AI search is no longer experimental on UK results pages. Google AI Overviews, AI Mode, ChatGPT search, Perplexity and Copilot all decide every day whether your business gets cited or quietly skipped, and the rules for getting picked up are not what most agencies are selling. This is the practical playbook for UK SMBs who want citations without buying into the GEO acronym circus.
What AI search optimisation actually means in 2026
AI search optimisation is the work of getting your pages cited, quoted and recommended inside AI-generated answers across Google AI Overviews, AI Mode, ChatGPT, Perplexity, Microsoft Copilot and the smaller LLM-powered search tools. It overlaps heavily with classic SEO, since AI engines pull most of their citations from pages that already rank in the top 10 organic results. The difference is in the unit of success. Classic SEO chases the click. AI search optimisation chases the citation, the brand mention, and the answer slot, knowing that the click may never come.
You will see this work labelled under several names: AIO (AI optimisation), AEO (answer engine optimisation), GEO (generative engine optimisation), LLMO (large language model optimisation), and increasingly just "AI search engine optimisation". Google itself has been clear that most of these acronyms are repackaging. There is no separate algorithm to "rank" inside AI Overviews, and there is no secret schema or hidden prompt that unlocks AI visibility. What works is unique content, sound technical hygiene, and a brand presence that AI models trust enough to cite. The rest is marketing for marketing tools.
For a UK SMB, the practical scope is narrower than the noise suggests. You are not trying to dominate every AI surface. You are trying to win citations on the queries your buyers type, in the AI features that show up most often for those queries, on the platforms your audience actually uses. That is a manageable list. If you would rather have that list built and worked for you than run it in-house, that is what our answer engine optimisation service does, and the generative engine optimisation side covers the ChatGPT and Perplexity surfaces specifically. If you are still deciding whether any of this is distinct from search work you already pay for, our comparison of AEO and SEO answers that first.
Six moves UK SMBs should make this quarter
1. Rank in the top 10 organically first
AI Overviews pull the bulk of their citations from pages that already sit on page one for the underlying query. If you are not in the top 10, the AI search work is almost moot. Fix the classic SEO basics before chasing AI features.
2. Lead every section with the answer
AI engines lift self-contained sentences. Open each H2 and H3 with a one-sentence direct answer to the heading, then expand underneath. This is how featured snippets won the last decade. The same pattern wins AI citations now.
3. Add schema that AI engines actually parse
Article, Author, Organization, HowTo, Product, BreadcrumbList. These are the structured data types Google explicitly uses to disambiguate your pages. Skip the exotic stuff. Five markup types installed correctly beats twenty rendered with errors.
4. Build off-site brand mentions
Around three-quarters of branded-query AI citations come from external sources: directories, reviews, forums, editorial coverage. Get your business on Crunchbase, Trustpilot, the relevant trade press, and recognised UK industry listings. AI models cite where the web already trusts you.
5. Show first-party expertise (real E-E-A-T)
Named authors with public credentials, original data, screenshots from your own dashboards, case study numbers, photographs from your site. Anonymous content generated by another LLM is the lowest signal you can send to a citing LLM.
6. Track citations, not just keyword rankings
AI answers are non-deterministic: the same query can return five different citation sets in a week. Measure share of voice (how often your brand is mentioned across repeat prompts) and citation rate, not just position. GSC, ChatGPT search referrer and Perplexity referrer give you a free starting view.
There is a seventh move, and it is a "do less" instruction: stop spending on services that promise AI ranking guarantees, GEO-only audits or LLMO playbooks. There is no recognised certification for any of these and no published study has tied them to lift. The agencies offering them are repackaging the same SEO work at a premium. Audit the deliverable, not the acronym.
Let the AI crawlers in before you optimise anything
The most common reason a UK SMB is invisible in ChatGPT and Perplexity has nothing to do with content quality: the site is blocking the crawlers. This is the first thing to check and the cheapest thing to fix.
AI systems use separate user agents from Googlebot, and a robots.txt written three years ago will not know about them. The ones that matter for a UK business are GPTBot and OAI-SearchBot (OpenAI), ClaudeBot (Anthropic), PerplexityBot, Google-Extended (Google's AI training and grounding control), and Bingbot, which still feeds Copilot. OpenAI documents its own crawler user agents and IP ranges in its official bots documentation, and the equivalents are published by each provider.
