Service · AI search
AI Search Optimisation (GEO): your business, cited in AI answers
GEO (Generative Engine Optimization), also called AI search optimisation, makes your business visible and cited in the answers of ChatGPT, Claude, Perplexity and Google AI Mode. We audit, then implement directly in your code: AI crawler access, structured data, entity identity, llms.txt.
Last updated:
Do these situations sound familiar?
A growing share of your buyers asks an AI assistant instead of scanning ten Google results. If the assistant does not cite you, you do not exist in that conversation.
-
ChatGPT never mentions you
Asked about your market, assistants cite your competitors or directories - but not you.
-
AIs confuse you with a namesake
Your brand shares its name with other products, and the models merge the entities.
-
Your CDN blocks AI bots
Cloudflare blocks AI crawlers by default: your site may be invisible to them without you knowing.
-
Your content only exists after JavaScript
AI fetchers do not always execute JS: they see a page without content.
-
No structured identity data
Without a complete Organization schema, an AI knows neither who you are, nor where, nor what you sell.
-
Your offers are unreadable by machines
Your prices exist as pixels, not data: an answer engine cannot repeat them.
What you get
-
01
A site accessible to AI crawlers
Explicit bot-by-bot robots.txt, verified CDN configuration, content present in the initial HTML.
-
02
An unambiguous entity identity
Complete Organization schema - legal identifiers, address, official profiles - so the models know exactly who you are.
-
03
An llms.txt that sums up your offer
Services, public prices and key pages in the format answer engines read first.
-
04
Content structured to be cited
Direct answers at the top of each page, complete FAQs in structured data, honest last-updated dates.
What we check
Everything that determines whether an answer engine can read you, understand you and cite you.
-
Crawler access
robots.txt, firewall and CDN (Cloudflare blocks AI bots by default), every user agent tested: GPTBot, ClaudeBot, PerplexityBot, Google-Extended…
-
Content rendering
What the initial HTML response contains without executing JavaScript: copy, prices, FAQs, offers.
-
Entity identity & namesakes
Organization schema, legal identifiers, sameAs to your official profiles, confusion risks with namesakes.
-
llms.txt & sitemap
Presence, accuracy and freshness of the files AIs use to map your site.
-
Content structure
Direct answer at the top of each page, heading hierarchy, marked-up FAQs and comparisons - the formats AIs quote.
-
Freshness signals
Real publication and modification dates, visible and in structured data - never fabricated.
Our method
The same cycle as technical SEO: audit, implementation in the code, follow-up.
-
1
Accessibility diagnosis
Can AI crawlers reach your site? robots.txt, CDN and rendering tested user agent by user agent.
-
2
Entity & structure audit
Structured data, namesakes, llms.txt, content structure: every gap documented.
-
3
Prioritised action plan
Every point ranked by impact and effort, with the exact fix to apply.
-
4
Implementation
Optionally, we apply the fixes directly in your code - as we did on our own.
-
5
Follow-up
AI bot crawl logs, periodic re-querying of the assistants, off-site recommendations (profiles, Wikidata).
The deliverable
A factual report: what blocks, what is missing, in which order to fix.
- AI crawler accessibility report: robots.txt, CDN, content rendering
- Schema and entity identity audit, namesakes included
- llms.txt, sitemap and freshness-signal check
- Content structure analysis of the key pages
- Prioritised action plan, ready to implement
- Video-call walkthrough, optional implementation by us
Every point in the report is verifiable: we show the request, the server response and the fix.
Where to start?
Free assessment
20 minutes on a video call: your market, what AI assistants answer about your sector today, and whether an audit is worth it.
Book a 20-min callGEO Audit
Complete audit and prioritised action plan. GEO is also included in Monthly management (€1,500 excl. VAT / month).
- AI crawler accessibility report
- Schema and entity identity audit
- Prioritised action plan
- Optional implementation in your code
GEO, AEO, LLMO, AI SEO: one practice, several names
The vocabulary has not settled yet; these terms all describe optimising a site for AI-based answer engines.
GEO - Generative Engine Optimization
The most widespread term: optimising so that generative engines (ChatGPT, Perplexity, Google AI Mode) find, understand and cite your site.
AEO - Answer Engine Optimization
The same goal, framed from the "answer engine" side: being the source the assistant quotes when it answers a question directly.
LLMO - Large Language Model Optimization
The variant centred on the language models themselves: what LLMs know about your brand, in their training data and at answer time.
AI SEO / ranking in generative AIs
The everyday phrasings of the same need: being visible in generative AI answers. That is exactly this service.
Our own site is our demonstration
We implemented it on seoforge.fr first. Every element of the method is live and inspectable - no promises, public code.
- llms.txt
seoforge.fr/llms.txt
Our summary for AIs: services, public prices, key pages - generated from the site's routes, so it cannot rot.
- robots.txt
seoforge.fr/robots.txt
Every AI crawler allowed by name - GPTBot, ClaudeBot, PerplexityBot… - with the same rules for all.
- Structured data
Complete entity identity
Organization schema with the company's registry identifiers, founder and official profiles, visible in every page's source - our answer to namesakes.
Frequently asked questions
GEO and SEO: do I have to choose?
No, they are complementary and SEO remains the foundation. A technically sound site - crawlable, fast, well structured - serves both. GEO adds what answer engines additionally require: explicit AI bot access, an unambiguous entity identity, an llms.txt and content formatted to be cited. Good GEO on a poorly ranked site compensates for nothing.
How long before it has an effect?
Technical accessibility is fixed within weeks: as soon as AI crawlers can reach the site, they read it. Citations in assistant answers take longer - several months - because they also depend on model update cycles and your off-site presence. Nobody can guarantee a citation, and we do not promise one.
Are the results measurable?
Partially, and we prefer to say so. What measures well: AI crawler access in the server logs, referral traffic from assistants, and what ChatGPT or Perplexity answer when asked about your market. There is however no mature equivalent of Google rank tracking yet. We track what is measurable and document the rest.
Is my site blocked for AIs?
Possibly without you knowing: Cloudflare, used by a large share of the web, blocks AI crawlers by default, and some firewalls do the same. A site that ranks perfectly on Google can be completely invisible to ChatGPT. It is the first thing the audit checks, user agent by user agent.
What is the llms.txt file?
A text file at the root of the site (your-site.com/llms.txt) that sums up your business, services, public prices and key pages in a format designed for language models. It is an emerging convention and inexpensive to set up. Ours is public: seoforge.fr/llms.txt.
Why trust you on GEO?
Because we implemented it on our own site first, and everything is publicly verifiable: our llms.txt, our robots.txt that allows every AI crawler by name, our entity schema with the company's SIRET, and our honest-dates policy. We sell the method we apply to ourselves - not client results that nobody can seriously prove yet.
Be present where your customers ask their questions
Describe your site and your market: an expert replies within one business day with a first AI-accessibility diagnosis.