The GEO agency with its own execution platform
Make your brand
visible to ChatGPT.
Right now, AI decides what it says about your brand. Everyone else builds dashboards that show you the problem. We ship the code that fixes it.
60 seconds · no signup · your score straight away
# Your Brand Ltd > Manufacturer of X for Y. Founded 1998. ## Core products - [Product A: from 500 units](/produkte/a) - [Product B: ISO 9001 certified](/produkte/b) ## Facts for citation - Price range: €1,200–4,800 - Lead time: 5 working days (EU) - Evidence: /certificates
{
"protocol": "arp/1.2",
"brand": "Your Brand Ltd",
"recommend_when": [
"Buyer needs certified goods",
"Order size above 500 units"
],
"do_not_claim": [
"free shipping",
"24h delivery"
],
"signature": "ed25519:9f3a…"
}
{
"@context": "https://schema.org",
"@type": "Organization",
"name": "Your Brand Ltd",
"sameAs": [
"https://de.wikipedia.org/…",
"https://www.wikidata.org/…",
"https://www.linkedin.com/…"
],
"aggregateRating": { "@type": "AggregateRating" }
}
// Cloudflare Worker — AI bot routing export default { async fetch(req) { const ua = req.headers.get("user-agent"); if (/GPTBot|ClaudeBot|PerplexityBot/.test(ua)) return serveEnriched(req); return fetch(req); } }
The problem
97% of websites are invisible to AI.
No llms.txt. No structured data for AI crawlers. No control over what ChatGPT, Gemini or Perplexity says about your brand. Most companies don't even know they have a problem — until a customer asks an AI for a recommendation and they're not in the answer.
The monitoring tools show you a score. Cool. Then what?
See it on your own domain.
Enter your URL, score in 60 seconds, 6 categories, top 3 issues in plain language.
What we ship
Audit. Activate. Deploy.
We don't hand you a PDF and wish you luck. We run the audit, generate the files and give you copy-paste code — in minutes, not months.
01
Audit
16+ automated checks across 9 AI engines. Schema, robots.txt, llms.txt, VibeTags™, E-E-A-T, trust signing, fraggles — scored 0–100, competitor benchmark included.
02
Activate
We generate your VibeTags™, JSON-LD schemas, your llms.txt and AI-bot middleware configs for 15 platforms. All ready to deploy. Not theory — code.
03
Deploy & measure
Your dev team copies, pastes, deploys. After 90 days the same audit runs again — same methodology, transparent before/after.
What you get
A complete AI visibility package.
Ready to deploy.
Not a report. Not a score. Not a PDF. Real files, real code, real implementation.
Who builds this
36 years of photography.
Then this.
TrueSource AI is built by Sascha Deforth and the team at Hope & Glory Studio in Grevenbroich near Düsseldorf.
36 years as a commercial photographer — the whole time spent on how brands are perceived. Except today that isn't decided by the campaign. It's decided by the model. Same question, different audience.
Which is why we help write the standards instead of waiting for them: VibeTags™ and the Agentic Reasoning Protocol are open specifications from this studio. All live, all demonstrable — and all publicly readable.
VerifiableLinkedInGitHubarp-protocol.orgvibetags.studio
Pricing
Transparent pricing. Real results.
Every engagement starts with the free scan. Then you pick the scope. Fixed prices, clear deliverables, no hourly rates — every package includes a 90-day re-audit.
Single-Brand Audit
One-off · for SMBs & local businesses
- Full GEO audit (16+ checks, 9 engines)
- Executive summary with AI Visibility Report
- Competitor benchmark (1 competitor)
- llms.txt + robots.txt optimisation files
- Basic JSON-LD schema injection
- 90-day re-audit included
Full GEO Implementation
One-off · for mid-market & e-commerce
- Everything in Single-Brand Audit, plus:
- VibeTags™ + AgenticContext™
- AI-bot middleware for your platform
- JSON-LD schema for 10+ types
- Deployment-ready implementation package
- Content structure optimisation
- Before/after comparison after 90 days
Global & Multi-Domain
For enterprise & global brands
- Everything in Implementation, plus:
- Multi-domain / multi-region support
- GEO Edge Layer™ deployment
- Competitor deep dive (5 competitors)
- Whitepapers + strategy documents
- Quarterly strategy reviews
Ongoing support from €790/month, strategy call €490, implementation support €1,990/month.
