AVE Studio · AI Visibility Engineering
A3 · Stage 2 — Fix: the diagnosis
Gap Analysis · GEO six-layer root-cause analysis
Being found, included and cited when an AI generates an answer — at answer time, from the live web.
Generative Engine Optimization: the definitionYou already know the AI doesn't recommend you — but not why. We uncover it question by question and issue the fix list — the working sheet of AI search optimization: what to rewrite in the code, and what in the content.
The machine-equity loop — Gap Analysis
- 1 · MEASURE — “Where do I stand today?” Per engine, with a range (Citation Tracker, Selector)
- 2 · FIX — “Why do they skip me?” Page by page, then fixed in code (Gap Analysis, Gap Closure)
- 3 · ANCHOR — “Where does the machine learn about me?” An unambiguous entity, present where models learn (Entity Anchor, Corpus Campaign)
- 4 · PROVE ↺ — “Did it move, and what did it bring?” Closed-book baseline and the delta at every new model generation (Model Memory Audit, Release Audit)
KG Doctor and the Claim Coherence Engine — the studio's own instruments — serve the entire loop.
Machine equity is the business asset that arises from AI systems being able to recognize, understand, recall, recommend — and, on a user's behalf, transact with — a brand. Composed of three layers: rented reach (retrieval visibility), owned memory (trained model recall), and machine buyability (agent-side transactability). Measured honestly only as a portfolio of ranged, per-layer metrics — never as a single blended score.
Machine equity: the definitionThe AI usually doesn't recommend another provider by accident: their content, data or site structure is measurably a better fit for the question. For every approved target question, the Gap Analysis identifies the competitor the AI engines most consistently recommend instead of you. We then examine that competitor's actual pages in detail and show, with evidence, what they do differently — and what you're missing.
01Content
Does your page answer the buyer's question in a form the AI can easily interpret and cite?
02Vocabulary
Do you use the same terms your buyers and the AI use to name the subject?
03Structured data
JSON-LD is the description written for machines. Does it clearly tell the AI what each element of the page means?
04External credibility signals
Which trusted sources link to you, and how credible do you appear to the AI as a result?
05Page type
Are you answering the question with the right type of page — or competing against a detailed service page with a blog post?
06Internal linking
Do your internal links lead the AI to your most important page, or is it buried in the structure?
We separate facts from assumptions
Every finding is tied to a concrete source: your own page, a competitor's page, or the measurement result. Anything without direct evidence is explicitly labeled "assumption" and never stated as fact. This reduces the risk of false, invented conclusions.
not a generic advice list
We examine the actual content and structure of the pages recommended instead of you, and document their practice with evidence. If the AI answers on a question are too volatile, we say so plainly: "no clear winner". We don't draw firm conclusions from uncertain data.
Every service runs on an application we built in-house for exactly this job — not manual spreadsheets, not general-purpose tools.
A six-dimension, evidence-linked analysis engine: every finding is tied to a source; anything unproven is explicitly labeled an assumption.
- Root-cause analysis per question: shows why the AI passes over your page. Findings across the six dimensions are ranked by severity.
- Competitor evidence: page by page, we document what the competitor the AI most consistently recommends instead of you actually does.
- Site-level technical findings: robots.txt, llms.txt — the file that guides how AI systems crawl your website — and the sitemap.
- A client-ready report: available on a shareable link, with the measurement date in the header so it's always clear what vintage of data it's built on.
- Analysis grounded in published research: we incorporate documented AI-visibility findings — such as the Princeton GEO research — with source attribution.
- The basis of the next step: the findings feed directly into the Gap Closure's executable fix package.
Reports are delivered in Hungarian and English.
How much does the gap analysis cost?
From HUF 170,000 net (approx. €425). The premium consulting market sells a "comprehensive audit + strategy" document at a multiple of this; here you get a page-by-page, question-by-question diagnosis measured against the actual AI answers, in six layers.
What do I receive?
A page-level, question-level breakdown of exactly where and why the AI passes you over: which page fails, in which layer (access, content, structure, entity), and who gets cited instead of you. Not a generic recommendation list — evidence, with addressed faults.
How is this different from generic AI-audit consulting?
The consulting genre writes documents: general recommendations and best practices. We measure your pages on your buyers' questions, and every finding comes with the actual AI answer it is visible in. The difference is evidence instead of opinion.
Why find the causes now? Until you know which layer your page fails on, every fix is guesswork — and development spent in the wrong place costs twice. The diagnosis is one-off work; its result underpins every step after it.