• by Laurie Sullivan  @lauriesullivan, May 23, 2023

Google continues to advance search advertising by expanding its suite of products through automation, generative artificial intelligence (AI) and large language models (LLMs).

Several announcements made Tuesday at Google Marketing Live support the move across Performance Max, Google Ads and Shopping.

Automatically generated custom advertisements being tested in Search and Shopping will roll out in the coming months. The feature will integrate directly into the AI-powered tool that Google calls Snapshot, also based on generative AI.

The additional information serves up when in the Search generative experience, an experiment and early step toward changing the way that people search and discover information and products, said Dan Taylor, vice president of global ads at Google.

“I feel like it’s a very transformational time,” Taylor said. "Not just for Google, but the entire industry.”

If someone searches for “outdoor activities” in Maui, for example, and then narrows the search to ask about activities for kids or surfing, they might see a custom ad for a travel brand promoting surfing lessons in Maui.

“It’s a new experience we’re experimenting with,” Taylor added.

As search evolves, so will advertising. Google will bring AI to campaign, marketer, and merchant workflows with help from large language models and generative AI.

The change forges nuances in automation and campaign setup and creation to improve performance for all channels across Google Performance Max.

Google introduced two experiments in Search Labs that will explore the integration of Search and Shopping ads directly into the AI-powered snapshot and conversational mode. The formats use generative AI to create relevant, high-quality custom ads.

In the conversational experiences for Google Ads, the marketer adds a landing page from the company’s website. The Google bot crawls the page and AI summarizes the content. It then generates relevant and effective keywords, headlines, descriptions, images and other assets for the campaign.

Marketers can review and edit the suggestions. This experiment is now in a global open beta for English, with plans to expand to additional languages this year.

Avoidance of repetitive, automatically created assets over time that are based on the same signals being fed into the system relies on a diverse set of keyword, headlines, and descriptions. The automatically created assets and the generative AI will be grounded in the assets the advertiser provides.

The feature can be considered a natural-language conversational experience that is designed to start campaign creation and simplify the production of search ads.

There’s an unlimited number of responses that marketers can generate, but there’s a fix number that can be inserted into the rotation.

“We are experimenting with limits as part of the beta,” said Brian Burdick, senior director of Search Ads Automation. “On the automatically generated assets, we will generate them on [the marketer’s] behalf and rotate them at run time. Then you’re able to see all those and/or disapprove them.”

Automatically created assets in the campaign level run continually after a campaign is set up to provide even more combinations of ads based on pre-approved headlines, landing page and queries.

The campaigns adapt to changing consumer demand. Google said the early adopters of automatically created assets for headlines and descriptions see 2% more conversions at a similar cost per conversion among ad groups with responsive search ads.

Google is working on providing a way to inform the consumer that AI has generated the created assets. Two possible identifiers are watermarks and metadata. The metadata, in the initial rollout -- especially in images -- will have a distinguishing mark of some sort.

Performance Max will soon give marketers this feature across all channels. Advertisers will have the ability to create professional-grade videos, images, and text creatives that correspond to YouTube, Display and Search. Google will pull in the details from a brand’s website.

About 25% of YouTube video advertisers use a video ad created by Google’s AI. Adding at least one video in an advertisers Performance Max campaign delivered 20% more conversions in YouTube compared with horizontal videos alone.

Google also included ecommerce in today’s announcement. Data suggests that having the correct mix of images drives business.

There is an increase in shopper demand for products when ads have multiple images -- creating as much as a 76% increase in impressions and 32% increase in clicks. But creating images is expensive and time-consuming. Product Studio for merchants helps them create and edit scenes, and improve the quality of lifestyle images for free.

Jeff Harrell, senior director of shopping at Google, said the company is working with a few smaller merchants “struggling to create the correct assets needed to better engage with users. But we see this as a larger opportunity that embraces all.”

Merchants in the U.S. can access Product Studio from Merchant Center Next -- an improved platform that lists products across Google -- starting in the next few months. These features will become available to merchants using the Google and YouTube app on Shopify.