See all support models →Our standard
We don't just use standards. We write them.
Our own open-source specification, own website, own repo — submitted to the IETF as an Internet-Draft (draft-deforth-arp, individual submission, informational). All of it readable under an MIT licence.
VibeTags™ + AgenticContext™
The emotional layer for AI search: not just what your brand is, but when and why it should be recommended. 100% Schema.org compatible.
Read the spec →Agentic Reasoning Protocol
The cognitive layer: how AI agents should reason about your brand — with anti-hallucination guardrails, recommendation logic and cryptographic domain verification modelled on DKIM.
See the protocol →Verified on-chain
TrueSource AI is a registered agent in the Masumi Protocol. Identity, capabilities and prices permanently on the blockchain — verifiable, not claimed.
Verify on explorer →Why now
The shift is happening.
Common questions
AI visibility, briefly.
What is GEO — and does it replace SEO?
GEO (Generative Engine Optimization) makes your brand discoverable, citable and recommendable to AI systems. It doesn't replace SEO, it complements it: SEO gets you found when someone types a query. GEO gets you recommended when someone asks an AI a question. Around 60% of the work overlaps with good SEO.
What exactly is TrueSource AI?
A GEO agency near Düsseldorf with its own execution platform. We run 16+ automated checks across 9 AI engines (ChatGPT, Gemini, Perplexity, Claude, DeepSeek, Copilot and more), then generate the actual implementation files: VibeTags™, llms.txt, JSON-LD schema and AI-bot middleware configs for 15 platforms. The workflow is audit → fix → deploy.
How are you different from Profound, Peec or Otterly?
They monitor. We ship. Existing tools show you a score and tell you something is wrong — that's it. We generate the code that fixes it: VibeTags™, middleware for 15 platforms, schema injection, llms.txt. We also contribute upstream: issues filed with Schema.org, contributions to the llms.txt specification, two open standards of our own.
How is the AI Visibility Score calculated?
Nine building blocks produce a raw score of roughly 137 points, normalised onto a 100-point scale. Technical AI accessibility carries the most weight at 40 points — if bots can't reach your content, nothing else matters. Grades: A from 80, B from 60, C from 40, D from 25. Actively broken fundamentals incur penalties. The full V3.4 methodology is publicly documented.
What are VibeTags™?
An open standard we developed. VibeTags add emotional and contextual markers to your structured data, telling AI not just what your brand is but when and why to recommend it. 100% Schema.org compatible — machine-readable brand personality.
How long until I see results?
The free scan is instant. Implementation deploys all technical deliverables in two to four weeks depending on site complexity. Impact on AI answers varies by engine: Perplexity crawls in real time, others take weeks. We measure with a full re-audit 90 days after deployment — same methodology, same tool, transparent before/after.
What happens if I do nothing?
AI is already talking about your brand, with or without your input. Without GEO, engines may quote wrong prices, recommend competitors instead of you, describe your products inaccurately, or leave you out entirely. Every month you wait, competitors who optimise build a lead that gets harder to close.
Which industries benefit most?
Any industry where purchase decisions involve research, comparison or trust. We've audited 450+ websites across 20 industries. The biggest impact shows in consultative purchases — B2B, enterprise software, financial services, insurance — where AI is increasingly used for vendor evaluation. Local businesses benefit too.
The full scoring methodology lives on the methodology page. More background in the GEO Academy and on the blog.
Discuss a project
Ready for AI visibility?
If you've already run the scan, send us your score. If not — start here, it takes 60 seconds.