This is an AI readiness audit example run on my own site, published with the failures intact. I scored solangerainha.com against the five-check instrument I publish, and the results were uncomfortable in a way I didn’t predict: my blog and framework pages scored 8 to 9 out of 10, while the pages that actually get me hired scored 6 and 7.
Publishing a framework is cheap. Anyone can write a scoring rubric, and nobody checks whether the author’s own work would pass it, so I ran my AI Readiness Audit Framework against my own site, scored it the way I’d score a client’s library, and published the numbers before fixing anything.
The headline: a mean of 7.8 out of 10 across the audited pages, with two pages sitting in the salvageable band: both of those are pages a hiring manager lands on, and my highest-scoring page teaches the framework, and my portfolio came bottom.
How I scoped it, and why not every page
The framework prescribes its own scope, and I followed it rather than inventing a more impressive-sounding method. Score the head in full, sample fifteen to twenty pages from the tail, time-box the pass, and fix nothing while scoring.
For this run I scored the head: the pages carrying real business value for someone deciding whether to interview me. That’s the homepage, the CV, the portfolio, and the flagship framework page. I added one blog article as a tail sample, choosing a piece from March 2026 rather than a recent one, on the reasoning that older work shows what my standards looked like before the framework hardened.
Full disclosure on the limits, because an AI readiness audit example that hides its methodology is worthless. Five pages is a head-and-sample pass, not a census of the library, and this audit measures whether a page should be citable. Whether it is cited needs separate tracking, which is a separate exercise with its own scoreboard.
The scores
Five checks, zero to two each, out of ten. Checks 2 and 4 grade substance; 1, 3 and 5 grade structure.
| Page | 1. Front-loaded answer | 2. Citable substance | 3. Standalone sections | 4. Attributed evidence | 5. Retrievable structure | Total | Bucket |
|---|---|---|---|---|---|---|---|
| AI Readiness Audit Framework | 2 | 2 | 2 | 2 | 1 | 9 | AI-ready |
| ICP pivot article (blog) | 2 | 2 | 2 | 1 | 2 | 9 | AI-ready |
| Homepage | 2 | 2 | 2 | 1 | 1 | 8 | AI-ready |
| CV | 1 | 2 | 2 | 1 | 1 | 7 | Salvageable |
| Portfolio | 1 | 1 | 2 | 1 | 1 | 6 | Salvageable |
| Mean | 1.6 | 1.8 | 2.0 | 1.2 | 1.2 | 7.8 |
Three pages cleared the AI-ready band and two landed in salvageable. Nothing scored AI-invisible, which is the one result I’d have been embarrassed by, and the only part of this I’d call reassuring.
The finding I didn’t expect
I assumed the blog would be my weak spot, because it’s the largest surface and the hardest to keep consistent. The opposite turned out to be true.
The pattern is that content pages outperform conversion pages: the framework page scored 9, the sampled article scored 9, and the two pages carrying the most weight in a hiring decision scored 7 and 6. Someone evaluating me typically moves from the homepage to the CV to the portfolio, and that path runs directly downhill in citability.
The reason is uncomfortable and, in hindsight, obvious. I built the blog and the framework pages after developing the AEO methodology, so they inherited it, while the CV and the portfolio were built earlier, as design objects, optimised for a human recruiter scanning quickly. Nobody applies an extraction framework to their own CV, so I optimised the pages that teach the framework and left the ones that get me hired alone.
The pages where I demonstrate the method score 9. Where I actually ask for the job, it’s 6 and 7. I had never once thought to point the framework at my own portfolio.
Where the points went missing
Two checks account for almost all the lost marks, and the pattern is consistent enough to fix in batches rather than page by page.
Check 4, attributed evidence, scored 1.2 out of 2 on average. This is my worst check, and it’s an unflattering result given that the orphan statistic sits at number two on my own list of the five failures that tank a page score. My site is full of orphan statistics, and they’re all mine: “48% marketing-sourced pipeline” appears on the homepage and the CV with no link, no methodology, and no date; same for “+217% reach,” “35% of website traffic,” “+28% YoY organic.”
The irony is that the methodology exists. I wrote a full article explaining how the 48% was calculated, including what it does and doesn’t mean, I just never linked the number to the explanation.
Check 5, retrievable structure, also scored 1.2. This one is a genuine tension rather than an oversight, and I want to be honest about the trade-off instead of pretending it’s a simple fix.
My headings are written in a voice: “Not everything. Just the highlights.” “Where I’ve shipped.” “The toolbox.” “Three systems, built in production.” They’re distinctive, they sound like a person, and they’re part of why the site doesn’t read like a template. For retrieval, they’re close to useless, because they match nothing anyone would ever type and tell a model nothing about what sits underneath them.
The blog article scored 2 on this check precisely because every heading there is a question: “How do you triage an entire content library during a pivot?”
What this AI readiness audit example put on the fix list
The framework says to cross each score with business value and spend the week in the salvageable-and-valuable quadrant.
The CV and the portfolio come first. Both sit at 6 to 7 with the highest business value on the site, which is the exact quadrant the framework says to prioritise. Three fixes each, batched by failure type rather than page by page.
