Kit / Sample audit

AEO One-Shot: run on our own site

The same audit we would run on your pages, run on kit.sdvsignal.com and both compare pages — with the fixes applied the same evening.

Date: 2026-09-20. Tool: checks/aeo-checks.sh from this product, unmodified. Raw checker output. Findings are pinned to that run; the pages have changed since.

Last updated: 2026-09-21

The scoreboard

Check	Before	After
FAIL findings across the three pages	7	0
`/pricing.md` and `/pricing.txt` for agents	missing (404)	both 200
Compare pages: JSON-LD	none on either	Article + FAQPage + BreadcrumbList on both
Compare pages: comparison table	prose only	a real 7-row table on each
Compare pages: word count	181 and 178	713 and 718
Compare pages: FAQ block	none	5 questions each, schema matching the visible text
Compare pages: question-shaped H2s	0 of 2	3 of 4
Freshness signal	absent on all three pages	visible “Last updated” + `dateModified` on all three
AI crawler access (8 bots)	all allowed	unchanged, all allowed

What was actually wrong, in the order it mattered

The compare pages carried no schema at all. Comparison content is the format most often quoted back by AI answers, and ours was invisible to it. Both now carry Article, FAQPage and BreadcrumbList in one @graph.
Both were under 200 words. Too thin to cite. Each is now ~715 words, and the added words are a table and five real questions, not padding.
No table on a comparison page. A prose paragraph about differences extracts badly; a row with two cells extracts exactly.
Nothing was dated. Undated content loses to dated content when an assistant has to pick.
No machine-readable prices. An agent asked “what does this cost” had to parse the marketing page. pricing.md is now one table of every SKU, price, deliverable and clock, with pricing.txt serving the same bytes for tools that only look for .txt.

What we did not do, on purpose

No HowTo schema. The checker warns about it on every page. Adding a HowTo to a comparison page would be schema that does not describe the page, which is the kind of thing that gets a site ignored rather than cited.
No Article schema on the home page. It is a product page, not an article.
No keyword padding. The word count went up because a table and an FAQ were added. If a fix does not also make the page more useful to a person reading it, it is not a fix.

The mechanical half is what you see above. The judgement half of a real one-shot — ten of your queries checked by hand in AI Overviews, ChatGPT and Perplexity, and which competitor got named instead of you — is not in this sample, because it is about someone else’s pages, not ours.

Can I buy this?

Yes — $99, one document, two business days from a complete intake. No call, no retainer, no monitoring fee. What you just read is the format you get back, run on our own site so you can judge it before you pay.

Buy the AEO One-Shot — $99

The judgement half — ten of your queries hand-checked in AI Overviews, ChatGPT and Perplexity, and which competitor got named instead of you — is the part that is not in this sample, because it is about your pages rather than ours. Nearest neighbour if your question is about your Claude Code config rather than about AI answers: the Claude Code Surface Audit $79.

Questions people ask about this

What is the AEO Visibility One-Shot?

$99, one document, two business days from a complete intake. Ten of your queries hand-checked in AI Overviews, ChatGPT and Perplexity — including which competitor got cited instead of you — up to five pages audited, two real rewrites and an ordered fix list. No call, no retainer, no monitoring fee.

What does an AEO audit actually check?

The mechanical half is a checker run over your pages: whether JSON-LD is present and of a type that describes the page, whether a comparison page carries a real table rather than prose, word count, a visible “Last updated” alongside dateModified, question-shaped H2s, machine-readable pricing at /pricing.md, and whether the eight AI crawlers are allowed. The judgement half is the query check, which a script cannot do.

What happened when you ran it on your own site?

Seven FAIL findings across three pages went to zero the same evening. The compare pages carried no schema at all and were under 200 words each; both now carry Article, FAQPage and BreadcrumbList in one @graph and run about 715 words, with a seven-row table and five real questions. /pricing.md and /pricing.txt were 404 and are now 200. The scoreboard above is that run.

Why not add HowTo schema to every page?

Because a HowTo on a comparison page is schema that does not describe the page, and that is the kind of thing that gets a site ignored rather than cited. The checker warns about it on every page and we left it off on purpose. The same rule is why the home page carries no Article schema: it is a product page, not an article.

Can I buy the One-Shot?

Yes. It is $99 on the Kit page and the checkout is live. This page is the proof — the same format you get back, nothing gated. Two business days from a complete intake.