SEO Evolution Analyzer

A tool that measures AI-search readiness should be willing to be measured. This post is the raw result of running the seo-ea.com analyzer on seo-ea.com: what the homepage scores today, the fixes we shipped to get there, the points we are deliberately not chasing, and the honest number that still looks bad — our own citation rate of 0%.

What does a 98/100 readiness score actually mean?

The readiness score is a 25% weighted blend of four categories: AEO (answer engines), AIO (AI Overviews), GEO (generative engines), and crawlability. On the analyzer's label scale, scores of 80% and above are "Excellent", 60% to 79% are "Good", 40% to 59% are "Fair", and anything below 40% is "Needs work". The homepage of seo-ea.com currently returns AEO 95, AIO 100, GEO 96, and crawlability 100, which rounds to the 98/100 headline.

A high readiness score is not a vanity metric we tuned. The same page scored 94/100 a few days before this post was written; the difference came from two structural fixes described below. Nothing about the analyzer was changed to make its own page score higher — the checks are the same rule-based signals every URL gets, which is the entire point of eating our own dogfood.

How did we raise our own score from 94 to 98?

The first gap was AIO: the homepage had Organization, WebPage, and FAQPage structured data, but no Article-type schema, so the AIO category topped out below 100. We added an Article block describing the on-page editorial section ("Why AI-search visibility matters") with author, publisher, and publish and modification dates, mirroring exactly the guidance the learning center gives to everyone else. AIO went from 91 to 100.

The second gap was AEO: the page teaches a procedure — analyze a URL — but had no HowTo schema. We added a HowTo describing the real three-step flow visible on the page: enter a URL, run the analysis or citation test, and apply the fix plan. AEO went from 89 to 95. Both fixes followed the same principle we push in the GEO guide: declare what the page is and does in machine-readable form instead of leaving it to an engine's guess.

Every fix we shipped to our own homepage is a fix any site can ship: correct structured data, literal percentages, honest blockquotes, and one descriptive H1.

Where do we still lose points on purpose?

The analyzer still reports two warnings on the homepage, and both are intentional. The first is QAPage schema, which is designed for community Q&A pages where different people post questions and answers — something this page does not have. The second is hreflang, which matters only for multilingual sites; this one is English-only. Marking up either would make the page score higher while being semantically wrong, and a transparency report built on dishonest markup would defeat itself.

There is a broader lesson here that we repeat whenever someone asks why their score is not 100: the goal is not the number. The goal is whether the structured signals honestly describe the page. A perfect score on a page whose schema claims community Q&A that does not exist would be worse for long-term trust than a 98 with two true warnings. Google and the AI engines reward accuracy, not schema count.

Why is our own citation rate still 0%?

This is the honest number in this report, and it deserves a straight answer. A citation test sends realistic buyer-intent prompts (n=4 to 6 per engine) to real LLMs and counts how often the brand is named. seo-ea.com is a young domain with a limited footprint, and citation rate moves with two slow things: indexation, which takes days to weeks, and the training and retrieval cycles of the engines, which take weeks to months. A 0% rate on a site that has existed for weeks is normal, not alarming — the same way it was for every brand the analyzer has tested in its first month.

What matters is that we measure it on a schedule and compare like for like. The methodology in the statistics post applies to us too: run the same prompt set monthly, average multiple runs, and watch the trend rather than any single answer. We publish the current rate here so that when it moves, the movement is visible against a published baseline. If it is still 0% in three months, this post will be updated to say so and we will treat that as a finding, not a footnote.

What does the full site score look like?

The table below shows every public page of seo-ea.com as scored by the analyzer on the day of writing. All pages share the same crawlability result of 100; differences come from AEO, AIO, and GEO signals.

Two patterns are worth noting. First, the pages with genuine article content score highest on AIO, which is why we keep publishing the guides linked below. Second, every page passes crawlability at 100 because the site serves a proper robots.txt that allows all major AI crawlers, publishes llms.txt, and avoids client-side rendering for critical content — the same crawlability gate that failed the site profiled in the JavaScript-rendering case study.

How can you run the same check on your own site?

The rule-based readiness scan is free and unlimited at seo-ea.com, and the citation test needs only a small daily allowance even on the free plan. Paste your homepage, read the crawlability and schema sections as a code review of your identity, and then run the free AI citation checker to see whether engines describe your category correctly today. The same page also explains the scoring weights in the compare tool if you want to see how the analyzer stacks up against alternatives such as Otterly or Peec AI.

If the report finds structural gaps — missing schema, weak headings, crawlability issues — the fix pack generates ready-to-paste JSON-LD and title and description alternatives, which is exactly the workflow we used on our own homepage. The reference vocabulary behind those fixes is the schema.org standard, and Google's guidance on JavaScript rendering and the rendering guidance on web.dev are the sources we trust when deciding what must be server-rendered.

What will we do next?

The sitewide work is captured in our internal audit roadmap, and the short version is: keep publishing content that earns AIO, keep the crawlability gate clean, and re-run the citation test monthly so the 0% becomes a measured trend instead of a single snapshot. If you want the same discipline, the fastest start is a free scan of your most important page and a look at the three biggest signals it flags first. Structural readiness is a habit, not a launch event.

What should you not copy from this report?

Every fix here is public, but the fixes only work when they describe something real. The Article schema on our homepage is valid because the page genuinely contains an editorial section written for readers; bolt the same schema onto a product page with no article-like content and you have created markup that misrepresents the page. The HowTo schema is valid because the three steps it names are literally visible on the page. The lesson generalizes: structured data is a declaration, not a decoration, and engines increasingly penalize declarations that do not match the page they sit on.

It is also worth remembering what a readiness score does not measure. A 98 on structure says nothing about authority, which is built by being mentioned and linked from places engines trust over months. Our own 0% citation rate is the proof: the structure is in place, and the retrieval side is still catching up. If your score is high but your citations are low, the fix is not more schema — it is the slower work of content depth, third-party mentions, and time. Readiness and citation are two different clocks, and this report is deliberately honest about which one is still on zero.

We publish our own number before it is flattering. If you would not do that, your measurement process is not ready to be trusted either.