Your website probably looks fine. It loads, the photos are decent, the phone number is correct, and it did its job for a decade. But "looks fine to a human" and "readable by the machines deciding who gets recommended" are now two different tests, and most established Canadian companies have never run the second one. This is that test: a self-serve audit you can work through in an afternoon, no developer required, that answers two questions. Can machines read your site? And does it answer the questions your buyers actually ask?
Quick answer: An AI-era website audit checks five things: whether AI crawlers can access and read your pages, whether structured data states your facts in machine-readable form, whether your content answers real buyer questions directly, whether your business identity is consistent across the web, and whether the leads your site earns actually get captured and followed up. Most established company websites fail on crawlability and content depth, and both are fixable without rebuilding from scratch.
Why does a website that looks fine fail an AI-era audit?
Because the audience changed. Ten years ago your website had one reader: a human with eyes, patience, and context. Today it has two, and the second one, the AI crawler, decides whether you appear in the answers buyers now read before ever visiting a website. A crawler cannot admire your hero photo. It reads text, structure, and data, and if those are thin, you are thin, no matter how sharp the design is.
The pattern we see in audit after audit: a company with 20 years of real expertise and a website that communicates almost none of it. Products with no specifications on the page. Services described in three sentences of brochure language. No answers to the questions the sales team fields every single day. When ChatGPT or Google AI Mode goes looking for a source to cite, it finds nothing to work with and cites a competitor instead. Not a better company, a more readable one.
Can machines actually read your site? The technical checks
Start with the plumbing, because if crawlers cannot get in, nothing else matters. Each of these takes minutes to check:
- robots.txt. Visit your-domain.ca/robots.txt and look for blocks on AI crawlers such as GPTBot, ClaudeBot, or PerplexityBot. Some default configurations turn them away without anyone deciding to. Our guide to AI crawlers and robots.txt walks through exactly what to look for.
- Text that exists without scripts. If your site renders everything through scripts, some crawlers see an empty page. Quick test: view the page source and check whether your actual product text appears in it.
- Schema markup. Structured data that states your business name, services, products, and service area as facts. Paste a key page into a schema validator; if nothing comes back, machines are guessing about you. Plain-English detail in our schema markup guide.
- An llms.txt file. A plain-text summary of who you are and what you sell, written for AI systems. An emerging standard, cheap insurance, one hour of work. See our llms.txt explainer.
- Page titles and headings that say something. "Home" and "Products" tell a machine nothing. "Skid steer grapples for 60 to 90 hp machines" is a fact it can use.
Does your site answer the questions buyers actually ask?
This is the half of the audit most owners skip, and it is the half that decides citations. Write down the ten questions your sales team answered most often last month. Real ones, in the buyer's words: what fits what, what it costs to run, how long it lasts in a prairie winter, what the lead time is, what the warranty covers, how freight works. Now search your own website for the answers.
Most established companies find two or three answered, seven or eight missing entirely. That gap is exactly what an AI engine sees. Here is the difference between the two readers your site now serves:
| Audit area | What a human visitor sees | What an AI crawler sees |
|---|---|---|
| Design and photos | Professional, trustworthy look | Almost nothing; images carry little citable meaning |
| Three-sentence service page | A tidy overview | Too thin to cite for any specific question |
| Specs in a PDF download | Available if they dig | Often unread; facts locked away from the answer |
| Buyer guide with question headings | Genuinely useful reference | Liftable answers, quoted nearly verbatim |
| Schema and llms.txt | Invisible | Ground-truth facts about your business |
| News page last updated 2021 | Mild doubt | A staleness signal weighing against citation |
Want this audit done properly, with your real buyer queries?
Our free AI visibility audit runs the questions your buyers ask across ChatGPT, Perplexity, and Google AI answers, checks your site the way the machines read it, and hands you the fix list in priority order.
Book Free AuditDoes your identity check out across the rest of the web?
Engines corroborate before they recommend. If your website says one thing and the rest of the internet says another, or nothing, you lose confidence points you never knew existed. Three checks:
- Entity consistency. Your exact business name, address, phone, and description should match on your site, your Google Business Profile, directories, and social profiles. "Acme Mfg." in one place and "Acme Manufacturing Ltd." in another reads as two uncertain entities, not one solid company.
- Reviews with substance. A steady flow of reviews that mention specific products and outcomes beats a wall of five-star ratings with no words. Machines read the words.
