LLMs Drive Traffic, But What About Engagement?

Nikka Sulton

LLMs Drive Traffic, But What About Engagement?

AI-driven referrals are rapidly increasing, yet shifts in user behaviour are reshaping how traffic is directed to websites and influencing brand visibility. As large language models (LLMs) like ChatGPT, Gemini, and others become increasingly central to how people search for and consume information online, businesses are facing a new set of challenges and opportunities.

LLMs are now commonly used for everything from answering complex queries to generating creative content, and their role in everyday internet use is expanding. For brands and website owners, this represents both an opportunity to tap into new traffic channels and a growing concern over the rise of zero-click results. These zero-click results, where users get answers directly from search engines or AI platforms without clicking through to a website, are becoming more prevalent, and they present a potential threat to traditional web traffic.

The initial excitement surrounding LLMs often revolved around their ability to simplify workflows and provide fast, reliable answers to user queries. The convenience of receiving instant answers, without the need to visit multiple sources, made these models a powerful tool for improving user experience.

However, for businesses that depend on website traffic to drive leads, sales, and maintain visibility in the digital marketplace, the implications of these technologies are more complex and multifaceted. While LLMs can function as robust referral engines, directing users to websites that provide relevant content and information, their growing capabilities may also significantly reduce the need for users to click through to external websites altogether.

LLMs have the capacity to deliver answers, solutions, and recommendations directly within their own interfaces, making them capable of fully satisfying user intent without the user having to leave the platform. For example, when users ask detailed questions or request in-depth insights, LLMs can generate comprehensive responses directly in the chat window or search interface. This reduces the necessity for users to visit a website to gather additional information, posing a challenge to brands that rely on website visits for their online presence.

As AI continues to advance and refine its capabilities, the shift towards more direct responses and zero-click results could lead to businesses losing visibility in search engine results. Website owners and digital marketers will need to adapt to this evolving landscape by finding new ways to maintain their relevance in an AI-driven environment. This might involve optimising content for AI platforms, ensuring that content is structured in a way that is easily digestible by LLMs, or focusing on creating unique and engaging experiences that cannot be replicated by an AI tool alone.

Furthermore, while LLMs can create opportunities for brands to be discovered through AI-generated content, the nature of user intent is likely to evolve. With AI answering questions and providing information instantly, users may have less reason to actively search for websites in the traditional sense. Instead, businesses may need to focus on ensuring that their content is integrated into AI-driven ecosystems, maintaining brand presence through partnerships, and leveraging AI to drive interactions that lead to measurable business outcomes.

In the face of these changes, staying ahead of trends in AI and digital marketing will be crucial. Brands that embrace AI-driven solutions while adapting their strategies for the shifting digital landscape will be better positioned to thrive as user behaviour continues to evolve.

The promise of referral traffic: A new discovery channel

Many large language models (LLMs) are specifically designed to provide comprehensive and accurate responses to user queries. These models also have the ability to cite their sources, giving users the opportunity to explore topics in greater depth. This creates a significant opportunity for brands to be discovered by users who may not have found them through traditional search engine results.

Consider a scenario where a user asks an LLM for recommendations on the best sustainable coffee brands. After processing vast amounts of information, the LLM might provide a list of brands, along with links to their websites. This allows the user to explore the brands’ offerings, read reviews, and potentially make a purchase—all within a single platform.

This new method of discovery essentially replaces the traditional experience of clicking through multiple search results to compare options. Users can now receive curated responses directly from the LLM, streamlining the process and improving efficiency.

Similarly, a user researching a medical condition might receive a well-organised summary from an LLM, along with links to reputable healthcare websites for more in-depth information. By providing both immediate responses and helpful resources, LLMs create a more user-friendly experience.

The key difference between traditional search and LLMs lies in the way users interact with the technology. Unlike traditional search, where users actively input specific keywords, LLMs respond to more natural language queries. This shift in user behaviour allows LLMs to uncover brands and resources that might otherwise be less visible in standard search results, giving smaller or niche businesses the opportunity to be discovered by a wider audience.

As LLMs continue to evolve and become more integrated into various platforms and applications, the potential for referral traffic could expand exponentially. For example, imagine an LLM embedded within a social media platform, recommending products based on user conversations, or a virtual assistant suggesting relevant services based on the user’s calendar entries. These scenarios demonstrate how LLMs can serve as a dynamic and effective discovery channel for brands.