- Source the first-party numbers. Every headline metric gets a link to its methodology where one exists, and a date and scope where it doesn’t. The 48% links to the methodology article, the others get “measured over five months, LinkedIn/Instagram/Facebook combined” style qualifiers inline.
- Fix the front-loaded answer on the portfolio. “Not everything. Just the highlights.” is a stylish non-answer sitting where the answer should be. It stays as the visual headline, with a genuine 40-to-75-word answer block immediately underneath saying what the work is, who it was for, and what it produced. That’s the same discipline I argue for in writing openings that survive the scroll, which I apparently applied to my articles and not to my own portfolio.
- Add question-format subheadings where they don’t cost the voice. Not everywhere; the section labels on the CV can carry a question underneath them, keeping the personality in the display heading and giving retrieval something to grab.
The blog gets one batch pass rather than a rewrite. The tail sample scored 9, so there’s no crisis, but the one check it dropped was attributed evidence for want of a table. The triage in that article is four categories described in prose, which is the “comparison trapped in prose” failure from my own list. Batch fix: find the comparisons across the library and build the tables.
What I’m leaving alone, and why
An audit that ends with “fix everything” is one nobody acts on. These are the deliberate omissions.
- The homepage stays at 8. It’s already in the AI-ready band, and the two points it’s missing sit in the same voice trade-off as everything else. Pushing it to 10 would mean flattening the headings that make it feel like me, for a marginal citability gain on a page people mostly reach by name rather than by question.
- The framework page keeps its heading style. It scored 9 and lost its point on check 5 for the same reason. I’ll take a 9 with a voice over a 10 that reads like documentation.
- I’m not auditing the long tail page by page. The sample told me what I needed: the blog inherited the methodology and behaves. Enumerating 50+ more articles would buy me a slow confirmation of that, which is a mistake I’ve made before and named in the framework.
- The orphan-stat fix doesn’t extend to every number on the site. Where a figure is decorative, adding a source line costs more clarity than it buys credibility.
What running an AI readiness audit example on my own work taught me
Three lessons, and two of them are criticisms of my own instrument.
- The business-value axis needs to be explicit about conversion pages. The framework says to cross the score with traffic, buyer intent, and strategic relevance. On a personal site, the highest-intent page isn’t the highest-traffic page, and I nearly skipped auditing the CV because it felt like a document rather than content.
- Check 5 needs a voice caveat. As written, the retrievable-structure check treats non-question headings as a straightforward deduction. In practice there’s a real trade between distinctiveness and extractability, and a framework that scores you down for having a voice needs to say out loud that taking the deduction is sometimes correct. I’d rather lose a point on purpose than pretend the tension doesn’t exist. The deeper argument for why substance leads and structure follows is in the AEO/LLMO framework, and this is a case where the structure layer should yield.
- The substance checks held up. Checks 2 and 4 are where the framework claims the citation lives, and scoring my own site is the first time I’ve watched that claim get tested against work I couldn’t rationalise. Every page scored 2 on citable substance except the portfolio, which is a catalogue rather than an argument. That’s the framework working as designed: it caught the one page where I have specifics but no position.
Why publish the failures
Because the alternative is a framework nobody can verify, and this whole site is an argument that verifiable beats impressive.
A hiring manager reading my AI readiness audit framework has no way to know whether it survives contact with real work. Publishing an AI readiness audit example run on my own pages, with a 6 out of 10 on the portfolio I built and a self-criticism of the instrument itself, is harder to fake than a case study about a company I worked at. It also demonstrates what the framework is actually for, which is producing a short fix list you’ll action rather than a forensic report you’ll file.
I’d rather show you a 7.8 with the working attached than a 10 with none.
A framework that only scores well when its author applies it to work they chose in advance is a brochure.
If you want to check my scoring, every page is live and the instrument is published. Run it yourself and see whether you get the same numbers. That verifiability is the same principle behind treating content as a system rather than a set of one-off wins.
Scores reflect a pass run in late July 2026 against the published five-check instrument. Pages change; if you’re reading this later, the live pages may already differ from the numbers above.
Frequently asked questions
It looks like a scoring table and a short fix list. Each page gets read against five checks scored zero to two, for a total out of ten, then sorted into AI-ready (8 to 10), salvageable (4 to 7), or AI-invisible (0 to 3). Scores are then crossed with business value, and only the salvageable-and-valuable pages earn the week’s work. On my own site that process produced a mean of 7.8 and a fix list of two pages.
Because they were built earlier, as design objects for human recruiters, before the AEO methodology existed. Conversion pages tend to feel exempt from content rules, so nobody points an extraction framework at their own CV. They lost points on unsourced first-party metrics and stylish non-question headings, both of which are cheap to fix without redesigning anything.
Orphan statistics, scoring an average of 1.2 out of 2 on attributed evidence. Figures like “48% marketing-sourced pipeline” appeared with no link, no date, and no scope, even where a full methodology article existed elsewhere on the site. The fix is linking each number to its explanation rather than producing new content.
No. The framework deliberately ends with a short list rather than a complete one, because an audit that recommends fixing everything is one nobody executes. On this run, two pages went on the fix list, three were left alone with the reasoning documented, and the long tail was sampled rather than enumerated.
