5 Things you need to know

From LLM-powered platforms like ChatGPT to the AI Overviews in Google’s search results, AI search is already having an impact on brand visibility and clickthrough rates. But the rate of change may be about to speed up with the arrival of Google AI Mode.

Currently rolling out to 180 countries worldwide, AI Mode launched in the UK on July 28th.

Instead of providing users with a ranked series of links for users to click on to piece information together, AI Mode responds to queries with conversational paragraphs, backed by linked citations to sources. While Bing already offers something similar with Copilot search, the sheer market dominance of Google means that AI Mode is likely to dramatically increase the consumer adoption of AI search.

All of this gives marketers possibly their strongest incentive yet to adapt quickly, so they can maintain or even improve their brand’s visibility in search rather than risk watching it erode.

This urgent need to do something makes the idea of an easily implemented templated approach extremely tempting. And, predictably, there are already plenty of supposedly “AI-optimised” content templates, frameworks and page layouts out there, promising to improve the chances of your content being cited in AI-generated responses.

But before you invest in optimising your content with one of these template-based approaches, do they work?

The short answer is no.

The slightly more nuanced answer is, it’s complicated.

Here are five things you need to know to understand what really matters when it comes to AI search.

1. Why content templates won’t impress LLMs

When discussing AI search, we must always be careful to distinguish between Google’s AI products (AI Mode and AI Overviews in search) and LLM systems like ChatGPT, Claude, or Perplexity. The way these AI models operate and validate information is fundamentally different.

When an LLM processes content, it first strips away all HTML tags, CSS styling, and JavaScript. An LLM only sees plain blocks of text, without your carefully crafted page formatting, navigation structures, or visual hierarchy.

There can be no such thing as an AI-optimised template because, as far as LLMs are concerned, your template is invisible.

2. Why Google AI Mode is not like ChatGPT

Unlike standalone LLMs, Google’s AI products use a combination of Google’s traditional search infrastructure overlaid with LLM processing.

Google AI Mode and AI Overviews use a ‘grounding’ mechanism, checking its responses against Google’s search database to reduce the risk of hallucinations and improve accuracy. This is why your traditional SEO foundations still matter for visibility in Google’s AI products.

ChatGPT and similar LLMs can’t ground their information in the same way. Instead, they use vast amounts of training data from sources like Common Crawl, supplemented with real-time search to identify fresh sources and validate information.

However, this means ChatGPT is effectively using the same fallible internet as you or me, without the benefit of all the authority signals, knowledge graphs and so on that power Google’s search infrastructure.

This difference has strategic implications. Content that performs well in traditional Google search has a better chance of being referenced by Google AI products because both systems draw from the same underlying infrastructure. Strong E-E-A-T signals, topical authority, and technical SEO fundamentals are still relevant.

3. Schema markup: the indirect truth

As LLMs strip away any HTML code from the content before processing the raw text, this means any schema markup is also stripped away. Ergo, LLMs can’t process your schema information to understand how your content is structured.

Yet, Google, Microsoft, and ChatGPT have each said that structured data and schema markup does help LLMs to better understand digital content.

No wonder marketers get confused.

The truth is that schema markup has an indirect influence on how LLMs find and prioritise information. LLMs will use search data from Google, Bing or elsewhere, as well as information from key content partners to identify, prioritise and validate information before generating a response. And structured information such as schema can influence and feed into this third-party data.

For example, while ChatGPT’s LLM doesn’t directly ‘read’ the schema markup when processing content, its shopping results draw on third-party websites and data feeds that do use structured metadata to provide information on pricing, product descriptions, and reviews.

So, schema markup is important in the same way it always has been; providing algorithms with the structured data to define entities and relationships, provide context and build knowledge graphs. It’s just that you now also get an extra, indirect benefit from AI systems drawing on the same information via third party sources.

4. Why ‘above the fold’ doesn’t matter in AI Mode

Content marketing advice often recommends putting the most important information within the first 800 pixels of a webpage; above the fold.

The logic seems sound: a short summary or scene-setting opening quickly confirms to the reader what the content is about. So surely AI systems would also benefit from key information at the top of page that clearly signals the content’s topical relevance to a particular query.

Recently, we set out to test this hypothesis. AI Mode often includes text-fragment links in its responses, taking the user directly to the relevant highlighted text within the cited content. By measuring the exact pixel depth of each of these highlighted text fragments, we could see if AI Mode was more likely to surface content from within the first 800 pixels.

What we found was that AI Mode will pull information and text from anywhere on the page, regardless of pixel depth. The most relevant snippet counts, not where it sits within a page layout.

5. Descriptive subheadings do matter in AI Mode

While we found that pixel depth is not a factor, our research did reveal one interesting pattern. We found that AI Mode tends to select text fragments that consist of a subheading followed by the first line of accompanying text.

This suggests that AI Mode is using the subheadings to navigate the informational hierarchy within the content.

But then we already knew that descriptive headings and strong opening lines were important to Google. This isn’t new advice to help you optimise your content for AI. This just reiterates the importance of long-standing content best practice.

Conclusion: the real strategy for AI visibility

You might be disappointed to learn there is no magic template to optimise your content for AI. There is no silver bullet.

On the other hand, this means you don’t need to invest/waste lots of money restructuring your content.

What worked for SEO is still important for AI search, even if some of the impact is more indirect. Focus on information hierarchy, descriptive subheadings, and comprehensive topic coverage

The best approach is the same one that’s worked for years: create well-structured, authoritative content that genuinely serves user intent.

Need help improving your AI visibility? Get in touch