What the Brutally Honest audit checks

The non-commodity test. Does this page contain anything an AI couldn't have written by itself without ever knowing your business? Real prices. Actual service areas or delivery times. Photos you took. Your guarantee. Opening hours. A customer's own words. If the answer is no — if the page is well-written but generic, the sort of thing any competitor could publish word for word — it fails, and the grade is capped no matter how good everything else is.

Two kinds of AI crawler, and people block the wrong one. Training crawlers gather text to train future models — blocking those is a legitimate business decision. Search crawlers fetch your page right now to answer someone's question right now — blocking those removes you from those answers entirely. They're easy to confuse and often blocked together by accident. The audit reports which is which.

The title tag and meta description. The description doesn't affect ranking at all — it affects whether anyone clicks. A page can rank well and still never get visited because the description is blank, duplicated, or says nothing specific. The audit checks both, and flags a title that still says "Home".

AI-quotability. Hiding invisible text aimed at manipulating AI systems is an automatic, unappealable F-. The audit checks whether the page's content is structured so a machine can quote it cleanly: specific numbers rather than vague claims, facts in real tables, and sections that still make sense when quoted on their own.

Heading structure. A designer wants the footer links slightly bigger, so they mark them up as headings. To a person it looks fine. To a machine reading the structure, your page about your main service appears to have thirty-nine sections, most of them named after footer links that have nothing to do with the topic. The audit reads the outline the way a machine does.

The audit reports speed and must label clearly whether it's quoting real Chrome user data or a lab simulation, because the two often disagree dramatically. A lab run on a fast connection can look excellent while real users on mid-range phones on cellular data see something far slower. The audit refuses to conflate the two.

Schema markup. Scored low on purpose because the research shows adding these labels doesn't by itself get you cited more by AI. The audit also warns that marked-up reviews that don't appear on screen can trigger a Google penalty — markup must match what's visible.

The share preview. When someone pastes your URL into a message or a post, the title, description and image that appear are what they actually see of your business. The audit checks that these are specific and correct, not blank or generic.

Images, links, and accessibility. Alt text on meaningful images, descriptive link text (never "click here"), visible focus rings, and sufficient contrast. These are worth fewer points because a page can pass them all and still be a blank brochure to an AI — but failing them is a sign of carelessness the audit will name.