And why Google will fend off OpenAI and Perplexity

Every week seems to be packed with AI news, but the last two weeks have likely surpassed all others in the past year. Here’s a quick rundown before we dive into discussing the future of search in the AI Era, an approximately $200 billion question. But even before that, don’t miss our AI Trailblazers Summit on June 6th in New York. Details after the main story. Now for the AI news.

• OpenAI launches GPT-4o at its Spring Update with the "o" standing for "omni". This version is notable for its real-time interaction capabilities. Users can converse with it, have it assist in solving math problems by observing work in a window, and run voice driven translations in real-time (OpenAI demonstrated Italian to English translations). While some of these features are not new for those in the industry, packaging them together made for a good story.
• Google made a slew of announcements at Google I/O . These included AI Overviews in Search (lots more on that later), the Ask Photos feature, multi-modality capabilities, much larger context windows (essentially memory windows), and sophisticated AI agents capable of performing a variety of jobs. The announcements reminded us of how long Google has been working on AI capabilities and how easily it can make AI advancements reach hundreds of millions of people

• Microsoft launched a Co-Pilot+ personal computer, integrating its AI capabilities into the Windows operating system, at its Build Conference a few days later. It also announced the Recall feature giving users perfect memory of all their historic virtual activity. That certainly isn’t creepy :-) The new computers will arrive on June 18th and will use Qualcomm processors. Microsoft is expecting 50 million AI-powered Windows PCs to be sold in the next year.
• Google came back this week with Google Marketing Live to announce how AI is being integrated into all their ad products and in turn how advertisers will be able to buy sponsored ads against its AI Overviews in search engine result pages. And that’s where it gets exceptionally interesting for the marketers among us as it demonstrates that, indeed, a lot is about to change.
• Nvidia created even more envy with its stock market climb. As if Nvidia stock hasn’t surged enough in the last 18 months, yesterday it climbed another 10% after surpassing its first-quarter earnings expectations. What many don’t realize is that it’s Cuda that gives Nvidia its moat. Listen to this Acquired episode which captured the Nvidia story best.

Unpacking the Search Question

As I shared in my January predictions , search is changing. Companies like Perplexity and OpenAI (with a soon-to-be-announced search offering) are innovating aggressively, but Google isn’t letting them encroach on its massive cash cow. This week's announcements from Google are a significant move to becoming more of a walled garden, with much more to follow. Let's unpack what this means for how people find content on the Internet and its implications for search engine optimization (SEO), paid search, and the publishing industry. Search has always been of particular interest to me, and in my last corporate role, a dedicated 50-person SEO team was part of my marketing function. Here's what to expect.

Google algorithms aggressively deprioritize low-quality content.

This plan has been in motion for a while, but the Google March Core Update was the boldest move to date with more to come. To combat synthetic content, Google has refined its core ranking systems to more aggressively downplay low-quality content designed to boost search engine rankings. The algorithm now does more to prioritize content that meets real human needs, emphasizing expertise, experience, authoritativeness, and trustworthiness. This shift favors real-time content and discussion threads from platforms like Reddit. According to some site owners , this update has already resulted in significantly lower search impressions for many publishers and brands.

Google is also addressing site reputation abuse, where low-quality third-party (often AI-generated) content is used on reputable sites to boost rankings. Additionally, it is tackling expired domain abuses, where a once-respectable domain is purchased by a third party to publish low-quality articles for SEO gains. Why does this all matter? Because as Peter Caputa of Databox said, it signals that Google is intentionally evolving from being a traditional search engine to becoming an answer engine that operates a little bit more like a walled garden. The way it treats other sites and what it values is rapidly changing.

Content libraries get smaller and more focused as a result.

Companies leveraging SEO often maintain extensive content libraries, with each piece contributing to the site's overall reputation and driving traffic. However, these libraries, which traditionally include substantial evergreen content, now face challenges under Google's updated quality criteria. Businesses will find that significant portions of their content will no longer perform well, rendering active management of these libraries less effective. This will lead to them needing to depend more on tools like TheWordsmith.ai , one of our AI Showdown winners at the AI Trailblazers SXSW Dinner.

