Will this improve my Google rankings?

Not directly, and I won't claim otherwise. Some of the work overlaps with technical SEO — structured data, clear semantics, fixing what's outdated — and that helps search generally. But this service targets what AI assistants say about you, which is a different surface with different mechanics. If rankings are your goal, ongoing SEO is the right service; they're often bought together, and I'll tell you which one your situation actually calls for.

Can you guarantee my company appears in ChatGPT answers?

No, and neither can anyone else. Answers vary between runs, get reassembled over time and depend on phrasing, language and whether the assistant searched. What can be done is making your facts unambiguous and confirmed by independent sources, so that when the assistant does search, the right version is what it finds. That shifts probability, not certainty.

How long before anything changes?

Source corrections take effect after a recrawl — usually weeks. Building visibility where there was none takes months, because it depends on third-party confirmation accumulating. Claims sitting in a model's training data may not change until that model is retrained, which is outside anyone's control. The first three months produce a baseline and little else, and I say so before we start rather than after.

What if the AI says something wrong about us that isn't fixable?

That happens, and identifying it early is part of the audit. If a claim comes from the model's memory rather than a retrieved page, source corrections won't reach it. The workaround is making the correct version dominant in retrievable sources, so search-grounded answers stay right while the model's memory lags. Since a growing share of answers involve search, that covers most of the practical exposure.

Do we need this if our SEO is already good?

Possibly not, and the audit will tell you. Sites with strong technical foundations often already appear in AI answers — in which case the useful question isn't visibility but accuracy: is what's being said correct and current. That's a smaller piece of work, and I'd rather scope it that way than sell you the full engagement.

Can we do this ourselves?

Much of it, yes. The audit method, the prompt set, the source tracing and the monthly measurement are all documented in detail in my articles, and several clients have taken the audit report and implemented it themselves. The parts that are genuinely harder alone are the structured data implementation and knowing which errors are worth pursuing.

AI-Ready SEO & GEO Optimization — Future-Proof Your Website for AI Search

Key Features of AI-Ready SEO and GEO

Audit against a fixed prompt set

Ten to twenty-five prompts run across ChatGPT, Perplexity and Google's AI answers, frozen so results stay comparable month to month.

Source tracing for every wrong claim

I trace every error to the page or dataset it came from, and separate what's fixable from what isn't.

Structured data and entity signals

Organisation and service markup that makes you a distinct entity to a machine, not a member of a category.

Correction of retrievable sources

Your own pages fixed and dated, third-party listings corrected where possible, and corrected URLs submitted for recrawl.

Measurement you can check yourself

A monthly log of presence, errors and citations — not a dashboard you have to trust.

AI-Ready SEO and GEO: two different problems

People increasingly ask an assistant instead of searching. The answer they get names two or three companies, and the buyer usually checks one of them. That creates two separate problems, and most work sold under this heading conflates them.

The first is visibility. Whether you are named at all when someone describes their situation to ChatGPT, Perplexity or Google's AI answers. This depends on how unambiguously your site states what you do, for whom and where — and on whether independent sources confirm it.

The second is accuracy. What gets said about you when you are named. Outdated prices, services you dropped, a capability credited to a competitor, or a caution advising the buyer to verify your credentials. This is a different problem with a different fix, and it costs more than absence — a wrong answer travels further than no answer.

I work on both, and I keep them separate, because the diagnosis differs. If you appear in two answers out of twenty, that's visibility. If you appear in fifteen and half of them describe you incorrectly, that's accuracy, and adding content won't touch it.

What you see	What it means	What I do
You're never named in relevant answers	Visibility: sources don't state clearly enough what you do	Rewrite the facts as checkable statements, add entity markup, build third-party confirmation
You're named but described incorrectly	Accuracy: a retrievable source carries the wrong claim	Trace the source, correct or outweigh it, prompt a recrawl
Your capability is credited to a competitor	Entity confusion: the two of you aren't separated	Entity markup, consistency across profiles, factual comparison content
The description is right but two years old	Stale sources outweigh your current site	Find and retire the old sources, date the current ones
The assistant advises caution about you	Either accurate feedback or thin data	Establish which — the two have nothing in common but the symptom

What the work actually consists of

A baseline audit against a fixed prompt set. Ten to twenty-five prompts written the way your buyers actually phrase things — branded, category, comparison and problem-based — run in a clean session across ChatGPT, Perplexity and Google's AI answers, in every language your market uses. For each one I record whether you were named, what was claimed, whether it's accurate, which sources were cited, and the date. The set is then frozen, because changing it between runs destroys comparability.

Source tracing for every error found. Errors come from somewhere. Usually an outdated directory listing, an old press item, a review profile written at registration, or a competitor's comparison page. I find which one, and separate what can be fixed from what can't — because a claim living in a model's training data doesn't respond to source corrections until the model is retrained, and effort spent there is wasted.

Structured data and entity signals. Organisation and service markup, links to the profiles representing you elsewhere, consistency of name, category and location across everything. This is what makes you a distinct entity to a machine rather than a member of a category — and it's what stops your capabilities being attributed to a similarly named competitor.

Correction of retrievable sources. Your own pages fixed and dated. Third-party listings corrected where they accept requests. Sources that can't be changed outweighed rather than argued with. Corrected URLs submitted for recrawl, including to Bing — which matters more than its traffic share suggests, because several assistants ground their search in it.

Content written to survive summarisation. Assistants don't quote, they compress. What survives is anything stated as a checkable fact: who you serve, where, what you technically do, what you don't. Aspirational language disappears and gets replaced by the generic description of your category. Part of this work is rewriting the distinguishing claims into a form that can be repeated.

Measurement you can verify yourself. The prompt set re-run monthly, the citation report in Bing Webmaster Tools, and a referral segment for traffic arriving from assistant domains. You get the log, not a dashboard — so you can check the numbers rather than trust them.

What I don't promise

Nobody can guarantee what an assistant will say, and anyone who does is selling you an outcome they don't control. Specifically:

No guaranteed placement in answers. Answers vary between runs, are reassembled over time, and depend on phrasing and language.
No fixed timeline for corrections. Source changes take effect after a recrawl — weeks, typically. Claims sitting in a model's training data may not change until a retrain, on the platform's schedule.
No claim that this replaces search work. AI visibility rests on the same technical foundation as search. If your pages aren't indexed or load badly, this work has nothing to stand on.

What I can do is make your facts unambiguous, correct what's correctable, and measure the result honestly enough that you can tell whether it moved.

Who this is for

Businesses where buyers research before contacting: professional services, B2B, considered purchases. Companies already selling into more than one language market, where the answer differs by language. And anyone who has found an assistant describing them incorrectly and wants to know whether it can be fixed.

Who this isn't for

Businesses whose customers don't research online at all. Companies with no website worth optimising — this rests on technical foundations, and it isn't the first thing to spend on. And anyone looking for a guarantee: the honest version of this service doesn't include one.

Prices

Baseline audit — from €250. The prompt set, the run across platforms and languages, the source tracing, and a written report separating what's fixable from what isn't. It's a complete deliverable — several clients have taken the report and implemented it themselves.

Ongoing work — from €800 per month, with a three-month minimum, because nothing measurable happens faster than that. Includes implementation, source correction, monthly re-runs and the measurement log.

Final figures depend on the number of languages, the size of the site, and how much source correction sits outside your own domain.

AI-Ready SEO & GEO Optimization