You might be bracing for a dystopian tale of Large Language Models (LLMs) overthrowing the mighty Google and its dominant position in search marketing. We aren’t quite at this stage yet, but we’ve decided to shed light on what’s happening and launch the new Coppett Hill AI Index, which will track the intersection of generative AI and search marketing every month.
Why does this matter?
It seems like every other day there’s a new headline proclaiming the death of search engines such as Google and Bing, and the rise of ChatGPT, Perplexity and other LLMs. Add in a judge ruling against Google in a major case over its search monopoly, and the DOJ reportedly planning to demand the sale of its Chrome browser, and it’s no wonder CMOs are questioning (and being questioned about) their search strategies.
At Coppett Hill, we decided to sift through the rhetoric and flashy headlines to see where the data actually leads us. It’s important to acknowledge that there’s no single comprehensive data source available; most sources provide only fragments of the story. By piecing together various information sources available, as well as our own data, we’ve worked to present a clearer and more complete picture.
There are two primary ways generative AI is currently reshaping search and search marketing:
LLMs taking traffic from search engines as users change their online research behaviour.
Search engines’ AI overviews reducing clicks to advertiser websites through both paid and organic rankings.
The Case for LLMs Disrupting Search
There’s no denying the appeal of using LLMs for web searches. Why wrestle with robotic searches like “best restaurant London cheap” when you can ask an LLM chatbot, “Where can I find affordable but good restaurants for a casual meal with friends in central London?” LLM’s tools offer conversational, intuitive ways to search—and that’s shaking things up.
A recent survey by Evercore asked over 1,300 U.S. respondents about their search habits. They found that 8% of people now use ChatGPT as their primary search engine, up from just 1% in June. That is a significant rise. Meanwhile, amongst those surveyed, Google’s share of users fell from 80% to 74% during the same period.
In fact, Gartner has predicted that by 2026, traditional search engine volume will drop by 25% due to chatbot-like LLM applications.
While that might sound dramatic, let’s examine the data. One of the most direct ways we can measure this shift is by looking at LLM referrals - instances where users engage with AI-generated responses by clicking through to websites.
To be clear, LLM referrals occur when a user asks a language model a question and clicks on a link provided in its response. Is this the future of search? Should CMOs be racing to maximize their visibility in LLM-generated results?
Introducing the Coppett Hill AI Index
To help answer this question for our clients, at Coppett Hill we’ve developed a new monthly AI Index: a structured framework designed to measure AI-driven traffic and give marketing teams a better understanding of how AI search impacts website visibility and engagement.
Our AI Index covers a mix of B2B and B2C companies across multiple sectors, and covers two trends:
AI referral traffic from large language models (LLMs) to advertiser websites.
The presence of Google’s AI Overviews on the search results page for keywords which drive organic search traffic to advertiser websites.
What’s happening with LLM traffic to websites?
Since October 2024, our AI Index highlights a sharp increase in LLM-attributed traffic, particularly as a proportion of overall organic search. While LLM-driven referrals are still a small proportion of overall traffic, their rapid growth suggests that marketing teams will need to rethink their organic strategies in the coming years.
Coppett Hill LLM Referral Traffic Tracker
Are Search Engines Really Losing Ground?
While LLM-driven search is showing signs of growth, does this mean search engines are losing their grip?
Let’s consider Google, as the dominant search engine in Europe and the US. So far, the numbers don’t point to a company in trouble. In Q4 2024, Google’s ad revenue remained strong, reaching $72 billion - a 10.6% year-over-year increase from $65 billion in Q4 2023. This growth pushed total ad revenue for FY24 to approximately $256 billion, up from $237 billion in 2023, reinforcing Google's dominance in digital advertising.
However, if the trends we’ve seen in the last three months continue, Google and other search engines should be worried – fewer users starting their research journeys via search engines means fewer organic search visits for advertiser websites, but also fewer paid search clicks earning revenue for the search engine. It is also reasonable to think that search engines will be more focused on protecting paid search clicks than organic search visits, so revenue data isn’t going to tell us the full story.
The Big Picture
The story isn’t as simple as LLMs vs. search engines. Yes, LLMs are gaining traction, and some users are migrating to these platforms as a starting point for online research. But the overall impact on search engine volume and revenue is not yet dramatic – albeit this depends heavily on the industry. For instance, few users currently turn to LLMs to book flights or shop for products, but many rely on it for tasks like researching and learning (where some advertisers such as education providers and advice sites are reporting a more significant impact).
Search Engines’ Double-Edged Sword
If LLMs represents external competition, search engine AI Overviews (AIOs) pose a more complex challenge from within. By integrating generative AI into their own search results, search engines have created a double-edged sword. On the one hand, these summaries improve the search experience by providing concise answers directly on the results page, saving users from sifting through multiple websites. On the other, they diminish the need for users to click through to brand websites, cutting into organic traffic and pushing those hard-earned organic rankings further down the page.