Two traps catch people out:
- Your CDN may be blocking them without your knowledge. Cloudflare now ships controls that block AI crawlers by default on many plans, which means a site owner who never edited robots.txt can still be invisible to AI engines. Cloudflare documents the behaviour and the toggles in its AI Crawl Control docs. Check the dashboard, not just the file.
- Blocking training is not the same as blocking citation.
Google-Extendedgoverns whether your content feeds Gemini and grounding responses. Disallowing it is a legitimate commercial choice, but make it deliberately, because it also removes you from the surfaces you are trying to win.
The second access problem is rendering. AI crawlers are far less patient with JavaScript than Googlebot has become. If your key content only appears after a click (accordion panels, tabbed specification sections, "read more" toggles that load on interaction) or is injected client-side, assume it will not be extracted. Serve the substance in the initial HTML. Fetch your own page with curl and read what comes back; if the answer is not in there, no amount of schema will rescue it.
What Google's official guidance actually says
Google has published two documents that should anchor every UK SMB's AI search strategy. The first is the May 2025 Search Central post "Top ways to ensure your content performs well in Google's AI experiences", and the second is the 2026 guide "Optimizing your website for generative AI features on Google Search", now the canonical reference for AI search work and currently the top organic result in the UK for this topic. You can read the May 2025 source on the Google for Developers Search Central blog. Both documents say the same five things in slightly different words.
One: focus on unique, valuable content for people. Two: provide a great page experience (Core Web Vitals, mobile usability, HTTPS). Three: ensure Googlebot and Google-Extended can access your content. Four: manage visibility with preview controls if you want to opt out of specific surfaces. Five: there is no special markup, content type or formatting trick that unlocks AI Overviews. Pages are eligible because they meet the same quality bar as the rest of search.
The 2026 guide adds one practical clarification. The AI features blend signals from the underlying ranking system, freshness, structured data, and the broader corpus the model has been trained on. There is no extra "AI rank" to climb. That is why most "GEO" advice boils down to good classic SEO, with a small adjustment for extraction-friendly writing.
We covered the broader announcement when it dropped, with a summary of Google's framing and what it means for UK marketers: see our piece on Google's New Guide to Optimising for AI Search. This playbook is the operational follow-up to that news.
The on-page format AI engines actually extract
If you read enough citations across Google AI Overviews, ChatGPT search and Perplexity, a clear pattern emerges. The lifted text is almost always a self-contained sentence or a small list. The surrounding paragraph rarely makes the citation. Writing for AI extraction is therefore mostly a matter of front-loading every section with the answer.
A few formatting habits that consistently show up in cited content:
- Definition sentences: "X is Y." Open any section that defines a concept with this structure. AI engines lift definitions almost verbatim.
- Numbered lists for processes: step one, step two, step three. AI summaries love sequence.
- Bulleted lists for comparisons and specifications: the AI strips the formatting but keeps the parallel structure.
- Short paragraphs: two to three sentences. Long blocks get summarised away.
- Direct question headings: use the question form for H3 inside FAQ sections, and answer in the first sentence underneath. This remains the highest-citation format in Google AI Overviews, and it still works even though the FAQ rich result itself has been retired.
- Named entities over pronouns: spell out the subject (your brand, the product, the place) rather than leaning on "it" or "they". A self-contained sentence with the entity stated in full is easier for an AI model to parse and lift verbatim, which raises your odds of being cited.
- Tables for data: AI engines parse table cells and frequently surface the underlying numbers.
Avoid the opposite habits. Long anecdotal openings, vague language ("various factors", "many things"), conclusions buried in the last paragraph, and headings phrased as cute non-questions ("Let's talk about X") are all easily skipped. If you remember nothing else from this section, remember to put the answer in the first sentence.
Schema markup that earns its keep
Five schema types do most of the AI search work. Adding them is a one-off job and almost always handled by your CMS plugin (Rank Math, Yoast, Schema Pro on WordPress, or built-in for Shopify themes). Verify in Google's Rich Results Test before you assume it is working. If WordPress is your CMS, our WordPress SEO guide covers the plugin setup end to end.
- Article or BlogPosting: headline, author, datePublished, dateModified, image. Required for every editorial post.
- Author: name, url, jobTitle, sameAs (LinkedIn and other authoritative profiles). Pairs with Article for E-E-A-T signal.
- Organization: legal name, logo, sameAs, contactPoint. Set once at site level.
- HowTo: for any procedural content. AI engines love structured steps.
- Product and Review: ecommerce only, but mandatory if you sell. Include price, availability, aggregateRating.
