Google expands Gemini-built ads across AI Mode and Search

Slogan.website Research Desk·

Google said on May 20, 2026 that it is testing new ad formats built with Gemini for AI Mode and bringing additional AI-assisted ad experiences to core Search in the coming months. For marketing teams, the practical takeaway is simple: Google is moving sponsored discovery closer to conversational search, product guidance, and lead qualification.

This matters because the update spans multiple surfaces at once. In its official announcement, Google introduced Conversational Discovery ads and Highlighted Answers for AI Mode, plus AI-powered Shopping ads and Business Agent for Leads for Search. Google also said these formats will remain labeled as sponsored, and that brands should strengthen their foundation with AI Max for Search, AI Overviews-eligible campaign types, and Performance Max where relevant.

What changed

Google's official product post breaks the launch into four main pieces:

| Format or feature | Surface | What Google said on May 20, 2026 | Why it changes workflow | | --- | --- | --- | --- | | Conversational Discovery ads | AI Mode | Ads can answer a specific user question with tailored creative and an AI explainer. | Creative has to map to longer, situational prompts, not just short queries. | | Highlighted Answers | AI Mode | Eligible ads may appear inside AI-generated recommendation lists. | Brands may win visibility earlier in research journeys. | | AI-powered Shopping ads | Search | Gemini can surface relevant products and write a custom explainer for the shopper. | Feed quality, product data, and asset coverage matter more. | | Business Agent for Leads | Search | A user can click into a chat-based lead interaction powered by website information. | Landing pages and site content now affect pre-lead conversations more directly. |

Google also said ads in AI Overviews are already eligible from existing Search, Shopping, Performance Max, and App campaigns in supported contexts, but there is still no direct placement targeting or opt-out specifically for AI Overviews. The new May 20 update extends a broader trend in which AI-generated search experiences are becoming another layer of ad inventory and user intent interpretation.

Why it matters

For marketers, the biggest shift is that ad relevance is being judged against both the query and the AI-generated context around that query. Google's AI Overviews help page says ads can be matched to the user's request and the overview content, which expands the number of research moments where a brand can appear even when the original query is not a standard bottom-funnel search.

That creates upside and pressure at the same time.

The upside is reach into more exploratory behavior. A buyer comparing options, researching a problem, or narrowing a shortlist may now encounter sponsored suggestions inside AI-assisted journeys instead of only on a standard results page. For teams working on search, retail media, lead generation, and growth, that can open new assisted-conversion paths.

The pressure is operational. These formats depend on clean product data, clear landing pages, strong brand positioning, and trustworthy site content that Google's systems can interpret. If your assets, feed attributes, or conversion setup are weak, AI-assisted surfaces will magnify the weakness rather than hide it.

This is also where search and GEO work start to overlap. If AI Mode, AI Overviews, and Search ads increasingly depend on structured context, marketers need better source pages, better brand mention tracking, and better internal knowledge of where the site answers real buyer questions. Slogan.website's GEO Visibility Checklist, Generative Engine Optimization guide, and brand mentions tracking article all become more relevant under this model.

Who is affected

The teams most affected are:

  • Search and PPC managers running Google Ads at scale.
  • Ecommerce and retail teams relying on Shopping feeds and product detail pages.
  • Lead generation teams in education, B2B software, local services, and other considered-purchase categories.
  • Marketing operations teams responsible for conversion tracking, feed governance, and landing-page consistency.
  • SEO and GEO teams that need to understand how commercial visibility is shifting inside AI-assisted search experiences.

Smaller advertisers should pay attention too. Google's own guidance for AI Max for Search points marketers toward text customization, search term matching controls, and landing-page-derived asset generation. That suggests more campaign performance may depend on the quality of your site and first-party inputs, not only on manual keyword curation.

What to do next

Use this short workflow before these new formats roll further into market:

  1. Audit existing Search, Shopping, and Performance Max campaigns for conversion hygiene, asset completeness, and current landing-page quality.
  2. Review whether your most important pages answer high-intent buyer questions clearly enough to support both organic visibility and AI-assisted ad interpretation.
  3. If you run ecommerce, check product titles, descriptions, pricing, availability, and image coverage before relying on AI-powered Shopping placements.
  4. If you generate leads, map which site pages would feed a chat-style pre-lead experience and remove vague or outdated claims.
  5. Build a reporting baseline now with spend, CTR, conversion rate, CPA, and assisted revenue so you can spot whether AI-assisted placements improve incremental performance later.
  6. Pair paid-search planning with internal GEO work using the GEO Visibility Checklist, then pressure-test budgets in the Marketing Budget Planner and business impact in the Marketing ROI Calculator.

A simple rule is useful here: if Google's AI has to explain your offer, your source pages need to be unusually clear.

What remains uncertain

Several important points are still unresolved.

First, Google described multiple formats as tests, pilots, or "coming soon" features in its May 20, 2026 announcement. That means timing, market availability, vertical eligibility, reporting detail, and auction behavior may still change.

Second, Google has not yet offered granular reporting for every AI-assisted placement. Its AI Overviews ads documentation says segmented reporting for ads shown within AI Overviews is not currently available. Marketers should expect measurement blind spots until Google publishes more detail.

Third, it is still unclear how performance will vary across geographies and business types. The announcement reflects Google's global product direction, but rollout depth can differ by country, language, campaign type, and commercial sensitivity. That matters for the high-value markets Slogan.website tracks, especially the United States, Canada, the United Kingdom, Australia, and major European markets.

The near-term conclusion is not to rebuild your account around a beta headline. It is to prepare the inputs that these systems depend on: structured offers, strong landing pages, trustworthy claims, clean feeds, and measurement you can defend. The brands that do that work early will be in a better position if conversational search becomes a larger share of paid discovery in 2026.