While still speculative, concerns about these effects are gaining traction among media executives and SEO experts. Forecasts suggest that organic search traffic to publishers’ websites could decline by as much as 20% to 60% due to AIOs. But, in true Coppett Hill fashion, let’s look at the data again to see how much weight these concerns actually hold.
Examining The Data
Firstly, a Statista report found that for news-related queries, the first organic search result is pushed down by an average of 980 pixels—equivalent to a full-page scroll. This makes it significantly harder for users to engage with organic links.
BrightEdge data from June 2024 highlights notable shifts in the deployment of AIOs, reflecting Google’s ongoing changes to this feature:
The prevalence of AIOs declined from 11% to just 7% of total queries, suggesting a more selective application, perhaps a result of well-publicised accuracy issues.
The impact varies significantly across industries:
Education queries saw a reduction in AIO appearances from 26% to 13%. This is one of the sectors where AIOs have been most frequent (sparking a recent lawsuit from US education provider Chegg, which claimed a 50% YoY reduction in website traffic in January 2025)
Meanwhile, AIOs almost disappeared for entertainment-related queries
These findings suggest that AIO impacts are industry dependent. It’s also clear that AIOs are still in a testing phase and their use is likely to evolve rapidly.
How do AIOs influence user behaviour? SEER Interactive analysed 7,800 Google queries from June to September 2024 to address this question. Their findings include:
Presence: AI Overviews appeared for only 7% of queries that feature paid ads, accounting for just 2.2% of total impressions, reaffirming their minimal impact on paid search performance overall.
Click through rate (CTR): where AIOs are present, they appear to have a significant impact on click through rates:
Paid CTR dropped by 12 percentage points (from 21.3% to 9.9%) when an AIO was present.
Organic CTR declined dramatically, by ~70%, from 2.94% to 0.84%, despite organic rankings remaining stable (average position 5.9 vs. 5.6).
Overall, all roads point to AIOs reducing CTR for both paid and organic media. However, there is a silver lining: being included as a source in an AI Overview significantly boosts both visibility and credibility. For example, SEER reports that websites cited as a source within an AIO saw their organic CTR nearly double, rising from 0.6% to 1.08%.
At Coppett Hill, we wanted to go beyond click-through rates and measure how frequently AIOs actually appear in search results for Google (by far the dominant search engine among our clients). Using the same mix of companies as in our AI referral analysis, we tracked the share of advertiser organic traffic exposed to these overviews.
Coppett Hill AIO Tracker
Our tracker shows that the presence of AIOs when weighted by advertiser traffic (i.e. on the search terms that actually matter to advertisers) has been stable in recent months. AIOs are more likely to be found at true ‘top of funnel’ / research keywords. This highlights their growing role in driving ‘no-click’ user journeys, where searchers find answers directly in AI-generated responses without visiting a website.
Does This Mean the End of Organic Search Traffic?
Not yet. The impact of AIOs is currently limited in scope, appearing for just 12% of keywords on Google as of February 2025 (when weighted by traffic).
However, AI Overviews are driving an increasing number of no-click journeys, where users find answers directly within search results without visiting a website. As search behaviour shifts, featuring within an AIO is one of the few ways to mitigate the impact.
Being featured as a source in an AI Overview can increase visibility and organic CTR. This sentiment was echoed by Liz Reid, head of Google Search, who said: “The links included in AI Overviews get more clicks than if the page had appeared in a traditional web listing for that query.” This creates an opportunity for brands to adapt and optimise their content for AIOs rather than viewing them solely as a threat.
So, What Should You Do?
For most advertisers, there’s no immediate need to overhaul your search marketing strategy—yet. However, AI’s growing role in search means businesses should be proactive in understanding and adapting to these changes.
What Should Businesses Be Doing?
Stay ahead by testing and adapting – Google remains the largest AI-powered marketing channel, and businesses need to treat it as an evolving space. Testing how AI-generated results affect search intent, rankings, and click-through rates is essential to understanding what drives visibility in this new landscape.
Monitor LLMs’ role in your audience’s journey – Businesses need to track how their audience engages with LLMs and ensure their content remains visible in AI training data to maintain future discoverability.
At the same time, it’s crucial to continue monitoring the situation. Stay informed about the evolution of AIOs within your industry and be ready to adjust your strategy as necessary.
The Final Word
AI Overviews and LLMs are reshaping the search marketing landscape, but they don’t signal the end of search engines. Google’s dominance in search remains strong, and for most industries, traditional search marketing continues to yield strong results. However, given the rate of change, we expect 2025 to be a year of significant transformation. We’ll be publishing our AI Index monthly to help you keep on top of the latest trends.
If you’d like to see how your business benchmarks within our AI trackers or discuss your company’s approach to AI disruption in search marketing, please contact us.
All views expressed in this post are the author's own and should not be relied upon for any reason. Clearly.