One correction worth making loudly, because a lot of published advice is now out of date: FAQPage no longer earns a rich result in Google Search. Google stopped showing the FAQ rich result on 7 May 2026 and removed the supporting documentation on 15 June 2026; the old FAQPage documentation URL now redirects to the removal note in the Search Central changelog. If an agency is still selling FAQ schema as a route to rich snippets, they have not read the changelog.
That does not mean you should strip the FAQ sections off your pages. The question-and-answer format is still one of the most extractable structures for AI engines, and question-form headings still win AI Overview citations. What changed is the payoff: you are now writing FAQs for the answer engines and for the reader, not for a blue-bordered SERP widget. Keep the content, drop the expectation.
Skip the exotic. Schema types like SpeakableSpecification, ClaimReview and FactCheck are narrow-use and frequently misimplemented. They will not lift a UK SMB's AI visibility.
Off-site signals are doing more work than you think
Research from across the AI search space puts the share of branded-query AI citations coming from third-party sources at roughly 75 percent. For a UK SMB that means the work happens off your own website at least as much as on it. The good news is that this maps onto basic UK digital PR and directory housekeeping, both of which compound across SEO, AI search and conversion.
A practical UK off-site list:
- Crunchbase and LinkedIn company page: consistent founding date, headcount, headquarters, sector tags. AI models pull from these constantly.
- Trustpilot, Feefo, Google Business Profile: active review profiles with replies. Volume and recency matter more than score.
- UK trade press citations: commentary in The Drum, Marketing Week, Internet Retailing, Econsultancy. One genuine quote a quarter is more valuable than ten guest posts on low-tier blogs.
- Niche forum presence: Reddit, industry Slack groups, Stack Exchange where relevant. ChatGPT and Perplexity cite Reddit heavily on commercial queries.
- YouTube transcripts: an underrated one. AI Overviews on this very topic cite YouTube explainers alongside written sources, because the transcript is clean, structured text with a named speaker attached.
- Consistent brand entity: name spelt the same way everywhere, identical address, single canonical logo. AI models build a confidence score on entity consistency.
None of this is new. What is new is the weight AI models put on consistency and external trust signals. A polished website with thin off-site presence will be cited less than a modest website that shows up consistently across the wider web. Our guide to generative engine optimisation and LLM visibility goes deeper on the brand-mention side of this work.
How to measure AI search performance without burning budget
The free starting stack covers most of what a UK SMB needs in the first six months.
- Google Search Console: the Performance report blends AI feature impressions and clicks into the standard organic data. Watch for queries that gain impressions without clicks, which usually means you are being cited but not clicked.
- Referrer analysis in GA4: filter for referrers including chatgpt.com, perplexity.ai, copilot.microsoft.com and gemini.google.com. The traffic is small but the visit quality is usually high. We set this up step by step in our guide to tracking AI search traffic in GA4, including why a large share of it lands as direct rather than referral.
- Manual prompt audit: once a month, run your ten most commercial queries through Google AI Mode, ChatGPT search and Perplexity. Note whether you are cited, what the citation says, and which competitor appears next to you.
- Brand mention tracking: a Google Alert on your brand name and a free Mention.com account is enough to catch new third-party citations.
Paid tools are useful at scale. Otterly, Profound, Peec.ai and others run automated AI citation tracking and brand share-of-voice analysis. Expect to pay roughly £100 to £400 per month depending on prompt volume. For most UK SMBs this is a year-two purchase, not a launch-day one.
What not to do (the snake oil checklist)
The AI search space is full of paid playbooks that will not move the needle. A few patterns worth refusing:
- Paying for GEO-only audits. If the agency cannot show you the technical SEO, content and authority work underneath, they are charging premium for the acronym.
- Buying FAQ schema as a rich-result play. That result no longer exists in Google Search. Anyone quoting for it is working from 2024 notes.
- Stuffing fake first-person anecdotes. AI models are trained to spot pattern-matched experience content. Inventing case studies to look authoritative degrades trust over time.
- Mass LLM-generated content for LLM citation. The loop is self-defeating. Generated content rarely earns external citations, and AI engines have started down-weighting low-effort generated pages.
- Hidden text tricks. Whether for classic SEO or AI extraction, hidden text remains a manual action risk and AI engines now flag it as one of the lowest-trust signals on a page.
- Chasing every citation regardless of value. Some AI Overview citations send no traffic and target queries with no commercial intent. Pick the queries that matter to your pipeline, then measure those.
