Sample audit

A real AEO audit, published in full

This is a real audit of one of our own sites. fdmini.com is a shop we run, so nothing had to be anonymised and nothing has been. Every figure carries the UTC day it was measured — a real audit is a snapshot of one moment, and so is this page.

Measurements of fdmini.com were taken 2026-08-20–2026-08-22 (UTC); each figure below carries its own date.

Subject
fdmini.com
Assembled
2026-08-22
All dates UTC
UTC

How to read this

A paid audit has ten deliverables. Some are probes and some are judgement, and pretending otherwise is how audit tooling oversells itself. Every section below is tagged with which it is. The four that a person writes are left blank here rather than filled with an example — a specimen of our prose would tell you nothing about your site, and it is exactly the place where a sample page is tempted to invent something.

MeasuredMeasured, then judgedWritten by the analystHow we measure

Where this site stands today

Context first, because a citation rate means nothing without it. fdmini.com is small: Domain Rating 0 on Ahrefs' 0-100 scale, which would place it 102 on our 101-domain SaaS leaderboard — that is, off it. Month to date it drew 17 clicks from 1,729 Google impressions, a 0.98% click-through rate at an average position of 22.8. It ranks somewhere for 174 queries. This is the position most sites are actually in, which is why we picked it.

Domain Rating by Ahrefs, read 2026-08-22. Leaderboard snapshot 2026-08-05.

Positions that moved this month

  • 1/6 scale in inches38 to 1
  • 1/6 doll size39 to 7
  • 1/12 scale in cm31.5 to 11

Average position, first day of the month against the latest. Search Console data through 2026-08-20.

01

Brand AI mention baseline

Measured

5-10 core buyer questions run across ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews, with screenshots of each answer — your starting position, on the record.

5 buyer questions were swept across every engine on our roster on 2026-08-21. 3 came back with fdmini.com cited by at least one engine — a 60% citation rate. The split by engine is below.

Tracked queryCitedSweptNo verdict
Google (Gemini grounding)250
ChatGPT (web search)150
Perplexity150
Claude (web search)250
Google AI Overview152
Google AI Mode150

"No verdict" means the engine answered but the answer carried nothing scoreable either way. It is reported separately and never folded into "not cited" — those are two different facts and only one of them is bad news.

Source: TriRank's own sweep of this site, 2026-08-21. A paid audit runs 5-10 questions chosen with you; these five are the ones this site already tracks.

02

Competitor citation analysis

Written by the analyst

Who the engines cite instead of you on those same questions, and how often each competitor shows up.

Who the engines cite instead of you, on those same questions, and how often each competitor comes back. It needs a category and a competitor set agreed with you before the first query is sent, so the analyst writes it. Deliberately blank here.

03

AI answer source map

Written by the analyst

The domains and page types the answers actually pull from in your category — the shortlist of places worth being. For each tracked query it also names the third-party pages every engine on our sweep roster actually cited, grouped by host with citation counts. We do not broker paid placements.

The domains and page types the answers actually pull from in your category — the shortlist of places worth being. Reading a stack of AI answers and deciding what kind of source each citation is, is judgement, not parsing. Deliberately blank here.

04

Reddit / forum placement opportunities

Written by the analyst

The specific threads and communities the engines are already citing in your space, and which are realistic for you to earn.

The specific threads and communities the engines already cite in your space, and which of them you could realistically earn. Written after reading the answers from section 01. Deliberately blank here.

05

Technical accessibility check

Measured

Whether GPTBot, PerplexityBot and Claude-Web can actually reach your content — robots rules, server-side rendering and page speed.

The homepage answered HTTP 200: 906 ms to first byte, 987 ms in total. robots.txt answered 200 and carries a sitemap directive. None of the four crawler tokens we check is disallowed. The server returns finished HTML — 5,976 visible characters before any JavaScript runs, with no client-side-rendering signals at all.

Crawler tokens checked

  • GPTBot · not blocked
  • ClaudeBot · not blocked
  • PerplexityBot · not blocked
  • Google-Extended · not blocked

Source: the same technical check the free audit runs, executed against fdmini.com on 2026-08-22.

06

Content AI-readiness score

Measured, then judged

How extractable your key pages are: answer-shaped openings, self-contained sections, and the passages an engine could quote.

Measured: 5,739 visible characters across 964 words, one H1, 8 H2s and 25 H3s, with no skipped heading levels. Under our rule-based scoring this homepage takes 97 out of 100 for technical SEO and 58 out of 100 for answer-engine readiness.

Not measured: whether those passages are quotable. Whether a section answers its question in the opening sentence, and whether it still makes sense once an engine lifts it out of the page. That is a read of the writing, not a parse of the markup, and the analyst does it.

Source: homepage parse, 2026-08-22. A paid audit scores the pages you care about, not only the homepage.

07

Schema audit

Measured

Which structured data you have, which is malformed, and which types are worth adding for your page templates.

Present on the homepage: Organization, WebSite. Absent: FAQPage, BreadcrumbList. Nothing malformed. FAQPage is the one worth adding first on a site like this — it is the markup that lets an engine lift a question and its answer together, intact and attributable.

Found

  • Organization
  • WebSite

Worth adding

  • FAQPage
  • BreadcrumbList

Source: homepage JSON-LD parse, 2026-08-22. Homepage only — a paid audit walks one page per template.

08

Brand entity consistency

Written by the analyst

How consistently your brand is described across your own site and the third-party sources engines read — the naming and description drift that confuses retrieval.

How consistently the brand is described across your own site and the third-party sources engines read — the naming and description drift that confuses retrieval. It means holding your wording against other people's, which needs a person. Deliberately blank here.

09

Vertical new-question radar

Measured, then judged

The emerging questions being asked in your vertical that nobody has answered well yet — the openings worth writing into.

Measured: fdmini.com ranks on Google's first page for 54 queries over the last 28 days. 3 of those are being watched for AI citations. 49 have never been swept by any engine, so their citation status is unknown — not negative, unknown. 2 are confirmed gaps: page-one rankings that no engine cites.

Confirmed gaps

QueryGoogle positionImpressions
1/6 scale in cm961
paintable bookshelf7.910

Not measured: which questions are emerging in this vertical that nobody has answered well yet. That comes from reading the category, not from querying it.

Source: Search Console through 2026-08-20, 28-day window.

10

30-day action plan

Written by the analyst

A prioritised, sequenced list of what to change first, with the expected effect and effort of each item.

The prioritised, sequenced list of what to change first, with the expected effect and effort of each item. Everything above exists to produce it, and it is entirely a judgement call. Deliberately blank here.

One thing no tool can hand you

Bing Webmaster Tools publishes an AI Performance report: the queries Copilot cited you on, with volumes. Microsoft's own answer is that it has no API. Someone has to open the console, export it and read it — which is why it belongs in a human audit and not in a dashboard.

22,781
Copilot citations, 90 days
80.8%
Informational intent
0.7%
Commercial intent

Across 90 days on eonebill.ai — published here with that site's permission — that export carried 22,781 Copilot citations, and the intent split was 80.8% informational against 0.7% commercial. The answers it was cited in were overwhelmingly people asking what something means, not people comparing products. The two most-cited queries were both plain definition lookups.

One site, one engine surface, one quarter, and we have not established why it grew. It is evidence about where citations come from, not a rule about where yours will.

Want this run on your site?

Same ten deliverables, your domain, your buyer questions, and the four blank sections above filled in by a person.