A Guide for Decision-Makers — demo on mercator-ocean.eu
No technical background required · 5 minutesThis short guide helps you explore the demo on your own. It explains what you are seeing, why it matters, and how to read the results.
This is not a slide deck. It is a live audit pipeline running on a real website — mercator-ocean.eu — in front of you. You will watch the machine collect, structure, analyse and report, step by step.
We call our approach an observatory, not a one-off audit: every measurement is captured at a baseline date and can be re-measured later, so that recommendations can be proven — before and after. Measured, not theoretical.
The first screen asks for three things:
Why it matters: one submission launches the whole pipeline. No manual sampling, no guesswork — the evidence base is built from the site itself.
You will see five stages run live:
| Stage | What it does |
|---|---|
| Rendering-aware collection | Reads pages the way a real browser does — essential for modern websites built with visual builders (JavaScript-generated content) |
| Structured normalisation | Turns every page into clean, structured data: text, metadata, links, structured data |
| Central index | Builds one searchable index of everything collected — the single source of truth for all five tracks |
| Specialised analysis | Five dedicated engines analyse their track in parallel, each producing sourced findings |
| Report generation | Assembles everything into a structured report with prioritised recommendations |
Why it matters: you see the work happening — counters, progress, statuses. Transparency by construction: the report is produced from what you just watched, not from a template.
The report opens with an executive summary, then five sections:
Recommendations are prioritised P0 / P1 / P2 (impact × effort), so implementation teams know exactly where to start.
Why it matters: every finding cites its source (page URL). You can verify any claim — this is an audit you can trust, not an opinion.
Search is changing: questions are increasingly answered directly by AI engines (ChatGPT, Perplexity, Google AI Overviews). A brand absent from these answers loses visibility even when its classic rankings are strong.
The GEO Observatory measures this presence:
The benchmark is dated and replayable — the same questions can be re-run quarterly to measure progress. This is the only credible way to deliver on GEO: measured, not theoretical.
The report is produced as a draft. It becomes final only after human validation — an expert reviews the findings before anything is exported. No recommendation reaches you as "final" without this review.
Why it matters: AI prepares, humans decide. You get the best of both: the speed of automated analysis and the accountability of expert judgement.
One click exports a self-contained dashboard (a single file): interactive, offline, no server, no licence. You keep it forever and can share it internally — full reversibility.
Is the data public? Yes — the demo processes public web content only. No personal data.
Is the crawl respectful? Yes — it follows the site's rules (robots.txt), with a polite pace and a limited sample. The site remains fully operational.
Is this a commitment? No — this demo is a proof of method on a sample. A full engagement would cover the complete site with monitoring and quarterly re-benchmarks.
Where does the platform run? On European (France-based) infrastructure — data stays in Europe.
What happens after this demo? The video walkthrough (available in the technical proposal) shows the same journey; the platform and method are documented in the Yellow Paper annex.
We built the platform that audits your website — and we measure, not theorise.