- According to Google, AI Mode has more than 1 billion monthly active users, AI Overviews more than 2.5 billion. Google is consistently building advertising around these AI answers.
- Google is testing new ad formats such as Conversational Discovery Ads and Highlighted Answers in AI Mode in the US. They are based on AI Max and Performance Max.
- Google will automatically switch Dynamic Search Ads to AI Max by February 2027. Google cites an average of 7% more conversions or conversion value at a similar CPA or ROAS.
- Many new features launch first in the US or in English. However, clean data, strong assets and clear guidelines can already be prepared today.
At Google Marketing Live on May 20, 2026, Google showed how advertising is supposed to work in a search that increasingly ends with an AI answer. For anyone who manages Google Ads daily, many of the announcements were no longer a surprise. But they clearly confirm where things are heading.
In this overview, we put into context the new features that have consequences for advertisers in German-speaking countries, and show what has already changed since May.
Google is consistently building search and advertising around Gemini
Google is investing heavily in its own AI infrastructure, from its own chips to the advertising platform. From Google's perspective, the use of AI search justifies this: At the I/O developer conference in May 2026, Sundar Pichai cited more than 2.5 billion monthly active users for AI Overviews. Around one year after launch, AI Mode has passed the mark of 1 billion monthly active users.
For advertisers, this means: A growing share of searches ends in an AI answer instead of a list of links. Anyone who wants to appear there has to set up their campaigns differently than before.
Ads in AI search become answers to the question
Google's guiding idea for the new formats: If people can ask Google anything, ads themselves have to provide an answer. Instead of fixed text ads, ad copy is created in real time to match the user's question.
Four new formats for AI Mode
- Conversational Discovery Ads: They answer a user's specific question with ad copy that Gemini creates for exactly this search. Google is testing them in AI Mode.
- Highlighted Answers: Relevant ads appear within the recommendation lists of AI Mode. This format is also being tested.
- AI-based Shopping ads: Gemini selects a retailer's matching products and writes a short explanation for each. They are expected to follow in the coming months.
- Sales assistant for leads: Prospects ask their questions directly in the ad, and the AI answers based on the advertiser's website. This format replaces the rigid lead form and is also expected to arrive in the coming months.
In addition, Google is expanding the Direct Offers pilot program, which has been running since January 2026: Retailers store discounts, add-ons or vouchers, and Gemini serves the matching offer at the moment of purchase intent. In Germany, ads currently appear above or below the AI Overviews; the new formats are launching in the US first. What this means for your campaigns is described in our article on ads in AI Mode.
Keywords take a back seat to AI Max and Performance Max
To appear in these formats, Google recommends AI Max for Search and Shopping campaigns as well as Performance Max as the foundation. The selection of individual keywords thus loses importance.
The consequence can be seen with Dynamic Search Ads (DSA): Google is automatically switching them to AI Max. This was originally planned for September 2026. In June, Google extended the deadline for DSA to February 2027, while automatically created assets and the campaign-level broad match setting have been migrated since September. Google cites an average of 7% more conversions or conversion value at a similar CPA or ROAS when all AI Max features are used.
However, there are also reports from practice that AI Max directs specific product searches to general landing pages, which is exactly the opposite of what DSA reliably solved. That is why, before every switch, we check which landing pages AI Max selects.
With AI Brief, you give the automation guardrails
Automation brings reach, but makes brands in sensitive industries cautious. With AI Brief, advertisers describe in natural language what their ads should and should not say, which audiences they want to reach and which brand rules apply.
From this, the AI derives guardrails for the campaigns. This does not replace a team's experience, but it makes the guidelines binding for the automation.
With the Universal Cart, Google becomes a cross-retailer shopping cart
In early 2026, Google introduced the Universal Commerce Protocol (UCP), an open standard for shopping via AI assistants. In addition to Google, the founding members of the technical committee include Shopify, Etsy, Target and Wayfair. In April 2026, Amazon, Meta, Microsoft, Salesforce and Stripe joined.
Building on this, Google announced the Universal Cart at Google Marketing Live:
- a shopping cart that works across different retailers and in services such as Search and Gemini,
- with payment via Google Pay in a few steps or handover of the items to the retailer's website.
Google is testing the first checkout features with retailers such as Nike, Sephora, Target, Walmart and Wayfair. Checkout via UCP is expected to come to Canada and Australia in the coming months, followed by the UK. There is no date for Germany yet. What is behind the protocol is explained in our article on the Universal Commerce Protocol.
Google is aligning YouTube and Demand Gen more strongly with sales
YouTube is meant to shed its image as a pure brand awareness channel. Google is therefore combining Demand Gen more closely with Search and Performance Max and has introduced several new features for this:
- Google Maps: Through new inventory in Google Maps, Demand Gen reaches people who are exploring their surroundings.
- Product feeds: They run on more surfaces and in additional industries, such as automotive. Videos from Merchant Center are to be served dynamically in Demand Gen campaigns in the future.
- Creator videos: Videos from creator partners can be used as ads directly when setting up the campaign.
- Measurement by campaign type: A dedicated attribution model shows what Demand Gen has contributed. This allows a direct comparison with social media campaigns. Uplift experiments can measure how Demand Gen complements Performance Max.
How to align YouTube with measurable results is shown on our page on YouTube advertising.
In Asset Studio, images and videos are created directly in Google Ads
Asset Studio is becoming the central workshop for creatives in Google Ads. From a briefing, brand guidelines, the website and the goals, it generates assets in different formats and themes that can be revised in natural language. In the future, videos are to be created with Gemini Omni, Google's new multimodal model.
The new one-click testing is practical: New assets can be tested directly against each other to find out which ones best achieve your goals. Google announced the features worldwide from summer 2026, initially in English; the video features with Gemini Omni have no fixed date.
Review remains important: If all advertisers create their creatives with the same tools, ads only differ in what does not come from the tool, such as your own photos, real customer voices and a clear offer.
Meridian brings marketing mix modeling to Google Analytics 360
The biggest structural change in measurement: Meridian, Google's open-source model for marketing mix modeling (MMM), is becoming part of Google Analytics 360. Cross-channel analysis, including outside Google, thus moves from the data science department into a tool that marketing teams already use.
In addition, there are two new signals:
- Attributed Branded Searches measure the searches for your brand that a specific ad triggered.
- Qualified Future Conversions use user signals to estimate which conversions a campaign will still bring over a longer period. This helps with long decision journeys for which the usual conversion window is too short.
Both signals evaluate campaigns that do not sell immediately more fairly. The prerequisite remains clean conversion tracking.
Ask Advisor connects Google's AI assistants
Finally, Google has merged its assistants: Ask Advisor is an AI agent that works across Google Ads, Google Analytics and the Google Marketing Platform. An agent for Merchant Center is to follow soon. Ask Advisor gives recommendations and can set up campaigns from a description in natural language.
The tool launches as a beta for English-language accounts. There is no date yet for German-language accounts.
Conclusion: Your edge lies in your strategy again
Google Marketing Live 2026 makes it clear: Google is taking over more and more operational work. Keyword research, asset testing and fine-tuning bids are increasingly handled by AI.
This makes everything Google cannot automate the difference: a clear strategy, complete first-party data and guidelines that keep profit in view. Many of the new formats will only reach Germany later. You can build the foundations for them today: clean tracking, values based on margin, well-maintained feeds and your own assets.