Most of the rest of the work is what any honest SEO programme would tell you to do anyway. Write better than the competition, get cited by reputable sources, and make sure the crawlers can reach your site.
A 90-day AI search optimisation plan for a UK SMB
If you are starting from a standing position, the rough sequence below is enough to move the needle without specialist tooling.
- Days 1 to 15: check crawler access first. Read robots.txt, check your CDN's AI bot settings, and curl your top pages to confirm the substance is in the initial HTML. Then audit your top 20 commercial pages for Core Web Vitals, valid schema, and a direct answer in the opening sentence.
- Days 16 to 30: rewrite the openings of those 20 pages so every H2 and H3 is answered in its first sentence. Keep the FAQ blocks, framed as extraction fodder rather than a rich-result play, and verify remaining schema in the Rich Results Test.
- Days 31 to 45: off-site sweep. Update Crunchbase, LinkedIn company page, Trustpilot, Google Business Profile and any trade directories. Brand spelling, address, founded date, all consistent.
- Days 46 to 60: publish two pieces of content with genuine first-party data (a survey, an internal benchmark, anonymised client numbers). These earn the citations that purely opinion-led posts cannot.
- Days 61 to 75: baseline your AI visibility manually. Run your ten most commercial queries through Google AI Mode, ChatGPT search and Perplexity. Save the answers and citations.
- Days 76 to 90: review. What moved in GSC? Which queries now have AI Overviews? Which of your pages got cited? Use the answers to plan the next quarter.
This is the work. There is no shortcut, no secret prompt, and no certified GEO badge that changes the outcome. There is only a disciplined 90-day cycle, repeated.
If you want a partner to run this work for your site, our team handles SEO services for UK businesses, including the technical audits, schema rebuilds and content programmes that put pages into AI citation range. If you are deciding what to publish next, our breakdown of the content types that win in AI search is the companion piece to this one. For broader context on whether outside SEO support is the right call at all, our guide to measuring SEO ROI walks through the honest answer, and UK SMBs starting from scratch will get value from our SEO for small business UK owner's guide.
Frequently asked questions
Is AI search optimisation different from SEO?
Mostly no. AI search optimisation shares around 80 percent of its work with classic SEO: technical health, content quality, authority and structured data. The remaining 20 percent is extraction-friendly writing, citation tracking, AI crawler access and off-site brand presence work. Any agency selling it as a separate discipline is overcharging.
How does AI search optimisation actually work?
AI engines answer a query by retrieving pages they can crawl, extracting self-contained passages from them, and citing the sources they trust most. So the work runs in that order: make sure the AI crawlers can reach your content, write passages that can be lifted whole, and build enough external brand presence that the model treats you as a credible source. There is no ranking dial to turn beyond those three.
Will AI Overviews kill my organic traffic?
Probably reduce, not kill. UK click-through rates on queries with AI Overviews are typically 30 to 50 percent lower than on queries without, but high-intent commercial queries tend to be less affected because users want to compare options before purchase. The right response is to double down on bottom-of-funnel content and brand-led queries.
Is FAQ schema still worth adding in 2026?
For AI extraction, yes. For rich results, no. Google retired the FAQ rich result on 7 May 2026 and removed the documentation the following month, so FAQPage markup will not put a dropdown under your listing any more. The question-and-answer content itself is still among the most citable formats in AI Overviews, so keep writing FAQs, just stop treating the schema as a SERP feature play.
How long does it take to appear in AI Overviews?
For a page already ranking in the organic top 10, two to six weeks after the AI feature stabilises on your query is typical. For a new page, expect the same time it usually takes to reach the top 10 plus a couple of weeks. There is no fast track.
Do I need to optimise separately for ChatGPT and Google AI Overviews?
The underlying signals overlap heavily. Extraction-friendly formatting, schema, external brand mentions and content quality all carry over. The differences are mostly in access (each platform runs its own crawler, so check them all) and in source preference, with ChatGPT leaning more on Reddit and Wikipedia while AI Overviews lean on Google's own index.
Is SEO dead in 2026 now that AI search has arrived?
No. SEO is evolving, not dying. AI Overviews and chat-based search pull their citations from the same index that classic ranking uses, so the work that wins organic positions (technical health, content quality and authority) is exactly what earns AI citations. The discipline is widening to include extraction-friendly writing and citation tracking, but the demand for trustworthy, well-structured content is rising, not falling.