In this context, LLMs have the potential to play a crucial role in how users discover new products, services, and brands. By offering users relevant, contextually aware recommendations, LLMs can provide businesses with a new avenue for growth and customer acquisition, even as traditional methods of search continue to evolve.

Tracking referral traffic growth and market shifts

After analysing referring traffic over the past 12 months, I noticed steady growth in LLM-driven traffic. Indexed against the previous year’s average, referral traffic from LLMs has increased eightfold since March 2024.

This surge can be attributed to the growing use of LLM tools and the continued inclusion of links, both as citations and direct links. However, the data revealed an unexpected dip in December, which contradicted expectations for strong holiday retail traffic growth.

Looking at the mix of LLMs driving this traffic, a few notable takeaways have emerged. ChatGPT has dominated referrals since August. Prior to this, referral traffic was not smaller but rather due to different referral URLs from OpenAI and Microsoft, which eventually shifted to ChatGPT (chatgpt.com vs. chat.openai.com). This migration highlights the complicated relationship between OpenAI and Microsoft, which runs Bing.

Although Perplexity’s referral traffic appears smaller, it has doubled over the past year. However, ChatGPT’s growth has been much faster, and it now holds a larger market share. Interestingly, referral links from Meta are nearly nonexistent. The data shows only one month with double-digit referring traffic numbers. While Meta has significant volume due to LLaMA being integrated across its apps, there has been minimal referral traffic from its platforms.

Shifting focus to the zero-click conundrum, this has become a critical issue in the rise of LLMs. For some time, LLMs have been designed to satisfy user intent directly within their interface. This mirrors traditional search features like Google’s answer boxes, featured snippets, or Knowledge panels.

However, with LLMs, the impact is even more pronounced. If a user asks for the capital of France, the LLM will likely respond with “Paris” without the need to link to a website. Similarly, if a user asks for a summary of a recent news article, the LLM can generate a concise overview, reducing the need to visit the original source.

This feature, while highly convenient for users, presents a challenge for websites relying on traffic for revenue, lead generation, or brand visibility. If the LLM can effectively answer a question or fulfil a request without requiring the user to click away, the website misses out on potential visitors.

This highlights the need for a new approach, one that goes beyond simply focusing on click-through traffic. Since LLMs are designed to answer questions and keep users engaged within their interface, visibility within these models has become even more crucial. Even if there’s no direct attribution, being featured within the LLM’s responses can significantly impact decision-making.

Navigating the new landscape: How your brand can adapt

As LLMs change how users discover content, refining your approach is key to staying competitive. Here’s how you can adapt to the shift.

First, focus on expertise and original research. While LLMs are excellent at synthesising information, they typically rely on existing content. By offering unique insights, conducting original research, and showcasing deep expertise in your niche, you’ll increase the likelihood of being cited as a source. This, in turn, can help you earn valuable referral traffic.

Next, optimise for natural language queries. Understanding how users phrase questions naturally is crucial for effective content creation. By optimising your content to directly answer these types of queries, you can increase the chances of an LLM citing your website. This approach helps make your content more likely to be discovered within these models.

Another key strategy is to build brand authority. A strong reputation and authority within your industry can significantly increase the chances of an LLM recommending your website as a trusted source. Focus on building credibility and thought leadership to position your brand as an expert within your field.

It’s also important to explore new forms of engagement. As LLMs evolve, new ways to interact with them may emerge. This could involve developing specialised content formats or leveraging APIs to provide information directly to LLMs, giving you more opportunities to be featured in responses.

Monitoring LLM performance and attribution is equally important. Actively tracking how LLMs reference your content will help you understand how much traffic you are receiving from these sources. By analysing this, you can refine your strategies over time to improve visibility and engagement.

In addition, consider the user experience. While LLMs are designed to provide answers directly, even if users don’t immediately click through to your website, the information they receive might lead them to explore your brand or services at a later stage. That interaction, even without a click, can still prove valuable in the long run.

Embracing the evolution of LLMs is crucial. These models are fundamentally changing how users find and consume information, with rising referral traffic but also the challenge of zero-click results. Brands that create authoritative content, optimise for natural language queries, and monitor attribution will be better positioned to remain visible— even when clicks aren’t guaranteed.

Adapting to this shift isn’t optional; it’s essential for long-term digital success. Being proactive and refining your strategies to align with these changes will help ensure your continued relevance and visibility in a rapidly changing landscape.