When we laid out our strategic bet on Google’s enterprise ecosystem over a year ago, we argued that the AI race wouldn't be won by flashy consumer chatbots. it'd be won on infrastructure, integration, and data.
The shift towards practical, working AI agents announced at Google I/O is the ultimate validation of our strategy.
The death of traditional keyword mechanics is actually the best thing to happen to brand strategy in a decade.
Here's why the death of traditional keyword mechanics is actually the best thing to happen to brand strategy in a decade.
From Media Mechanics to Brand Architecture
For years, digital marketing's been bogged down by the minutiae of bidding mechanics. Match types, bid adjustments, and campaign-level exclusions are gone. AI Max for Search essentially automates these away.
But Google has replaced those manual levers with something far more strategic: AI Brief.
Instead of writing endless ad copy permutations, marketers now use natural language to programme Google's AI. They define an overarching brand proposition, specific audience guidelines, and strict tone of voice rules.
For the first time, there is a direct, causal relationship between your ability to articulate a brand's core truth and the performance the AI can deliver.
In 2026, the primary differentiator isn't platform mechanics, it's brand architecture. If you can't define your brand proposition with absolute precision, the machine has nothing to optimise.
The Rise of Information Retrieval Engineering
The traditional sequence of "Crawl, Index, Rank" is gone.
It's being replaced by "Query Fan-out, Sub-query Retrieval, and Answer Generation" as search becomes a conversational experience.
With Gemini's automated shopping features and the persistent Universal Shopping Cart, AI agents are increasingly researching, comparing, and checking out on behalf of users. To be cited in Google's AI Overviews or recommended by these active shopping agents, your technical setup must evolve.
To be cited in Google's AI Overviews or recommended by these active shopping agents, your technical setup must evolve. At Jaywing, our LLM Visibility Study analysed over 200,000 unique prompts.
It proved that LLMs look for ‘evidence pages’ grounded in structured data. Our research showed that semantic, highly structured content structures are cited 11.4% more than standard pages.
We aren't optimising for a static search engine results page anymore. We're engineering content to be retrieved by autonomous machines.
Immaculate First-Party Data is the Steering Wheel
An AI agent's only as good as the data you feed it. At GML, Google introduced powerful new routing and data steering tools. These include server-side Google Tag Gateway (GTG) with Confidential Matching and Product Value Adjustments (PVAs) inside Merchant Centre.
With ongoing signal loss, having a pristine first-party data pipeline in Google Cloud and BigQuery is no longer a luxury. Features like PVAs allow us to instantly feed high-margin, high-repeat, or overstocked product signals directly to Google’s bidding engine.
The agencies still trying to manage campaigns through manual spreadsheet uploads are going to get left behind. The agencies building robust, real-time data pipelines to steer Google's algorithms will win.
The Jaywing View
The transition to AI Max and agentic commerce isn't something to fear. It's a massive opportunity to elevate our industry.
By automating the mechanical work, Google's cleared the deck. The winners of this new era won't be the ones holding onto the manual levers of 2015. It'll be the brands and agencies who master the first-party data pipelines to feed the AI, and the creative brand strategy to guide it.
We built Jaywing to sit precisely at that intersection of data science and creative brand strategy. The game hasn't changed. It's just finally being played on our turf.