When AI handles the tactical work, the advantage lies in your data: what’s new at Google in 2026

At Google Marketing Live in May 2026, Google is unequivocal: “the only way to win in the age of AI is with AI”. Gemini has gone from being just another feature to becoming the system that underpins the company’s entire advertising, commercial and measurement architecture.

At LaMagnética, we’ve been watching this change unfold in our clients’ accounts for months. This post is a roadmap of what’s happening — and what you should act on now.

No-click search is no longer a threat but has become the playing field

At Google I/O in May 2026, Sundar Pichai announced that AI Mode had surpassed one billion monthly users in just twelve months since its launch, and that AI Overviews had already reached 2.5 billion. It is the fastest-growing search system in Google’s history.

Behaviour within this environment bears no resemblance to that of traditional search. Queries in AI Mode are, on average, three times longer than traditional searches: people describe what they need rather than typing in individual words. And ‘brainstorming’-style searches are growing much faster than general queries, suggesting discovery much earlier in the funnel.

The consequence is the famous “zero-click”. According to Seer Interactive’s study of 25 million impressions, 93% of queries in AI Mode do not generate any clicks to an external site. A randomised field experiment (Agarwal and Sen, published on SSRN in April 2026) found that, when an AI Overview appears, organic clicks fall by 38% and the zero-click rate rises from 54% to 72%.

Here lies the paradox that interests us most: traffic falls, but sales need not necessarily follow suit. Appearing within the AI summary acts as a highly selective filter. When a user does click, it is usually on a brand search or direct navigation: the decision has already been made. The metric that matters is no longer ranking but rather the share of mentions within AI responses. Seer’s data suggests that brands mentioned in AI Overviews capture a significantly higher paid CTR on the same queries.

2026 Benchmarks: automation has stabilised costs

For the first time in five years, WordStream’s Google Ads benchmarks for 2026 show an overall reduction in cost per lead, driven by AI Max and Performance Max. Some key figures:

AI Max and AI Brief: from manual control to guideline-based control

AI Max has officially exited beta and become the central automation engine for Search, Shopping and Travel. The promise, according to Google’s data: AI Max campaigns deliver on average 27% more conversions than manual campaigns, and combining it with PMax delivers 15% more volume whilst maintaining a similar ROAS.

One detail worth noting in your calendar: Google is phasing out Dynamic Search Ads in September 2026, and the official migration path is AI Max (DEMG). Accounts that make the switch early — by defining clear conversion windows, listing negatives in advance and separating product categories — are seeing CPA improvements of 30–40% in the first six weeks.

These engines are managed via AI Brief, where you describe your brand’s guidelines in natural language across three key areas:

• Messages: what the text generator should say or avoid (for example, omitting prices or promotions in regulated sectors).
• Alignment: the semantic boundaries that guide the algorithm towards priority intents.
• Audience: the key profiles so that the system can adapt the value proposition and tone.

For retail, AI Max for Shopping dynamically transforms Merchant Center feeds into ads that respond to complex conversational searches, capturing long-tail queries that traditional Shopping overlooked.

YouTube, creators and agent-based lead generation

Google has integrated Gemini’s conversational capabilities with YouTube and Maps. Demand Gen campaigns are already being served on the local layer of Google Maps, and now comes Affiliate Partnerships Boost: a pilot scheme that allows organic affiliate videos from the YouTube Shopping programme to be promoted as paid ads.

The case of Jimmy John’s with creator Adam W illustrates why this matters: by giving the influencer creative freedom (“brand guidelines rather than rigid scripts”), the YouTube Short achieved a CPV 62% more efficient, a completion rate 3.7 times higher than the benchmark, and a CTR 158% higher than the brand’s conventional ads.

In lead generation, the Business Agent for Leads replaces static forms with a Gemini-powered chat that conducts three to five rounds of conversation to qualify the user — much like a sales representative’s screening call would. The interesting part: when transferring the lead to the CRM, it attaches the full conversation and a Gemini summary detailing identified pain points, intentions and objections.