This shift will prompt a move away from generating large volumes of content towards maintaining a smaller library of higher-quality content. Marketers will need to focus on fewer pages that provide significant human-centric value to readers. These high-value pages will continue to be well-indexed by Google and may even achieve higher rankings. Despite the improved quality, the overall value of SEO to the company is expected to decline due to the reduced number of pages. Additionally, managing this streamlined content will likely require smaller SEO teams.

Google be selective about where it wants to be an answer engine.

For those of us with years of experience in the Travel and Financial Services sectors, we have already observed Google's evolution in this direction. As it has moved towards being an answer engine, Google increasingly provides the content users seek directly on the search results page. For example, typing "Give me the NBA scores from last night" displays the games and scores in the SERP before showing a list of links to sites like ESPN. Similarly, looking up a stock quote yields the answer right in the search results. Historically, this feature was limited to a few categories with structured information such as scores, stocks, and weather.

But in some cases, Google has gone further as in travel where it built out its own travel portal enabling hotels and airlines to list their deals. Travel aggregators like Expedia and Booking.com have been forced to advertise on these pages alongside the hotels and airlines that are their own suppliers. Google's size allows it to be both a provider of traffic to traditional aggregators and a travel aggregator itself. This leads to travel aggregators having to pay twice for traffic—both in the traditional sponsored links slots in the SERP and now on the travel portal pages too. Jeff Hurst, a former Expedia Group executive I worked with, testified during the US v. Google antitrust trial that VRBO’s ad spend on Google increased from $21 million to $290 million over five years with no increase in traffic. Now all of that cost cannot be attributed to the travel portal, but it shows how much more difficult Search may get depending on the moves Google makes.

With the introduction of Google AI Overviews, we can expect Google to expand into many more categories to answer more complex questions, potentially creating more specialized pages like their travel portal. Google has signaled this direction with a new tagline: "Let Google do the Googling for you." The multi-billion dollar question is which industries Google will move into next and how aggressively. While Google might say it will be where they can add the most value to the consumer experience, it may not be that simple.

Google is not stupid, they won’t kill their paid search business.

While Google transforms Search to keep pace with innovations from startups like OpenAI and Perplexity, it will be very careful about protecting its paid search cash cow. For instance, Perplexity announced a month ago that it would soon begin accepting advertising, and OpenAI has yet to even launch its search product, which could be subscription-based. Meanwhile, Google wasted no time in launching AI Overviews and Sponsored Ads that can run within these overviews. They’re smart to develop their product and monetization models simultaneously.

More broadly, Google would not have announced that AI Overviews would be made available to every user on the Internet if it believed it would materially compromise the performance of its paid search business. That would make no sense. Furthermore, Google shared that AI Overviews on their SERP pages are actually increasing engagement and user satisfaction, leading to better quality clicks to destination sites. So while there may be a compromise in quantity, quality may be improving similarly to Google's ongoing organic search strategy. This is good for users, good for Google, and arguably good for advertisers who provide a higher quality, deeper experience when users click through.

There’s a chance that AI Overviews won’t have a meaningful impact on paid search volume at all because they don't reduce the number of clicks on paid results. Gartner, which predicted that Search Engine volume may drop by 25% in 2026, could be right (unlikely though)but, regardless, it doesn’t mean Google’s revenue will shrink.

AI Overviews and Performance Max may work well together.

For those less familiar with Performance Max, it is a Google ad product where decision-making around ad copy and creative, ad placement (search, display, YouTube, etc.), audience targeting, and optimization is entirely handled by Google's machine learning. This reduces the involvement of digital marketers. Performance Max has had a mixed history, initially falling short of Google's claims, but it has recently become more sophisticated.

In a future where more answers to queries are provided directly on the SERP pages, Google may not want the search industry to focus obsessively on how much of their traffic and performance comes from paid search. This approach would relieve some pressure on Google and provide more flexibility, as Google can decide where to place ads without advertisers worrying about traffic predictability being disrupted by new generative AI features in search results. As long as advertisers achieve the performance they need predictably and repeatably (at a price Google finds acceptable), they shouldn't worry too much about how new generative AI features like AI Overviews change the search experience. Google is moving in this direction by automatically placing ads from existing advertising campaigns, not giving the advertisers a choice on whether to participate or not.

