Generative Engine Optimization, Measured
SWIM IN THE SPOTLIGHT
An AI answer names two or three companies and stops. Thysanos measures whether those engines can read your site at all, fixes what blocks them, and re-measures every month, in every language your customers ask in.
The same audit we ran on 16 major European tour operators.
ChatGPT
GPTBot · ChatGPT-User · OAI-SearchBot
Claude
ClaudeBot
Perplexity
PerplexityBot
AI Overviews &Gemini
Google-Extended
Microsoft Copilot
Bingbot
See exactly how AI sees you before your competitors do.
Paste your URL above and a full report opens instantly: your visibility score per engine, what AI currently says about you, and the gaps costing you customers.
01
Your readiness score, out of 100
scored across eight checks
02
Whether AI crawlers can reach you
GPTBot, ClaudeBot, PerplexityBot
03
Your effective score after access
what the engines can actually read
04
The gaps, ranked by impact
with the fix for each
The free report tells you what the engines can read. Measuring what they actually say about you, and who gets recommended in your place, runs on a live prompt set across six engines and comes with an engagement.
AI VISIBILITY SCORE
78
ChatGPT
88
Gemini
74
Perplexity
71
WHAT AN AI READ
"We asked an AI assistant to describe a major German tour operator from its own About page. It returned the site's Google Analytics script. No company name, no destinations, no people."
Real result, July 2026. The site was not blocking anything.
10 blue links
Customers scroll, compare, and click around
One AI answer
The assistant names 1–3 businesses. You must be one of them.
80%
of consumers now rely on AI summaries for at least some of their searches — and most never click past the answer.
SEO got you found on Google. GEO gets you recommended by AI.
Thysanos is the Generative Engine Optimization service of SyeniteLabs. We measure whether AI assistants such as ChatGPT, Perplexity and Gemini can reach, read and cite a website, implement the technical and content changes that fix what blocks them, and re-measure monthly against a fixed prompt set so progress is provable rather than claimed.
WEEK 1
We build a prompt set from how your buyers actually ask, freeze it, and run it across six engines. Then we request your pages as GPTBot, ClaudeBot and PerplexityBot to see what each one really receives. Two scores come out: your technical readiness, and what survives after crawler access.
WEEKS 2-6
Crawler rules, server-side rendering, structured data across ten schema types, llms.txt, entity markup, and content restructured so individual passages survive being lifted out of a page. You get the files and an implementation guide, or we deploy them.
ONGOING
The identical prompt set runs every month against recorded history, so you see the delta rather than a fresh snapshot. Source overlap in AI answers between consecutive days runs 34% to 42%, which is why a single screenshot proves nothing and a fixed monthly measurement proves everything.
THE NUMBERS
The shortlist is being written inside AI, and it closes before you hear about it.
Generative AI chatbots are the single biggest influence on B2B vendor shortlists, ahead of software review sites, vendor websites and peer recommendations. Ninety-five percent of the time, the vendor that wins was already on the shortlist the buyer wrote on day one, before speaking to anyone. A third of buyers now purchase from a company they had never heard of until an assistant named it. Eighth place on a results page still put you somewhere a buyer scrolled past. Eighth place against an answer naming three companies reaches nobody.
of B2B shortlists shaped first by AI chatbots, ahead of review sites at 15.1% and vendor websites at 12.8%.
G2, 2025
B2B buyers bought from a vendor they had never heard of before AI named it.
G2, March 2026
of deals are won by a vendor already on the buyer's day-one shortlist.
6sense
conversion value of an AI search visitor against a traditional organic one.
Semrush, 500+ topics
WHAT YOU GET
Everything your AI presence needs
Four phases, 83 itemised deliverables, a named consultant on your account throughout.
We request your pages as each named AI agent and read what comes back, rather than inferring it from a robots.txt file. Both common failures, a CDN returning 403 and a page that renders empty, are invisible to file inspection.
Organization, Article, FAQPage, HowTo, Product, Service, LocalBusiness, BreadcrumbList, Review and a stable entity `@id` so the markup describes one thing rather than several.
llms.txt, llms-full.txt, agent-card.json, clean markdown mirrors and a parseable pricing file, so agents evaluating you on a buyer's behalf can read what you charge.
Answer-first restructuring into self-contained passages, headings phrased the way people ask, sources and statistics on every claim.
We make sure you get recommended in the language of your clients. Tailor-made optimizations for your region/market.
A frozen prompt set re-run every month against a history database, reported as a trend rather than a snapshot. The baseline is yours to keep and re-run.
WHAT WE WILL NOT DO
The shortcuts we refuse
Never a crawler-only version of a page. Engines detect it and the penalty is removal rather than a ranking dip.
No bought forum posts, no seeded reviews. A poisoned entity is harder to fix than an absent one.
Google names AI-targeted content variants as scaled content abuse. We write for readers and organise for extraction.
GEO is the practice of structuring a website so AI assistants can reach it, extract passages from it, and cite it in their answers. It covers crawler access, machine-readable structure, and whether the wider web describes your brand consistently enough for a model to resolve it.
There is peer-reviewed measurement behind it. A Princeton and Georgia Tech study published at KDD 2024 tested nine optimization methods across roughly 10,000 queries and measured visibility improvements of 22% to 41%. Citing sources, adding statistics and adding quotations were the strongest levers. Keyword stuffing was the weakest and measurably reduced visibility.
Not necessarily, and this is the most common thing we correct. Your CDN or WAF can refuse the same crawler your robots.txt invites. In our 2026 study of sixteen European tour operators, a market leader with the highest technical score in its field scored zero in practice because Akamai answered GPTBot with a 403.
Technical fixes get picked up in weeks and entity work compounds over months. Your first re-measurement lands 30 days after the baseline, which is the first point at which a change is provable rather than anecdotal.
The deployment, yes, and we write the implementation guide for exactly that. What internal teams are rarely set up for is the standing measurement: a fixed prompt set through several engines every month, recorded against history, showing which competitor took your citations.
Quoted per engagement after the free report, because a named consultant works your account and the hours follow the size of the job. Percentage of turnover, milestone and performance-linked structures are available for businesses whose revenue is seasonal or high-volume.
See what AI says about your business in 15 minutes.
A short call, your live audit on screen, and a clear plan. No pressure, no pricing games.
Swim in the spotlight.
Thysanos is the Generative Engine Optimization service of SyeniteLabs.
© 2026 · THESSALONIKI