Paid Media
Update by Casey Anderson, Director of Paid Media
The major paid media platforms are racing to win on automated, AI-driven performance — Meta via its creative and bidding stack, Google with Demand Gen and new recommendation transparency. For advertisers, the playbook is: keep diversifying channels, feed the algorithms clean first-party data, and verify that automation is actually moving the needle. Here’s what’s new this month.
Google Marketing Live Takeaways
Each year at Google Marketing Live, Google unveils advertising platform updates that give us new features and insights into the direction they’re taking the platform. Here’s the biggest news affecting Google Ads from the May 20, 2026 event.
1. There is no escaping AI Mode. It is the new Search and therefore, the new search advertising. We’ve been testing AI Max for months in client search campaigns, which is said to give better ad ranking and better likelihood of ads showing up in AI Mode and AI Overviews.
Has it helped? The verdict is still out at this point, but there are a few pieces to this puzzle that have been filled in with Google Marketing Live announcements.
2. New ad and campaign types will fully leverage AI Mode and contextually align with what users are searching. Next gen ads like AI powered Shopping ads, Business Agent for Leads, Direct Offers and Conversational Discovery Ads are going to make the user experience smarter and more engaging.
3. “The best ads are just answers.” Seems obvious, but potentially difficult to execute without the help of AI Max.
With this knowledge, your target audience, the problems you solve, and where you want to show up, is a set of guardrails for how Google’s AI expands your targeting. In practice, this means that you get actual controls rather than hoping Google doesn’t send irrelevant traffic to your site.
AI Brief helps with this control. You can tell Google in plain language what your ads should and shouldn’t say. It’s a step in the right direction for B2B advertisers or any advertiser who wants to protect their brand. We think it’s worth testing.
4. Your feed should not be ignored. Feed health and site content are front-line performance levers. Gemini is going to summarize why your product is the right match for searchers. That summary pulls from your feed and your site. Lacking in the content department means you won’t show up.
5. Asset Studio can pull together full campaigns from a written brief. Have a creative production bottleneck? Leave it to Asset Studio to flesh out a more robust creative bank. Of course, Google’s generative creative has been somewhat questionable in the past, so before jumping into this, we (as always) recommend a testing phase.
Getting your ads to actually show up in AI Overviews is still going to require human optimizations. Sticking to the basics of what we’ve come to learn over time with search ads will remain the bread-and-butter of performance. Diverse content, clear and informative messaging, strong audience signals, and consistent monitoring will never go out of style.
SEO & AI
Update by Brett Woodward, Sr. SEO
Google Publishes Guide to Optimizing for Generative AI
Google has finally published an official set of considerations for AI search.
The guide includes comments on several topics:
- whether or not SEO is still relevant (it is)
- how to create valuable citation-worthy content
- advice for keeping content accessible to crawlers
- Google’s take on popular GEO tactics (in a section called “mythbusting”)
- tips for optimizing for AI agents
Besides doubling down on their “people-first” philosophy, Google’s guidelines also include some helpful info on how large language models (LLMs) go about generating answers and finding citations to back them up.
While it was probably coming either way, I’m viewing this move as a response to the large wave of GEO experts coming out of the woodwork who are building businesses around optimizing for the new era of search and cultivating a sense of fear to support it.
To be clear, we very much are in a new era of search. The landscape has changed for good; AI is synthesizing the web’s information in ways search engines haven’t done before and that’s led to a host of new user experiences and behaviors.
What bothers me is that the narrative I see being spun often seems to hinge on the idea that GEO is all about technical tactics aimed at the AIs themselves instead of improving the experience for the people we’re trying to reach.
I tell clients repeatedly that technical tactics can shift quickly as the tools and tech evolves, but building a great experience for your audience is solid ground you can always count on.
We absolutely need to understand how these technologies function and we absolutely need to ensure our content can be found, understood, and shared by those technologies. But I firmly believe that businesses should focus on creating a brand or product or experience worth mentioning before embarking on strategies to gain more mentions.
AI Citation Study Finds Accessibility & Search Rank Among Most Influential Factors
We marketers can and certainly should have an impact on helping businesses create a brand or product or experience worth mentioning. But at the end of the day SEOs are judged on their ability to earn:
1) visibility on the tools where people search, and
2) meaningful engagement stemming from that visibility.
So let’s talk about factors that influence AI visibility. One of the more interesting reports on AI citations I’ve seen lately came from well known SEO Cyrus Shepard. In his thorough meta-analysis, he reviewed over 50 studies on AI search citations done by leading marketers and connected the dots to show us what factors correlate most often with earning citations.
He found that the strongest indicators for whether content will be cited by AI comes down to accessibility and search rank. Specifically, if the content can be found and understood by AI, and how well that content ranks on traditional search engines.
Obviously the AIs need to be able to view the content before they can digest it and recommend it to users, and keeping content accessible to those systems should be paramount.
Regular search engine rankings are also critical. LLM outputs that didn’t come directly from their training data typically seek reputable, high-value content that they extract from search engine results.
Other strong indicators included how well the content matched the user’s query (key words and phrases), and how well the content’s format matched the user’s intent (the example he gives is listicles rank for “best of” searches).
All of the above is core to how the best SEOs have been advising for years. That means a good SEO strategy that makes content available to AI crawlers, rankable on search engines, and impactful to users will go a long way towards good AI visibility.
