Google Marketing Live has always been part product showcase, part signal about where the company sees advertising heading. The 2026 edition was no different, but the scale of what was announced – and the consistency of the thread running through it – made it worth paying close attention to. AI is no longer a feature sitting alongside Google Ads. It’s increasingly the infrastructure on which the platform is built on.
That shift carries real implications for how businesses use Google Ads. It doesn’t mean adopting everything Google announced. It means understanding what’s changed, working out which parts of it are relevant to your situation, and making deliberate choices about where to apply new tools rather than letting the platform’s defaults make those choices for you.
Search is behaving differently
The way people use Google Search has changed. AI Overviews – Google’s AI-generated summaries that appear above traditional results – now reach 2.5 billion users monthly. AI mode queries, where users interact with a more conversational version of Search, are running at three times the length of a traditional search. People are asking more specific questions, in more natural language, with more context behind them.
For advertisers, this creates a different kind of intent signal. A longer, more specific query tells you more about what a user actually wants. Google’s AI Max feature, which uses these richer signals to match ads to searches that wouldn’t have triggered them under traditional keyword matching, is reportedly driving 27% more conversions for campaigns using it. That’s a big number, though results will vary considerably depending on account structure, category, and how tightly campaigns are already managed.
The new ad formats Google is testing in AI mode – interactive ads that allow users to ask follow-up questions, direct offers matched to individual shoppers, and AI-generated product descriptions in natural language – continue the direction of travel. Search is increasingly a discovery surface, not just a place where people go to click on a known result.
The Universal Commerce Protocol signals where shopping is heading
Of everything announced at GML 2026, the Universal Commerce Protocol (UCP) is probably the piece that tells you the most about Google’s longer-term thinking. It’s a proposed industry standard that creates a shared language for AI agents and merchant systems – so that an AI agent browsing on behalf of a user can interact directly with a retailer’s live inventory, pricing, and loyalty data.
This is AI-to-AI commerce. Rather than a human conducting a search, clicking through to a site, browsing, and checking out, the future UCP imagines is one where an AI agent handles that entire process on behalf of the user – with the merchant’s own systems communicating back in real time. Google’s Universal Cart and Native Checkout features are part of a one-click checkout experience that removes friction from the moment intent is established.
None of this is fully live yet. The Universal Cart is rolling out in the US in summer 2026. But the direction is clear, and it matters for any business that sells online. The infrastructure Google is building is designed around AI-mediated transactions, not just AI-assisted search. Businesses that invest in clean, well-structured product data, accurate inventory feeds, and first-party data programmes now are building on foundations that will serve them well as these systems mature.
YouTube’s role in the purchase journey is more significant than many accounts reflect
Google’s figures put YouTube and Google properties together in 82% of purchase journeys, with the combined channel delivering 40% higher return on ad spend than other media and 1.3 times higher ROI than paid social in EMEA. DemandGen campaigns have been upgraded with Gemini to improve creator-led targeting, and creator endorsements are reportedly driving 13 times higher brand search likelihood.
Those are Google’s own numbers, so apply the usual caveats. But the broader point – that YouTube is a serious performance channel and not just a brand awareness play – is one we see reflected in client account data. It tends to be underweighted relative to what it can deliver, particularly for businesses where consideration time is longer, and trust is a meaningful purchase barrier.
If your current media mix is heavily weighted to Search and you’ve written YouTube off as too expensive or too difficult to measure, the 2026 announcements are a reasonable prompt to revisit that.
Measurement is getting more sophisticated – and that matters
Two new metrics were announced. Attributed Branded Searches (ABS) measures the short-term branded search uplift from campaigns – useful for understanding whether your activity is building brand familiarity, not just converting existing demand. Qualified Future Conversions attempts to predict purchases up to six months out, giving a longer view of campaign value than last-click conversion data allows.
These join a broader measurement update that includes Meridian, Google’s marketing mix modelling tool, now integrated into Google Analytics 360, and a unified cross-channel view pulling together data from Google, Pinterest, TikTok, and Snap. Google Tag Gateway adoption is showing a 14% average conversion uplift, and accounts using first-party data well are seeing an 11% incremental ROAS increase on average.
The consistent message across all of this is that measurement quality directly affects what AI features can do for you. The platform’s automation – bidding, audience matching, creative optimisation – relies on conversion data being accurate and comprehensive. If your tracking has gaps, your first-party data is thin, or your attribution model is set up in a way that doesn’t reflect how customers actually buy, the AI features you’re being asked to adopt will be working from a weak foundation.
Ask Advisor consolidates Google’s AI agent offering
Google has consolidated its collection of separate AI agents into a single tool called Ask Advisor. It connects Google Ads, Merchant Centre, and Analytics in one place, responds to natural language queries, retains context from previous conversations, and can generate creative assets and launch campaigns. The pitch is that it levels the playing field between smaller businesses and enterprise advertisers by giving everyone access to the kind of campaign management capability that previously required a large, specialist team.
That’s true to a degree. Ask Advisor will make certain routine tasks faster and more accessible. But the quality of what it can do is still dependent on the quality of the inputs – account structure, conversion data, business context. An AI agent optimising a poorly structured account with poor measurement won’t produce better results than a human doing the same thing. It’ll just do it faster.
The more useful framing for most businesses is that Ask Advisor is a productivity tool. It saves time on tasks that take time. It doesn’t replace the strategic judgment about which tasks are worth doing in the first place.
How to approach this without overextending
The volume of announcements at GML 2026 can make it feel like there’s a long list of things to implement immediately. There isn’t. Most of these features are either still in testing, US-only for now, or only relevant to certain types of accounts and business models.
The more useful approach is to audit where AI features could make a difference to your specific situation, rather than treating every announcement as something that needs to be acted on. AI Max is worth testing if your campaigns are capturing strong conversion data, and your current keyword coverage may be leaving intent unmatched. The commerce features – UCP, Universal Cart, Native Checkout – are worth watching closely if you run an ecommerce operation and want to be positioned well when they roll out more broadly. The measurement updates are worth acting on now, because cleaner data improves everything else.
Applying tools where they’ll make a difference is a more reliable route to competitive advantage than applying all of them at once.
The Last Word
Google’s 2026 announcements confirm that AI is no longer a layer added on top of the ads platform – it’s increasingly the platform itself. For businesses that invest in the right data foundations and apply new features deliberately, that shift creates real opportunity. For those who either ignore it or try to adopt everything at once without a clear rationale, it’s more likely to create noise than results. The businesses that will benefit most are the ones treating AI as a tool to apply with judgment, not a solution to hand things over to.
If you’d like to talk through what any of this means for your Google Ads account specifically, we’re happy to take a look. Get in touch with the Herdl team, and let’s talk strategy!