ServiceGrowth: SEO & ads
Be readable, and quotable, by AI search
Our AI search service (often called GEO, generative engine optimisation) makes your website easy for assistants such as ChatGPT search and Google's AI Overviews to find, understand and cite: pages that answer in their first sentence, structured data that matches what is on the page, the right crawlers allowed, a plain fact sheet for AI agents and same-day change notifications. It is built on solid SEO, because that is what these systems read first.
What we do for you
- An answer-first opening sentence on your key pages: the question is answered before anything else
- JSON-LD structured data (Organization, Service, FAQ, LocalBusiness where it applies) matching the visible content
- A robots.txt that explicitly allows the crawlers AI search tools use to show sites, with your choice on training crawlers
- An llms.txt fact sheet at the root of your site: who you are, what you offer, where, and the pages that matter
- IndexNow notifications on every change, alongside sitemaps and Search Console
- Facts kept consistent across your site so assistants do not find two versions of your hours, prices or address
Who it is for
- Fits: a business whose customers ask assistants for recommendations, comparisons or how-to answers
- Fits: a site that already has useful pages but buries the answer under long introductions
- Fits: owners who want to decide, knowingly, whether their content may be used to train AI models
- Does not fit: anyone looking for a trick that makes ChatGPT recommend them — no such switch exists, and we will not sell one
What AI search really needs from your site
There is a lot of noise about AI search. Here is what the platforms themselves say. Google writes that there are "no additional requirements to appear in AI Overviews or AI Mode, nor other special optimizations necessary": a page must be indexed and eligible to be shown with a snippet, and the usual SEO best practices remain relevant. Google adds that you do not need new machine-readable files or special schema.org markup for these features.
OpenAI, for its part, says its OAI-SearchBot crawler is "used to surface websites in search results in ChatGPT's search features", and that sites blocking it will not be shown in ChatGPT search answers. So the foundations are simple: be crawlable by the right bots, be indexed, and be the clearest answer. That is why this service sits on top of our SEO service rather than replacing it.
What we put in place, and why
| Element | What it does | Official basis |
|---|---|---|
| Answer-first opening | States the answer in the first sentence, so a reader (human or machine) gets it without scrolling | Our editorial rule, in line with Google's people-first guidance |
| JSON-LD structured data | Describes your business and page in a format machines read; only what is visible on the page | Google recommends JSON-LD and forbids markup about content users cannot see |
| robots.txt rules | Allows search crawlers such as OAI-SearchBot; lets you decide on training crawlers such as GPTBot or Google-Extended | OpenAI crawler documentation; Google common crawlers |
| llms.txt | A short Markdown fact sheet for AI agents at the site root | llmstxt.org — a proposal, not a search-engine requirement |
| IndexNow | Tells participating engines (Bing, Naver, Seznam.cz, Yandex, Yep) that a URL changed; shared between them | indexnow.org |
Search crawlers versus training crawlers: your choice
Two kinds of bots are easy to confuse. Search crawlers fetch pages so they can be shown and cited in answers. Training crawlers collect content that may be used to train future models. OpenAI separates OAI-SearchBot (search) from GPTBot (training). Google separates Googlebot from Google-Extended, a token that controls whether content may be used to train Gemini models; Google states that Google-Extended "does not impact a site's inclusion in Google Search nor is it used as a ranking signal".
By default we allow search crawlers, because blocking them removes you from those answers. For training crawlers we explain the trade-off and apply your decision. We also remind clients that robots.txt is not a privacy tool: Google says it "is not a mechanism for keeping a web page out of Google"; noindex or a password does that.
About llms.txt, honestly
llms.txt is a proposal published on llmstxt.org: a Markdown file, usually at /llms.txt, with an H1 naming the site (the only required section) and a short blockquote summary, followed by links to the pages that matter. It is designed for AI agents that need concise information in one place.
We add one to every site we manage because it is cheap, harmless and keeps your key facts in one checked place. We do not tell you it makes Google rank you: Google says no AI text file is needed for its AI features. It is a convenience for agents, not a ranking lever.
How we do it, step by step
- Inventory. We list the questions customers ask about your business and check how your pages answer them today.
- Facts sheet. We collect your facts once (offer, area, hours, conditions, contact) and you confirm them. Every page, the structured data and llms.txt are built from that single list.
- Rewrite openings. Key pages get a first sentence that answers; the detail follows. The same automatic checks as our SEO work apply: no unsourced figure, no copied text.
- Technical layer. robots.txt rules, JSON-LD, llms.txt and IndexNow are set up and tested.
- Measure. We follow your Search Console and, where visible in analytics, visits referred by AI assistants. Changes are re-measured after 28 days.
Mistakes we avoid
- Blocking every AI bot by reflex, which also removes you from AI search answers
- Adding structured data for reviews, ratings or prices that are not on the page
- Writing pages for machines (lists of keywords, repeated questions) instead of for people
- Promising citations in ChatGPT or AI Overviews: no one controls those answers
Questions we get
Can you guarantee that ChatGPT will recommend my business?
No, and be wary of anyone who does. We make your site crawlable by AI search bots, clear, consistent and well structured; whether an assistant cites you depends on its own systems.
Does Google need an llms.txt file to show me in AI Overviews?
No. Google says no special AI text file or markup is required; a page must be indexed and eligible for a snippet. We add llms.txt as a fact sheet for other AI agents.
Will allowing AI crawlers let them train on my content?
Not necessarily. Search crawlers and training crawlers are separate (for example OAI-SearchBot and GPTBot, or Googlebot and Google-Extended). We allow search crawlers and apply your decision on training ones.
Is AI search a separate job from SEO?
It builds on it. AI features read indexed, well-structured, useful pages, so the same daily SEO work does most of the job; this service adds the AI-specific layer.
How do you measure AI search?
With what is measurable: indexing and performance in Search Console, and visits from AI assistants where analytics records the referrer. We do not invent visibility scores.
Sources
- Google Search Central — AI features and your website (checked 2026-10-06)
- OpenAI — Overview of OpenAI crawlers (checked 2026-10-06)
- Google Search Central — Google common crawlers (Google-Extended) (checked 2026-10-06)
- Google Search Central — Introduction to robots.txt (checked 2026-10-06)
- Google Search Central — Introduction to structured data markup (checked 2026-10-06)
- llmstxt.org — The /llms.txt file (checked 2026-10-06)
- IndexNow — home and FAQ (checked 2026-10-06)