The 403 Nobody
Could See.
This service exists because of an audit we ran for a client who asked a simple question: “Why doesn’t ChatGPT know about my business?”
Their site was good. Their robots.txt politely welcomed every AI crawler. And when we tested what the server actually did, every single AI system — OpenAI, Anthropic, Perplexity — was being served 403 Forbidden by a firewall setting the owner didn’t know existed. The assistants couldn’t read a word of the site. Not because the business wasn’t good enough to recommend — because the door was locked and nobody could see the lock.
So we swept every site under our own management — more than a hundred live sites — and found over half of them silently blocking AI crawlers, most through defaults nobody ever chose. We fixed them the same day, verified every one, and wrote the whole method up in a step-by-step guide you can run yourself.
That’s the pattern with AI search right now: the problems are real, they’re invisible from the dashboard, and they’re fixable — if someone actually tests instead of assumes. That’s the job.
A New Front Door Opened.
Most Businesses Haven’t Noticed.
For twenty-five years, being found meant ranking in a list of blue links. That game still matters — we’ve played it at the highest level since 1996 — but a second game started, and it has different rules. When someone asks ChatGPT or Perplexity for “the best [what you do] near me,” there is no page two. There are two or three names, delivered with confidence, and everyone else might as well not exist.
Those names aren’t chosen by magic. Assistants assemble answers from what they can read (your site, if their crawlers can reach it), what they can verify (structured facts, consistent details across the web), and what they can corroborate (reviews, mentions, citations). Every one of those is a lever you can actually pull — and almost nobody is pulling them yet, because the tooling is young and the industry around it is mostly noise.
The noise is the other half of the problem. “AI SEO” is already filling up with agencies selling guaranteed ChatGPT rankings — a thing that does not exist and cannot be bought. What exists is unglamorous, verifiable work: access, structure, content, reputation, measurement. We’d rather tell you that plainly than sell you a dashboard.
Tested, Not
Assumed.
Everything we sell on this page runs on our own site first. Our blog’s GEO playbook is the same one we execute for clients, our structured data ships schema-first on every page, and every AI crawler that matters gets a 200 from us — you can test that yourself with five terminal commands. The approach has three pillars:
Access. If AI systems can’t read you, nothing else matters. We test every major crawler against your live site — homepage and deep pages, apex and www — and fix what’s blocking them at the firewall, not just in robots.txt. Tested from outside, verified with real HTTP responses.
Understanding. Machines shouldn’t have to guess what you are. We build a connected entity graph — your business, your people, your services, your location — in structured data on every page, with every fact matching what the page visibly says. It’s the difference between an assistant parsing prose about you and an assistant knowing facts about you.
Credibility. Assistants recommend what they can corroborate. Content that answers real questions in plain language, consistent details across every directory and profile, reviews where they count, and the emerging plumbing — llms.txt included — done honestly, with no pretending it’s magic.
"When it comes to digital strategy, he is truly one of the best in the world at what he does. I don't say that lightly."
The Work,
Named Plainly.
No mystery deliverables. Every item below is something we execute, verify, and show you the evidence for.
AI Visibility Audit
Live testing of every major AI crawler against your actual site — GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, PerplexityBot and Google’s systems — plus a review of what they find when they get in. You get the real HTTP responses, not a screenshot of a settings page.
Entity Schema Build
A connected structured-data graph across your whole site: business, people, services, hours, location, profiles. Every fact sourced from your visible pages — schema that lies to Google is worse than no schema at all.
GEO Content Strategy
Pages and articles engineered for how generative engines actually quote: direct answers, real questions, visible FAQs with matching markup, and topical depth that makes you the safe recommendation.
Technical Plumbing
llms.txt written and published, sitemaps and canonicals clean, crawlable architecture, fast pages. The unglamorous layer that decides whether retrieval systems bother with you at all.
Citation & Consistency Sweep
Your name, address, hours and story reconciled across every directory, profile and listing that AI systems cross-reference. Conflicting facts make a cautious machine skip you.
Honest Measurement
Monthly checks of what the assistants actually say when asked your customers’ questions, AI-referral traffic, and true search positions from real SERP checks — never vanity metrics dressed up as progress.
Two Games,
One Foundation.
AI assistants retrieve from the same indexes classic search built. GEO without real SEO is a megaphone with no signal — and both run best on a site built right from the first line of code. We do all three, as one system.
SEO & Performance
Nearly three decades of search. The rankings AI systems retrieve from start here — GEO stands on top of it, not instead of it.
Explore → THE VEHICLEBespoke Web Design
Clean, fast, semantic sites are what machines read best and humans convert on. When we build the site, the AI-readability is engineered in, not bolted on.
Explore →Ask an AI About Your Business.
Then Ask Us.
The first step is an AI visibility audit: can the assistants read you, what do they know, and what do they say. Real tests, real evidence, and a straight answer about what’s worth fixing.