The tipping point

Nick Mullen
Monday 11 August 2025

Recently, I’ve become increasingly intrigued by the moment when AI-driven interactions like chatbots and AI-powered search overtake traditional search engines in popularity. This tipping point feels like more than just a technical milestone, it marks a profound shift in human behaviour.

Whilst there has been a significant increase in the use of AI it is important to recognise that traditional search is still the dominant way in which users search. That picture is changing quickly. According to a recent Adobe report, AI search referrals surged by an astonishing 1,300% during the 2024 holiday season compared to the previous year. For me, the tipping point signals the transition of AI from a novelty, or trend, to a foundational element in how we access and process information.

Search has long been a cornerstone of the web development industry. It has shaped how we design, structure and optimise websites. But as a web developer, I see this shift as something that challenges the core principles of our field.

Search Engine Optimisation (SEO)

SEO is the art and science of tailoring webpages so they rank highly on search engines. While the exact algorithms remain opaque, we broadly understand that fast-loading, well-structured pages with regularly updated content tend to perform well.

Historically, strong SEO has been critical to success. You can create exceptional content, but without good SEO your pages risk being invisible. AI platforms, however, function differently. Like search engines, they crawl and index content, but rather than simply ranking links, they train models on this data to generate conversational responses. This means content with good SEO isn’t guaranteed to perform well in AI chat interactions. It is possible that content creators may optimise for Google, yet miss their audience entirely in AI platforms.

Gartner predicts that organic search traffic will decline by 25% by 2026.

To complicate matters further, the traditional search market has been dominated by a single player – Google. Tailor your content for Google, and you’re mostly covered. But AI is a fragmented space. There is no single dominant chatbot or AI search engine, and the number of platforms is growing. This makes optimisation far more complex and dynamic.

We’re entering a world where being good at SEO alone is no longer sufficient. We must now optimise for both search engines and AI systems. As traditional search cedes ground to AI, the overall effectiveness of SEO is likely to decline.

Web analytics

Web analytics has long helped us measure performance by tracking user behaviour on our sites, however, growing privacy concerns and regulations have led to more users opting out of tracking, especially younger audiences. As a result, we often end up analysing only a partial, and potentially skewed, dataset.

Even more concerning is the invisibility of AI interactions. If a prospective student asks ChatGPT a series of questions about one of our courses, and the answers come from our website, we won’t see any of that traffic. There’s no pageview, no click-through, no record, just invisible engagement.

For instance, in a US$60 million deal, Reddit licensed its data to Google for AI training, leading to a surge in visibility via AI answers but this traffic is often ‘invisible’ in traditional analytics tools. These posts may influence millions of AI-generated responses, yet generate no pageviews or referral data.

Current analytics tools are designed around monitoring what happens on our pages. They’re not built for a world where our content is lifted and used as an invisible source in off-platform conversations. As AI chat becomes more prevalent, analytics will become less representative of reality. At what point does it become misleading to make statements about the performance of our webpages based on web traffic analytics?

Digital marketing

Traditional digital marketing is built around driving traffic, then guiding users through a conversion funnel, from landing pages to final actions (like sales or sign-ups). Crucially, this model depends on data and control: we track what’s working, adjust our approach, and shape the user’s journey.

But today’s users increasingly expect a conversational experience, asking questions, receiving real-time responses, and engaging in two-way dialogue. And these conversations often happen beyond our reach, on platforms like ChatGPT.

If a user wants train times, they don’t click through to a website anymore – Google tells them directly. The same is happening with AI tools like Alexa and ChatGPT, where direct answers replace traditional website visits.

Since the launch of Google’s AI Overviews, several publishers have reported double-digit traffic drops. Users now often receive complete answers at the top of the page, bypassing the original source entirely and possibly the funnel entry. AI visibility doesn’t always equate to web traffic or measurable engagement.

So, how does digital marketing adapt to this new reality? When we lose visibility (analytics), control (conversion funnel), and even the ability to directly influence the conversation, what does effective strategy look like?

We’ve studied, trained and built entire careers around practices that are now shifting rapidly. This doesn’t mean everything becomes obsolete overnight, but it does mean we need to reassess the relevance and effectiveness of long-established practices in an AI-first world. Once again, the web industry must evolve.  After all, that tipping point might be closer than we think.

Related articles

Cookie preferences