Answer engines

AI SEO Services

Being the source an AI answer quotes, rather than the tenth blue link underneath an answer that already satisfied the question.

Book a Reading See what you get

[INSERT TRUST BAR]Client logos, years in operation, or a verified results figure.

The click you used to get is now an answer somebody else got credited for

A growing share of searches now end without a click. The person asks, an AI Overview or a chat assistant answers, and they act on that answer. Your ranking did not change. Your traffic did. The two have quietly stopped being the same measurement, and most reporting has not caught up.

What decides whether you appear in that answer is not what ranks you tenth. Answer engines pull from sources they can parse cleanly, attribute confidently and corroborate elsewhere. A page can rank well and never be cited, and a page can be cited constantly from position eight. That gap is the work.

What AI SEO actually consists of

Concrete work. There is no setting in ChatGPT that we have access to and you do not.

01 Baseline visibility measurement We ask the assistants the questions your buyers ask, at volume and on a schedule, and record who gets cited. Without this you are optimising against an outcome nobody is watching, and every later claim of improvement is unfalsifiable.

02 Answer-shaped content structuring A direct answer in the first paragraph, then the evidence. Models extract claims, not narrative build-up, so a page that withholds its conclusion for four hundred words is a page that does not get quoted.

03 Entity and attribution work Making it unambiguous who you are, what you do, and what you are an authority on, across schema, your own site and the third-party sources models corroborate against. Assistants cite entities they can resolve confidently.

04 Corroboration across sources A claim that appears only on your own site is a claim a model will hedge on. The same fact stated on sources it already trusts is one it will repeat. This overlaps with digital PR and it is the slowest part.

05 Crawlability for AI agents Whether the assistant crawlers can reach your content at all, what your robots directives actually permit, and whether the substance survives JavaScript rendering. We find blocked crawlers more often than we expect to.

06 Conventional SEO underneath it AI answers are drawn largely from pages that already rank. Nobody has found a shortcut around that, and an agency selling AI SEO detached from search fundamentals is selling you a layer with nothing underneath it.

How an AI SEO engagement runs

01 Prompt set and baseline We build the set of questions your buyers actually ask an assistant, run them across ChatGPT, Google AI Overviews, Perplexity and Gemini, and record who gets cited today. You get the baseline before anything is agreed. Week 1-2
02 Gap analysis Where a competitor is cited and you are not, we read why: the source they pulled, its structure, and what it stated that you do not. Week 2-3
03 Technical and entity fixes Crawler access, schema, disambiguation and rendering. These are quick and they are prerequisites, so they go first. Week 3-5
04 Content restructuring and creation Existing pages rewritten to answer directly; new pages for the questions no page of yours currently addresses. Month 2+
05 Re-measure against the baseline The same prompt set, monthly, against the same starting numbers. Movement is reported as citation share, not as a screenshot of one lucky answer. Monthly

Results, with the working shown

Every number here comes with the client name, the timeframe and the method. If we cannot show you how it was measured, we do not publish it.

[INSERT CASE STUDY 1]Client name, starting position, result, timeframe, and how it was measured.

[INSERT CASE STUDY 2]Client name, starting position, result, timeframe, and how it was measured.

See documented case studies

How we price AI SEO work

No published rates. This is a young discipline with a lot of confident pricing attached to it, so here is what genuinely moves the number.

01 Size of the prompt set Tracking twenty buyer questions across four assistants is a different measurement job from tracking three hundred. This is mostly a tooling and analysis cost and it is the one that scales.

02 Content volume How many existing pages need restructuring, and how many questions have no page at all. Usually the largest single line.

03 How resolvable your entity already is An established brand with consistent third-party coverage starts a long way ahead of a company that models cannot confidently distinguish from a similarly named one.

04 Whether corroboration is in scope Earning mentions on sources models already trust is the slowest and most expensive component. Some engagements need it. Many get a long way without it.

05 The state of conventional SEO If the site does not rank, the honest sequence is to fix that first. We will say so, and it changes what the first three months are spent on.

What you are committing to

Baseline before you commit Ninety days to start Then month to month The prompt set is yours

One thing we will not do is price this as a premium add-on to work we are already doing. Where it overlaps with your SEO retainer, it is the same work and it is billed once.

Questions buyers actually ask

Work that makes your content the source an AI assistant cites when it answers a question in your category. In practice it is measurement of where you are cited today, structuring content so claims can be extracted and attributed, making your entity unambiguous to a model, and making sure AI crawlers can reach the content at all. It sits on top of conventional SEO rather than replacing it.

The foundations are shared: crawlable, credible, well-structured pages help in both. The difference is the target. Traditional SEO optimises for a ranked position that earns a click. AI SEO optimises for being quoted inside an answer that may produce no click at all, which changes how a page should be written and how success has to be measured.

No. Nobody can, and an agency that offers it either does not understand the systems or is counting on you not checking. Model outputs vary between users, sessions and model versions. What we commit to is a measured baseline, a fixed prompt set, and monthly reporting on citation share against it, including the months it does not move.

By sampling rather than by anecdote. The same prompt set is run repeatedly across assistants on a schedule, and we report the share of responses citing you. A single answer is noise. A shift in citation share across dozens of runs is a signal, and it is the only honest way to report this.

Yes, and mostly it is the same investment. AI answers draw heavily on pages that already rank, so a site with no search visibility has very little for an assistant to cite. If your fundamentals are weak, we will tell you the first three months are SEO, not AI SEO.

They are overlapping names for closely related work: answer engine optimisation, generative engine optimisation, AI SEO. We treat this page as the commercial entry point and keep the detail on separate pages, linked below, rather than writing the same thing three times under three acronyms.

Technical and structural fixes can change citation behaviour within weeks, because models re-crawl. Building the corroboration that makes a model confident about you takes months, on the same timeline as earning authority anywhere else. Anyone quoting thirty days is describing the easy half.

Measured, or asserted

The failure mode in this category is not fraud. It is that nobody is measuring, so nobody can be wrong.

What you are comparing A typical AI SEO offer MyInscribe

Baseline Rarely taken Recorded before work starts, and yours

Evidence of movement A screenshot of one good answer Citation share across a fixed prompt set

Relationship to SEO Sold as a separate product Same foundations, billed once

What is promised Guaranteed AI visibility Nobody controls model output, and we say so

Find out whether AI search knows you exist

A free baseline: we run the questions your buyers ask across ChatGPT, AI Overviews and Perplexity, and send you who gets cited today. You keep the prompt set and the findings whether or not you work with us.

Get the AI visibility baseline Send your questions instead

hello@myinscribe.com · replies within a working day