Dan Starmer, Head of SEO, Catalyst Marketing Agency, on how businesses and marketers need to tap into the data to better understand customer intent and refine lead generation and conversion strategies.
It seems 38.48% of marketers see the measurement of AI search engine success a challenge. That’s huge. AI-search engines are growing stronger by the day, by the hour and by the second, eating terabytes of data and with each byte comes exponential knowledge. So, marketers have a job on their hands to keep apace, to adapt and to please their clients, external or internal.
So, if tech is progressing at such a rate that it’s impacting traditional search, how do marketers A) learn how to harness the power and benefits of LLMs and B) measure it. After all, if it can’t be measured, how do we set KPIs?
Furthermore, how do we maintain trust in what the results are showing us, when a recent study from the Tow Center for Digital Journalism suggests that “AI search engines are inaccurate 60 percent of the time.”
First up, where are we right now?
The rise of LLMs is reshaping how we find and consume information, shifting away from the traditional search engine model that relied on keyword optimisation and predictable user journeys.
Historically, marketers and SEO professionals would craft content around clear search intent, optimising for specific queries and guiding users toward an intended target; a landing page on a website.
However, LLMs provide direct, conversational answers without requiring users to click through to external sites, leading to a surge in zero-click searches, which have increased by 30% with the rise of AI-generated responses. They’re ‘answer engines’ rather than ‘search engines’.
This results in reduced traffic from organic search and a need to rethink content strategies. If users get what they need without clicking, conversion opportunities drop significantly. Instead of solely optimising for rankings, brands must focus on authority, brand presence and engagement. Structured data, first-party data and partnerships with AI platforms may become critical.
Likewise, LLMs increase competition for visibility, and marketers must compete for inclusion in AI summaries rather than just ad placements or organic rankings.
Additionally, value-driven, unique content that LLMs cite or reference will be key to maintaining visibility, and search teams will need to work much closer with their digital PR colleagues to ensure rich content is unique and picked up by high authority publications and sites. As search continues to evolve, marketers must adapt to a model which looks beyond traditional search engine result pages.
Can LLMs be wholly trusted?
A study titled Epistemic Integrity in Large Language Models highlights a significant issue of “epistemic miscalibration,” where LLMs exhibit a mismatch between their linguistic assertiveness and actual certainty. This misalignment can lead users to overestimate the reliability of the information provided.
Furthermore, in the medical field, the stakes are much higher. The paper Analyzing evaluation methods for large language models in the medical field: a scoping review points to the absence of a clear framework for evaluating LLMs in medical applications, raising concerns about their dependability in critical contexts.
The Techspot piece referenced above even goes so far as to suggest that LLMs are programmed to answer everything asked of them, and that “they report with complete authority that what they say is true even when it is not.”
But does this matter? We’re just trying to generate leads, right? Should we care about authenticity if our objective is to drive traffic and boost DA?
From an SEO professional’s perspective we have to understand and appreciate that there are risks with authenticity around anything that’s online, not least social media where the exploitative spread of misinformation has been well documented.
All we can do is continue to shape our content strategies that align with the traditional search engines and LLM’s and be committed to driving transparent results. We have always had to be agile, this is just something else to test that agility.
And, as my PR colleagues continually tell me, while there exist myriad ways to produce AI-generated content and then run it through another AI tool to purchase a bunch of “guaranteed backlinks”, somebody somewhere still wants to read informed, educational content too. Otherwise, the internet will just become full of content designed to take you somewhere else – without actually being read.
We also have to appreciate that LLMs provide richer but more ambiguous data compared to traditional models. Their responses are context-aware, generative and nuanced, offering deeper insights.
But they are lacking the predictability of keyword-based search results. While LLMs unlock deeper insights, refining trust, and verification remains key to harnessing their full potential.
Rethinking content strategies
Despite the arguments around authenticity, one thing’s for sure and that is LLMs are not only here to stay, with 57.6% of marketers reporting more SEO competition than ever before, they’re here to dominate. To continue to generate and convert leads, we must adapt and shift attention from just ranking on search engines to ensuring our content is cited, referenced and embedded within these AI overviews. Likewise, digital PR strategies must seamlessly integrate SEO, ensuring keywords are naturally embedded in messaging while also optimising for AI to recognise GEO-specific content.
Rather than relying on rigid keyword stuffing, we should focus on natural, context-rich content that aligns with AI-generated responses, making it more relevant and engaging for users. In addition, we should adopt a multi-channel approach to our content such as video and interactive media to improve accessibility and user, or customer, engagement.
LLMs also give priority to structured sources for generating responses, and from an SEO perspective we need to ensure our content does just that, such as optimising for featured snippets, while also authoring authoritative content to reinforce credibility.
Moving forward, to harness the power of LLMs there will be a greater dependence on digital PR and even paid search with fewer organic clicks for driving traffic and conversions.
Ultimately, the era of AI-generated results – or answers – isn’t a threat to lead generation but a huge, still-untapped opportunity for those ready to adapt.