What’s most important to note is that Google has many levers to pull. If it isn’t seeing the performance or revenue it expects, it can easily reverse course. As long as it maintains its distribution deals with companies like Apple to be on the iPhone, it won’t lose too much traffic either way.

Writing content, creating ads and experiences for AI agents.

An important point to consider is that, whether we like it or not, thanks to generative AI and innovations from companies outside the traditional search space, search is going to change. While Google will do everything in its power (and it has a lot of power!) to manage this change effectively, change is coming for anyone tied to this industry. These changes may be the most significant yet in Google’s history. Why? Because eventually (and who knows when), marketers will have to write content for every human being’s personal AI agent. SEO will evolve to target both humans and AI agents. AI Overviews will serve both humans and AI agents.

Most interestingly, as OpenAI has recently demonstrated, multi-modal capabilities will fundamentally change how we interact with the world, including how we seek answers to questions. Voice is invariably going to play a much larger role than before. AI Agents passively observing what we’re doing in the physical or virtual world will also play a role too. The search engine results page may eventually disappear; we just don't know when this will happen. And what’s certain in a multi-lap race, Google is most certainly further along than OpenAI, Perplexity or anyone else.

What comes next

Next week, I will keynote ADMERICA, the American Advertising Federation’s annual conference in Salt Lake City, Utah. I plan to slow down after June 6th before ramping up again for Cannes Lions. Let me know if you’d like to meet there!

But even before then, I’ll be co-hosting the?AI Trailblazers Summit ?in partnership with Amazon Web Services on June 6th at the stunning Neuehouse in Madison Square, New York. This meticulously curated forum will delve into the critical questions shaping marketing, product development, innovation, legal frameworks, and venture capital in the AI era.?We’re fortune to have an incredible list of corporate leaders joining us as speakers with many more as attendees. If you’re interested in attending respond to this email requesting an invite.

Speakers include:

• Allison Goldberg, SVP & Managing Partner, Comcast Ventures
• Bob Lord, former SVP of Cognitive Applications at IBM
• Cheryl Guerin, EVP Global Brand Strategy & Innovation, Mastercard
• Dave Lampert, CIO, Cadwalader LLP
• George Matthew, Managing Director, Insight Partners
• Jonathan Halvorson, SVP Consumer Engagement, Mondelez
• JJ Juergensen, MD, Head of Cyber Risk, Barclays
• Karen Levy, CIO, Debevoise & Plimpton
• Manish Goyal, Operating Partner, Berkshire Partners
• Marisa Thalberg, CMO, United Parks
• Matt Turck, Managing Director, FirstMark Capital
• Nihal Mehta, Founding Partner, Eniac Venture
• Patrick Garcia, SVP, Innovation at SRS Distribution Inc.
• Peter Lesser, CIO, Skadden Arps
• Robert Huntsman, Chief Data Scientist, Prudential Financial
• Sara Fischer, Journalist, Axios
• Sophie Kelly, SVP of Tequila Brands at Diageo
• Suzanne Vranica, Journalist, The Wall Street Journal?

You can also review the?agenda ?and?request an invite ?before we reach capacity.?

Where I’ve been

Below are a few photographs from the Adweek Vanguard Summit where I spoke to fellow marketing leaders on AI’s potential to transform or traumatize. It's been many weeks of non-stop travel for speaking engagements, consulting projects and board meetings.

What I’m reading

• Johansson says she is 'shocked, angered' over ChatGPT voice (NPR)
• With spatial intelligence, AI will understand the real world (Ted)
• The promise & the reality of GenAI agents in the enterprise (McKinsey)
• A.I.’s Black Boxes Just Got a Little Less Mysterious (NYT)
• AI is changing management — companies must decide how (FT)
• Administration Announces Ai Principles For Worker Well-Being (DOL)

Recent Savvy AI Articles

• AI Myths vs. Reality: What You Need to Know (May 2024)
• Smart Leadership for AI Adoption (April 2024)
• How serious are you about AI? (March 2024)
• From Moon Landings to Creative Renaissances (February 2024)
• Super Bowl Ads that matter in the AI Era (February 2024)

What I’m writing about this week

I'm in the process of writing my third book, Marketing with AI for Dummies . I’ve finished writing all the chapters and have just the introduction to complete and a round of edits to do. The book is available for pre-order on Amazon .