Agent-based commerce: the Universal Cart and who is the ‘merchant of record’

This is probably the most structural initiative. Google is building the infrastructure for agent-based commerce on three pillars (Search Engine Land): the Universal Commerce Protocol (UCP), the Agent Payments Protocol (AP2) and the Universal Cart.

The Universal Cart functions as an aggregator that transcends the boundaries of individual shops: the user adds products whilst browsing the web, chatting with Gemini, watching videos on YouTube or reading Gmail. In the background, the cart monitors price drops, reviews purchase history and even assesses technical compatibility between items from different suppliers. When the purchase is made, the UCP processes it via Google Pay in a couple of clicks, with BNPL options integrated through Affirm and Klarna.

The key point for brands: the retailer always remains the merchant of record. Brands retain ownership of transactional data, billing, margins and the after-sales relationship.

The rollout begins in the US during the summer of 2026 via Google Search and the Gemini app, with a gradual expansion to YouTube and Gmail, and to Canada, Australia and the UK before the end of the year. It is already natively compatible with brands such as Nike, Sephora, Target, Walmart and Wayfair, as well as shops on Shopify.

Ask Advisor and the new creative approach with Gemini Omni

Ask Advisor is the cross-platform agent that connects Google Ads, Analytics, Merchant Center and Marketing Platform within a single conversation. With a simple instruction, it can retrieve the product from Merchant Center, set up the ad group in Ads and identify the best landing pages using historical data from Analytics. It is currently in beta for English-language accounts.

On the creative side, Asset Studio integrates Gemini Omni to generate text, images and video by assimilating the brand’s style guide and brief, with one-click A/B testing. But the strategic tip we like best is the distinction between:

• Creative management (translating copy, adapting formats, cropping videos for Shorts): perfect for delegating to automation.
• Creative conceptualisation (emotional tone, brand narrative): this must remain in human hands. It is the best defence against growing consumer scepticism towards generic algorithm-generated content.

Scientific measurement: Meridian within GA360 and QFCs

The most significant structural announcement for analytics, according to Adswerve: Meridian, Google’s open-source Marketing Mix Modelling framework, integrates directly into Google Analytics 360. It brings a scientific methodology for measuring ROI — which previously required multi-million-pound investments and external consultancy — into the very same environment where raw behavioural data resides.

For complex journeys and zero-click scenarios, Qualified Future Conversions (QFCs) are here: a Gemini-based predictive indicator that analyses early-stage behaviour (recurring brand searches, navigation depth, return visits) to forecast a lead’s future value and integrate it with Smart Bidding. Critical for B2B and long consideration cycles.

Two technical notes:

• To avoid losing data due to cookie blocking, you must adopt Google Tag Gateway (server-side measurement) and consolidate the Data Manager API as the official ingestion channel, replacing the old Measurement Protocol.
• Google Ads’ global data retention policy reduces the retention period for granular historical statistics to 37 months from June 2026. Anyone requiring long-term year-on-year analysis should start structuring their own external databases now.

Action plan: four moves for this season

1. SEO no longer seeks clicks; it feeds the algorithm. Structured content, conversational FAQs, expert opinions and authentic stories teach Google’s AI what your brand is and why it deserves to be recommended. The new metric is citation, not ranking.
2. Audit and enrich your product feeds. Titles saturated with artificial keywords are counterproductive when AI writes the product summary. Add comprehensive conversational attributes: dimensions, use cases, compatibility, materials. A complete feed determines whether you are suggested or excluded in the Universal Cart.
3. Migrate your measurement infrastructure. Google Tag Gateway + Data Manager API, with real business metrics (margin per SKU, LTV) via Product Value Adjustments, so that Smart Bidding optimises profitability rather than empty volume.
4. Move from managing keywords to defining guidelines. The specialist’s value lies in structuring AI Briefs effectively and validating incremental impact with Meridian in GA360.

The underlying idea, repeated throughout the event’s presentations: when AI handles tactical execution, the advantage shifts from optimisation to first-hand data, the creative strength of the brief, and robust long-term measurement. The machine does the rest. Our job is to decide where to steer it.

Would you like us to review together how to prepare your marketing for the new AI ecosystem? Get in touch.