Google was never behind the curve, and now it is winning. Every new AI product triggers the same reaction, be it a new large language model (LLM) feature or an AI browser like Perplexity’s Comet, writes János Moldvay, Chief Data Science Officer at Funnel

Google was never behind the curve, and now it is winning. Every new AI product triggers the same reaction, be it a new large language model (LLM) feature or an AI browser like Perplexity’s Comet, and the industry declares that ‘Google Search is dying’ or ‘Google is not what it used to be’. Like plants drawn to the sun, marketers and technologists alike often chase the newest and shiniest thing. 

But the data tells another story. Research from Funnel last year found that 64% of marketers expect their customers to use traditional search engines less in the coming two to three years. This may sound alarming, but it’s quite the opposite.

Over the past year, Google has emerged as the biggest winner of the Search and AI race, even as it shifts focus from SEO to AEO. What looks like fragmentation of Search is, in fact, Google’s deliberate reinvention. Google has also always been an AI company, whether the industry chose to see it or not. Marketers would do well to remember Google is the top dog of adtech for a reason.

Search is a destination, not a navigation service

Zero-click search and proliferating AI overviews make it look like Google is cannibalising its own model, yet this is a deliberate transition. Google used to be more of an intermediary for web traffic, connecting users with the products or information they needed. Now it is an end destination, and advertisers are being drawn further into using it.

Where before marketers would bid for keywords, this is giving way to intent bidding, where marketers attempt to predict consumer choices. Google Performance Max and AI-driven campaign management have shifted advertisers away from buying specific placements because Google’s AI reads intent better than any human planner. Brands over-indexing on lower-funnel performance marketing will come unstuck as keywords are used less and less to determine consumer choices. 

This comes with obvious drawbacks, notably that Performance Max or Google Shopping are black boxes, making ‘intent’ difficult for marketers to judge and tricky to price. Within the black boxes, Google has more metrics to draw upon than its competitors to determine intent through experimentation and to roll back what doesn’t work, making it attractive to advertisers. This omnichannel approach, where Search, Shopping, Display and YouTube campaigns are brought under one roof for advertiser budgets and bidding, and are all optimised by Gemini, puts Google out ahead of OpenAI’s Ads Manager or Microsoft Advertising.

Google’s user and data advantage

Google Search has always been an AI product, whether we’ve recognised it as such or not. Self-proclaimed “AI experts” often forget that Google employees authored the seminal 2017 paper “Attention Is All You Need,” introducing the Transformer architecture, which laid the foundation for modern LLMs. 

The Page-Rank algorithm within Google Search is probably one of the most successful AI tools or AI products out there, certainly by number of user interactions. Ideas that Google has been scrambling to catch up with buzzy LLMs overlook the underlying head start it has both in terms of technology and user base. 

Let’s consider search queries, even now: several hundred billion every month, compared to the several hundred million you see with Perplexity, plus Gemini’s web-traffic share has grown from 5.4% in January 2025 to approximately 22.5% by December 2025.

Over the course of a year, search queries have become increasingly AI-driven, learning from every previous query, with Google’s AI Overviews now reaching 2 billion monthly users. Think of how many of us are locked into using Google Chrome or Gmail, how many user graphs Google has access to. Even if OpenAI were to mount a serious challenge, Google is already streets ahead in rich, real-time data on how users interact with marketing content. 

Ultimately, the success of AI is always about the quality of data. LLMs are becoming commodities; modelling is becoming a commodity. So, the success of a company like Google depends on user data accessibility. And they have more access than anyone else right now.

Adaptability is king 



As its lower funnel becomes more automated and less differentiated by brand, advertisers are maintaining - and often increasing - spending to ensure measurement infrastructure keeps pace. Zero-click journeys and agent-led commerce require measurement triangulated across marketing mix modelling, multi-touch attribution and incrementality testing, rather than being reliant on Google Analytics dashboards.

Challengers like OpenAI and Perplexity won one battle, innovating enough to push Google to reinvent its core product. But Google has reinvented itself, and marketers must now understand the new rules it is setting. Bet against it at your peril.

Marketers who write Google off will find themselves misallocating budgets and misreading results at exactly the moment the rules are changing. Those who understand what Google has become, an AI-native, intent-driven commercial infrastructure, will be best placed to compete.