority B. No Google Business Profile linked Linked from the site: Facebook, Instagram. A Google Business Profile is not among them, and for a business people search for by place it is the one that decides whether they appear at all.
What this means
For anything local, the Business Profile is the record most often read back. Unlinked, it is harder to tie to the site.
- Fails, priority C. Nobody is named in a way a machine can read No Person markup anywhere on the 15 pages crawled. Names may be written in the prose, and prose is where an assistant has to guess which name is the founder and which is a testimonial.
What this means
Structured data is the machine-readable version of facts already on the page. Without it those facts have to be guessed from prose, and guessing is where a business gets described as something it is not.
- Not checked. The site states no rating and no review count not checked Nothing on the crawled pages claims a rating or a number of reviews, so there was no claim to check against a source.
What this means
A rating on your own website is a claim, not evidence. What makes it evidence is a link to the place the reviews live, where a stranger can count them.
- Not checked. Whether a licence number is published not checked This business does not need a licence to operate, so there is no number to look for.
- Not checked. Whether the profiles match the site, field by field not checked No profile links were given, and searching for a profile and assuming it belongs to this business is how an automatic report says something false about someone.
What this means
Where a name, address, phone or list of services differs between a website and its profiles, the difference is what gets noticed, and it makes every other claim look less certain.
2 checks pass
- Passes. The site links to a review or business listing, so its claims can be checked somewhere else.
- Passes. Enquiries arrive saying what they are about.
Recommend
Would an assistant recommend this business?
Read and Trust are facts about one website on one day, and anyone with a browser can check every line above. Whether an assistant actually recommends a business is a different kind of question: one answer proves nothing, because it can change tomorrow. Answering it properly means asking the questions your buyers ask, week after week, watching who gets named, and telling a real shift from noise. That is the paid work, it is not part of this report, and this report is complete without it.
How this was measured
The website was read on 17 September 2026. 15 live pages crawled, plus a direct request as each of 6 AI crawlers. Every check above is a fact from that crawl or from a request made to the site, and each carries the evidence it came from.
Nothing needed an account, and nothing was measured that a visitor cannot see.
Nothing here was measured that is not publicly visible, and no account access was used. The site read for this report was https://blackslateroofing.com, observed on September 17, 2026.
Black Slate Roofing: if you would rather this came down, email support@areyouvisible.co and it is removed the same day, no questions asked.