- Third-party mentions. Association memberships, supplier pages, dealer listings, local press. Ask an engine "what do you know about [your company]" and see whether the answer is accurate, vague, or confuses you with someone else.
Does your site capture the leads it earns?
An audit that stops at visibility misses the expensive part. Every gap between "buyer found you" and "somebody responded" leaks revenue, and established companies leak more than they think because nobody owns the checking. Test it yourself, today: submit your own contact form and time the response. Call the main line at 4:55 p.m. on a Friday. If the form takes two days and the call rings out to voicemail, your marketing budget is filling a leaky bucket.
The fixes are systems, not effort: instant acknowledgment on every form, missed-call text-back so an unanswered ring becomes a conversation, and an AI receptionist for after hours. We covered the response-time math in speed to lead for Canadian B2B, and the broader toolkit lives on our AI automations page. One recovered order typically covers the entire fix.
How do you score the audit and what do you fix first?
Fix in this order, because each layer depends on the one before it. First the technical plumbing: robots.txt, rendering, schema, llms.txt. It is one-time work, mostly done in days, and everything else is wasted without it. Second, lead capture, because new visibility pouring into an unanswered phone helps nobody. Third, content depth: publish real answers to your buyers' top questions, steadily, in your own voice. Fourth, identity and proof cleanup across directories and reviews.
Then re-test quarterly. Ask the engines your buyer questions, note who gets named, check your response times. The companies pulling ahead in AI search are not smarter; they are the ones who measure. If the audit reveals a site too old to patch, that is a different conversation, and we wrote about it in whether your old website is costing you sales.
Frequently asked questions about AI-era website audits
What is an AI-era website audit?
It is a review of your website against the two audiences that now read it: human buyers and the AI systems that decide which companies get named in answers. It checks crawlability, structured data, content depth against real buyer questions, identity consistency across the web, and whether inbound leads actually get captured and followed up.
How do I check if AI crawlers can read my website?
Start with your-domain.ca/robots.txt and look for rules blocking crawlers such as GPTBot, ClaudeBot, or PerplexityBot. Then view the source of a key page and confirm your product text actually appears in it rather than being rendered only by scripts. Finally, run a page through a schema validator to see whether structured data exists at all.
My website looks modern. Why would it still fail this audit?
Because design and readability are different tests. A machine cannot be impressed by photography; it reads text, structure, and data. A beautiful site with three-sentence service pages, specs locked in PDFs, and no schema gives an AI engine nothing to cite, so it recommends a competitor whose site states facts a machine can lift.
What is llms.txt and does an established company need one?
An llms.txt file is a plain-text summary at your site root, written for AI systems, stating who you are, what you sell, where you operate, and where your key pages live. Adoption is still emerging, so treat it as cheap insurance rather than a magic switch: about an hour of work that forces useful clarity and costs nothing to maintain.
Can I run this audit myself or do I need an agency?
You can run the core checks yourself in an afternoon: robots.txt, page source, a schema validator, your buyers' top ten questions against your own pages, a test form submission, and a test call. An agency adds the parts that are hard to do alone, like running real buyer queries across multiple AI engines and benchmarking you against named competitors.
How often should an established company audit its website?
Run the full audit once, fix in priority order, then re-test quarterly. AI search behaviour and engine features are shifting fast enough that an annual check is no longer enough, but quarterly is sustainable and catches drift: new pages that shipped without schema, a redesign that quietly blocked crawlers, or response times that slipped.
What does the AlphaPixels AI visibility audit include?
We run the actual questions your buyers ask across ChatGPT, Perplexity, and Google AI answers, record which companies get named, review your site the way machines read it, and deliver a prioritized fix list. It is free, it applies anywhere in Canada, and it gives you a baseline you can hold any agency accountable to, including us.
Related reading
- Content Clusters for B2B SEO: How Canadian Companies Build Topical Authority
- Is Your Old Website Costing Your Established Company Sales?
- Answer Engine Optimization for Canadian B2B Companies: The 2026 Guide
The bottom line on auditing your website for the AI era
Your website was built for a reader who no longer arrives first. The machines that now stand between you and your next buyer read structure, facts, and depth, and they cannot cite what they cannot read. The good news is that nothing in this audit requires a rebuild of your business: the expertise already exists, the fixes are known, and most of the technical work is one-time. The companies falling behind are not the ones with old websites. They are the ones who never checked.
Run the checks yourself this week, or let us run the full version for you: start with the free AI visibility audit or book a free fit call with AlphaPixels and we will walk through exactly where your site stands and what to fix first.