Google Marketing Live 2026: what it means for PPC

Google Marketing Live 2026 was not a product showcase with a few interesting slides. It was a structural announcement: Gemini is now embedded in every layer of Google's ad system, AI Mode has crossed one billion users, and Universal Commerce Protocol changes where and how conversions happen. For PPC managers and agencies, that means rethinking campaign structure, data strategy, and creative workflow at the same time.

Key takeaways

• AI Mode has over one billion monthly active users and the average query is 3x longer than a classic search.
• Planning queries in AI Mode are up 80% over six months, meaning the purchase funnel shifts before a user ever clicks an ad.
• First-party data connected through Google Data Manager drives an average 11% ROAS lift.
• Universal Commerce Protocol (UCP) enables purchase directly inside an ad, without leaving to a website.
• AI Brief gives advertisers a structured way to govern Performance Max automation.

AI Mode is reshaping how people shop

The usage figures Google released for the first time after a year of AI Mode operation are hard to ignore. Over one billion monthly active users, queries running three times longer than traditional search, and follow-up queries growing 40% month over month. Planning queries (where to buy, how to compare, what to pack) are up 80% over the past six months, according to Search Engine Journal.

In practice, users start on traditional search and move into AI Mode for deeper exploration. Electronics, fashion, health, automotive: categories where a customer needs to be convinced, not just pointed somewhere. An ad that appears at the right moment in AI Mode reaches a user at a different decision stage than a classic text ad ever could.

The most common opening words in AI Mode are what, how, is, can, and notably I. People treat AI Mode like a personal advisor, not a search engine. That changes what good creative looks like and what product data needs to say.

Gemini inside the campaign: from creative to checkout

Google showed at Marketing Live how Gemini now touches every step of the ad process. On the creative side, Asset Studio arrives with Adobe and Canva integration and 1 Click A/B Testing for continuous asset optimisation. Gemini Omni generates variations directly inside the environment where campaigns run.

More significant is the shift in what ads actually say. AI-powered Shopping ads generate personalised product explainers in real time, directly inside the ad unit. A single static creative with one message for everyone will gradually lose ground because Gemini synthesises product data into answers tailored to the specific query, as Google Ads & Commerce Blog describes.

Conversational Discovery ads and Highlighted Answers are new formats appearing directly in AI Mode results, clearly labelled as ads. For advertisers, this is the first direct entry point into conversational search, a surface where ads simply did not exist before.

Universal Commerce Protocol: conversion where the customer already is

UCP is probably the biggest structural change for e-commerce advertisers. Purchase directly inside Search, Gemini, or Google Maps without a redirect to a website. Universal Cart unifies the basket across platforms and the retailer stays the merchant of record, as confirmed by Google Ads & Commerce Blog.

The launch partners are seven major brands (Nike, Sephora, Target, Ulta Beauty, Walmart, Wayfair, Shopify), but the platform is opening up. Google Pay is adding BNPL options through Affirm and Klarna and expanding into hotel booking and food delivery.

For agencies and in-house teams, the most immediately useful new tool is AI performance insights in Merchant Center, which tracks share of voice against competing brands directly in AI search surfaces. Currently rolling out in AU, CA, IN, NZ, and US, with global expansion to follow. Teams that build habits around this data early will have a real optimisation edge.

First-party data and AI Brief: where strategic control lives

With this much automation, the natural question is where the advertiser's strategic judgment actually matters. Two places stand out.

First, first-party data. Advertisers connecting their own customer data through Google Data Manager see an average 11% ROAS improvement. Dr. Martens reported a 16% gain using Performance Max with their own data, according to PPC Hero. The quality of data going into the system directly determines the quality of what comes out. This is where human work has the highest leverage.

Second, AI Brief. The new tool lets advertisers input goals, audience personas, and brand guidelines as a structured brief for Performance Max and AI Max for Shopping. The campaign then operates within defined parameters rather than running loose. For agencies managing multiple clients with distinct positioning, this is a practical way to maintain consistency without manually configuring every setting.

Demand Gen campaigns are expanding to Google Maps, tablets, and YouTube Pause ads, giving brick-and-mortar businesses a real-time high-intent capture tool. Launching Demand Gen with product feeds is one of the fastest ways to test the reach of the new formats.

What to do first

The shift to an AI-first ad environment is not a future concern. It is a question of what to do this month.

Concrete steps in order of impact:

• Connect first-party data through Google Data Manager if you have not already. Without your own data, you are asking Gemini to work with weaker signals.
• Audit your product feeds and add conversational attributes to product descriptions. AI search draws on these when generating explainers.
• Set up AI Brief for existing Performance Max campaigns and define brand guidelines as structured input.
• Launch or expand Demand Gen with product feeds and track performance on the new placements (Maps, tablets, YouTube Pause).
• Monitor share of voice in Merchant Center as soon as AI performance insights becomes available in your market.

If you want to coordinate these steps across campaigns and channels without adding manual overhead, Charlie's AI agents are built for exactly this kind of work. You can see how it fits together on the product page.