ChatGPT and other large language models (LLMs) are rewriting the rules of search and many brands don’t even realise they’re vanishing from search results. As LLMs shift search behaviour from keywords to questions, and show answers rather than links to websites, strategies rooted in SEO are losing their edge. CMOs and content strategists alike will soon be facing an uncomfortable truth: our content might still exist, but is it still visible?
To explore this turning point in digital content strategy, Juliet Stott sat down with Tom Salvat, a content intelligence expert and co-founder of Elelem.ai, a Generative Engine Optimization (GEO) platform designed to help brands stay visible in an AI-led world.
In this candid conversation, Tom unpacks the black-box nature of LLMs, explains why authenticity and utility still matter, and offers grounded, actionable insights for CMOs who want to stay ahead of the curve.
Juliet Stott (JS): How do AI models, specifically LLMs, search for content?
Tom Salvat (TS): These models, like ChatGPT, function similarly to Google in that they crawl, scrape, and index the web ahead of time so they can quickly provide answers to queries. But unlike traditional search engines, which give you a list of URLs to explore, ChatGPT aims to directly give you the answer.
JS: Do these models differentiate between content written by humans and content created by AI?
TS: From what I’ve been told by contacts at Google, they don’t differentiate. They need AI-generated content to scale and continually improve. It’s less about who created it and more about how effectively the content answers the query.
JS: There’s a shift from keyword-based searches to question-based searches. Could you elaborate on this?
TS: Absolutely. With traditional search, users typically input keywords to find information. However, the emergence of LLMs like ChatGPT means people now more naturally pose questions. This shift significantly changes content strategy, requiring businesses to create content explicitly structured around answering complete questions rather than targeting fragmented keywords. Moreover, the LLMs often reformulate the user input (whether keywords or fully formed questions) when trying to retrieve relevant information to respond with. In short: keyword strategies are no longer sufficient.
JS: What practical advice can you give to content creators to improve their content’s visibility in AI searches?
TS: If you think of traditional SEO logic – content that answers a user’s query effectively – you’re on the right track. Tools that analyse Bing’s hierarchy might also help, given Microsoft’s involvement with OpenAI. Ultimately, it’s about creating content that clearly answers common queries your audience has. Also, content intelligence tools like ours help identify questions customers are asking, as well as how LLMs are reformulating these questions to improve retrieval, which informs content strategy significantly.
JS: With the proliferation of AI-generated content, will original, journalist-written content hold its value?
TS: Yes, absolutely. High-quality, human-written content should become even more valuable because it builds trust. While AI-generated content can be commoditised, genuinely insightful and original pieces will cut through the noise more effectively. Journalists and brands that “own” their audiences without relying on search traffic will continue to prosper.
JS: What led you to establish your Generative Engine Optimization (GEO) platform Elelem.ai, and what’s the problem you aim to solve?
TS: We initially developed our previous company around content intelligence, helping businesses understand what their audience wanted rather than what the brands wanted to say. After a decade in content and AI, we saw an opportunity with the rise of ChatGPT and AI search. Businesses need to understand and optimise their content to ensure visibility within these LLM-driven searches.
JS: How exactly does Elelem.ai help marketers understand and leverage AI-driven traffic?
TS: Our platform clones a client’s website, making it optimised for AI to quickly retrieve answers. We also track how often the client’s site appears as a source for AI-generated answers, giving insights into brand citations and mentions. Furthermore, our virtual API ensures that any AI agent accessing the site shares the questions it seeks answers to, providing invaluable real-world query data.
JS: Do companies need a new kind of team member to handle this content optimisation for LLM search?
TS: Typically, it’s the SEO or content teams handling this, sometimes data analysts. But our tool simplifies this by directly providing briefs and scoring how effectively your content answers key queries, aligning closely with how an LLM assesses content relevance.
JS: Will traditional websites evolve drastically because of LLMs?
TS: Websites will undoubtedly change, but how quickly that happens depends greatly on human adoption. Despite ChatGPT’s usefulness, adoption rates remain surprisingly low. While change is inevitable, it’s hard to predict exact timelines due to the complex mix of economic, social, and technological factors.
JS: Finally, how can marketing agencies or smaller brands practically use your platform, especially those in niche B2B markets?
TS: Even niche B2B businesses can benefit significantly. We begin by identifying relevant keywords, then our tool shows precisely how visible the content is in AI searches, helping guide strategies to improve visibility. The system’s recommendations then directly feed into content optimisation, ensuring even smaller brands stay competitive in AI-driven search environments.
How to make your brand findable in AI search
As AI-powered search engines like ChatGPT, Gemini, and Claude increasingly shape how users discover content, a growing number of companies are exploring tools that can help them adapt. One such platform is Elelem.ai, which offers what it calls “Generative Engine Optimization” – an approach aimed at understanding and improving how brands are referenced in LLM-generated responses.
For CMOs, tools like this are appealing for several reasons:
- They provide real-time insights into how a brand appears in AI search results.
- They benchmark visibility against competitors and broader trends.
- They offer structured, data-led recommendations to improve content visibility in AI-generated answers.
- Some even attempt to simulate how language models “see” your content to guide strategic updates.
Elelem’s model is subscription-based, with pricing that scales based on usage:
- Entry-level packages start at $300/month
- Most users fall into a mid-tier around $600/month, which includes broader access and query tracking
- Enterprise-level pricing is available for those with more complex requirements
While Bright Star is not currently using this platform, we’re actively exploring it as part of a wider investigation into how clients we work with can stay discoverable in the fast-evolving world of AI-driven search.
Want to know more? Let’s talk.
