For twenty years, being found online meant one thing: ranking on page one of Google. Buyers typed a few keywords, scanned ten blue links and clicked. SEO was built for that world then, and it still matters today.
However, buyers no longer just search. They ask.
Rather than being served a page of links, ChatGPT, Gemini, Perplexity and Google’s AI Overviews hand them a short list of trusted names, and 94% of B2B buyers now use LLMs during the buying process. The shortlist is often written before anyone visits your website.
So what’s actually different about optimising for AI search, and where does traditional SEO still do the heavy lifting? Here’s how we see it at Hallam, as a B2B agency that specialises in both.
What does traditional SEO do well?
SEO earns organic rankings in search engines and turns them into clicks. It hinges on three core factors:
- Technical SEO: makes sure search engines bots can crawl, render and index your site.
- On-page and content work: matches pages to the keywords people use, with the depth to deserve a high ranking.
- Off-page work: earning quality backlinks through link building, signals that other sites trust you.
Success is easy to read: rankings, organic sessions, click-through rate and the leads those sessions produce. The unit of competition is the page, and the prize is the click.
None of that has gone away. Google still handles huge search volumes, and AI Overviews draw heavily on pages that already rank well. A site search engines can’t read is a site AI models will struggle with too.
What is AI search optimisation?
AI search optimisation, also called generative engine optimisation (GEO) or answer engine optimisation (AEO), is the work of improving how often and how accurately AI platforms cite, mention and recommend your brand.
The goal shifts from ranking a page to getting your brand named in the answer, often without a click. Nearly 70% of Google searches in the UK now end without one. If a buyer asks an AI tool to compare suppliers and you’re not in the answer, you’re not on the shortlist.
That means AI visibility depends on far more than your website. Models build their view of your brand from everything written about you, which makes it as much a brand challenge as a technical one.
What are five ways in which AI search differs from SEO?
1. The goal: named, not ranked
SEO competes for position on a results page, whilst AI search competes for a place among a handful of recommended names, and for being described accurately once you’re there. Fifth place on page one still earns some traffic. Missing from an AI answer earns you nothing.
2. The starting point: buying situations, not keywords
Traditional SEO starts with keyword research. We start AI search with category entry points: the real situations that make a buyer start looking, such as an audit deadline, board pressure or a product launch. Those situations shape the questions buyers put to AI, so we map them before writing anything.
3. The content: written for two kinds of prompt
AI platforms don’t all behave the same. Prompts to Google’s AI Overviews come from search, so they’re short and keyword-led, at around 25 characters. Prompts to ChatGPT and other LLMs are more than twice as long and conversational. AI Overviews favours concise, definitional content, while LLMs want deeper, explanation-first answers. Treat them as one channel and you won’t meet the needs of either.
4. The signals: off-site brand matters most
Backlinks have always been part of SEO. In AI search, a much wider set of off-site brand signals carries the weight. Earned media makes up 84% of the sources ChatGPT, Claude and Gemini cite (Muck Rack, 2026), and branded web mentions can be up to three times more influential than backlinks in driving LLM citations. Press coverage, G2 and Trustpilot reviews, Reddit threads, LinkedIn discussion, awards and YouTube all shape what large language models say about you.
5. The measurement: Share of LLM, not just rankings
A rank tracker can’t tell you whether ChatGPT recommends you. We measure AI visibility with Share of LLM, our own measure of how often AI models name your brand compared with competitors, and how they describe you. We then tie in Custom GA4 tracking then follows visits from AI platforms through to leads and pipeline.
Where do SEO and AI search overlap?
Much of what makes a site rank also makes it readable to AI: crawlable pages, clean structure, schema and structured data. Entity mapping helps models connect your content to your brand. Clearly authored, reference-grade expert written content earns rankings and citations alike.
Paid search belongs in the picture too. We plan paid and organic together under unified search, spot where they overlap, move spend away from terms a client already wins organically and report on overall search profitability rather than channel by channel.
The real difference is scope. SEO is one part of AI search optimisation, not the whole of it. If you do SEO well, you’ve built the foundation, but if you stop there you’ll ignore most of the signals models rely on.
How we do both: Total Search™
The brands people remember are the brands AI recommends most often. Our research across four B2B sectors found that brands in the top half for Share of Search were 2.5 times more likely to be in the top half for Share of LLM. Buyers and AI models form their view of you from the same sources, so they usually land on the same names.
In other words, mental availability (coming to mind when a person enters the category) and model availability (coming up when an AI is asked the same question) are built by the same levers. That’s why we deliver AI search through Total Search™, our methodology for making B2B brands visible everywhere buyers search. It covers four areas:
- Category authority: brand strategy, positioning, thought leadership and original research that shape how your category is understood.
- Content authority: pillar pages, guides and FAQs built to be the reference people and platforms rely on.
- Distributed authority: digital PR, reviews, community, social, events and YouTube that build credibility where you don’t control the conversation.
- Technical authority: technical SEO, entity mapping, schema, crawl health and server-log analysis, so AI and search engines can find and read you.
Every discipline is in house, working to one plan. Our Total Search™ dashboard tracks Share of LLM, citations, share of search, unified organic and paid, AI-referred traffic and YouTube in one view. Share of LLM was central to our winning entry for Best LLM Adoption & Integration by an Agency at the European Search Awards 2026.
So, do you need SEO or AI search optimisation?
Your brand needs both, working as one. SEO makes sure search engines and AI can find and read you. AI search optimisation makes sure that when a buyer asks, your brand is the one in the answer. There’s no shortcut: no brand investment means no visibility.
Want to know where you stand today? We can start with a Share of LLM baseline against your competitors. Get in touch or read our guide to AI search.
