Skip to content
Menu
As AI-powered tools such as ChatGPT, Claude and Google’s AI overviews, become more embedded in the search journey, brands need to consider how they structure and present their online content.
Fire on the Hill director Nikki Scrivener speaks to AJ Handley-Rowe, director at digital marketing specialists, Salt, about how brands can improve their chances of appearing in LLM-generated search results, why human-first content still matters most and where businesses should prioritise their efforts.
Nikki Scrivener: How important is content structure when it comes to appearing in AI and LLM search results?
AJ Handley-Rowe: It’s important, but it shouldn’t feel daunting – most of the principles behind it aren’t particularly new. AI crawlers tend to favour some of the things that search engines have rewarded for years – content that’s readable, scannable, well-structured, genuinely useful and written for humans.
That means breaking content up clearly, using logical headings, making pages easy to navigate and delivering genuinely valuable information.
One mistake brands often make is trying to write for AI at the expense of the audience. If the content itself isn’t interesting, useful or different, it won’t help you achieve your objectives. Everything still comes back to marketing fundamentals: right audience, right message, right time.
NS: Which types of content perform best?
AJHR: Summaries tend to do particularly well. AI tools favour concise introductions, executive summaries and clear snippets at the top of a page because they help users and crawlers quickly understand what the content is about.
That really just mirrors human behaviour. Most people want to get to useful information quickly, a bit like skipping straight to the ingredients on a recipe page. Content that surfaces key points early, with clear summaries at the start and useful wrap-ups at the end, performs better in both traditional SEO and AI-generated answers.
FAQs are also important. A large proportion (up to 50 per cent) of Google and AI searches are phrased as questions, so structuring content around the kinds of things people naturally ask is very effective.
For longer-form content like blogs, guides and educational resources, it also helps to think about the questions a reader might ask as they go – and answer them throughout the page. e.g.:
This makes the content easier to scan, more useful for readers and more snippet-friendly for AI tools. One of our clients recently added structured FAQ content to a small set of pages and quickly saw improvements in rankings, traffic and AI-driven brand mentions.
Comparison and decision-making content also performs well – things like product comparison tables, “this vs that” content, pros and cons lists, buyer guides and step-by-step decision frameworks. These formats align with how people use search engines and AI tools when they’re researching or evaluating options.
NS: Does that mean re-writing existing content from scratch?
AJHR: In most cases, existing content can be retrofitted and improved rather than discarded. The focus should be on refining and restructuring it – adding summaries, improving formatting, introducing FAQs, tightening relevance and improving internal linking.
Even content that feels outdated or unfocused can be improved enough to start generating value again. That makes this a relatively quick win for many organisations.
NS: Should brands start producing even more content?
AJHR: The issue isn’t really about volume – most brands are already producing blogs, case studies, guides and newsletters. The bigger challenge is whether the content is structured and targeted effectively.
The brands performing best tend to combine:
It’s not about just publishing keyword-heavy articles like the SEO of the early 2000s.. Quality, relevance and usefulness matter much more and also let you sleep well at night as content owners – you know you’ve done a good job.
NS: How should brands prioritise where to start?
AJHR: Rather than guessing what users want, companies should look at a combination of:
The goal is to identify the overlap between:
This helps to prioritise the pages most likely to generate impact quickly. Interestingly, the homepage is rarely the best place to start. They’re often broad and multi-purpose, whereas search engines and AI models tend to favour more focused pages built around a specific topic or intent.
NS: How important is the technical side of optimisation?
AJHR: Good technical SEO practices help search engines and AI crawlers understand and navigate a site more effectively.
That includes fundamentals such as:
Beyond that, structured data and schema markup – code that labels content clearly for crawlers – are also important. While schema alone isn’t a silver bullet, it does provide stronger signals to search engines and AI models about the type of content on a page, e.g. FAQ or authored content, and how it should be interpreted.
The best place to start is with an audit covering:
From there, businesses can identify both quick wins and longer-term opportunities.
NS: Are AI mentions actually driving meaningful business impact?
AJHR: Yes, although not always through direct clicks. One trend we’re seeing is that when brands start appearing more frequently in AI-generated answers, they often see an increase in branded search traffic.
Users might discover a company via an AI response initially, absorb the information and then return directly to search for the brand when they’re ready to take the next step. So, while click-through rates from AI interfaces themselves may be relatively low, visibility within those environments is still influencing awareness and consideration.
Technology may be evolving quickly but the fundamentals of good communication haven’t changed a bit.
NS: Want to assess how AI-ready your website content is?
AJHR: A combined SEO, AI visibility and content audit can help identify:
For SEO, AI, and all things digital, get in touch with Nikki or AJ to find out more.
