2025 is the most exciting moment in adtech in more than a decade.

Some folks (mostly outside the advertising and adtech ecosystem, though sometimes within) have predicted that AI will kill or significantly shrink the advertising market. I think this is highly unlikely for reasons I’ll explain. It is much more likely that AI will transform advertising. And that means that things we’ve taken for granted for 5-10 years are no longer obvious.

Five years ago, if you had invested in the Big Tech companies (Google, Meta, Amazon, Microsoft, Apple, almost all of whom have giant ad businesses), you’d have drastically outperformed the S&P 500, and those are still the Big Tech companies today (with the addition of Nvidia). It’s no longer obvious that these companies will be as dominant in five years as they are today. And there is one reason: AI, which could cement Big Tech’s advantages but also transform the markets they dominate and give rise to multi-hundred-billion-dollar competitors. 

The same goes for the various sectors of adtech. Just two or three years ago, if someone had said they were building a challenger to The Trade Desk, almost everyone in the adtech illuminati would’ve laughed. Of course, TTD still has a massive advantage due to some combination of its technology, team, relationships, and brand. But the notion that the way people buy media will fundamentally change — and that, with that change, the companies at the top of the adtech hierarchy will shift — is now far from inconceivable. 

Virtually everyone reading this newsletter should be excited about this. Arguing that AI is a mirage or a passing fad is equivalent to sticking your head in the sand. As Azerion’s Matt Newcomb pointed out to me, Google is estimated to have taken about five years to hit 1 billion queries per month. ChatGPT reportedly took four months. I now use ChatGPT as much as Google or possibly more. I use the former to glean information about virtually every facet of life, but most notably, for particularly complex queries that I wouldn’t expect Google to be able to handle. This represents a tremendous amount of value merely in the form of attention (which is particularly relevant to advertising). But the applications that will be built on this technology will be numerous, and some will be billion-dollar companies (think about the value generated by the last big technological paradigm shift to mobile).

If you’re bought in on the change, you might still object, “But the change may undermine my company’s position.” Perhaps — but then, the time has come to reinvent your company or join a new one. Give up short-term greed, and become greedy for the long-term gains of net new value creation. 

Advertising technologies that have been written off as boring — and companies that are regarded as having hit a relatively low ceiling — have the opportunity to leverage existing advantages (such as advertising transaction data or buy- and sell-side relationships) to lead the next wave of innovation. For example, booking direct sales for publishers is boring. Creating AI agents that transform the economics of the publishing business and allow publishers to expand their TAM, accessing spend from SMBs that is today only available to the walled gardens? Exciting. 

I’ll spill a lot more digital ink on this topic, but for now, I want to address four specific ways AI will change advertising. 

Search Ad Market Disruption

The global search ad market is about $350B. I don't think it will go away, but the modality will change from the Google Search we've known for 10-20 years to a conversational experience on AI platforms like ChatGPT. This is the biggest threat Google has faced since its inception. But I wouldn't necessarily bet against Google. They will likely figure out the paradigm shift. I'd just doubt their lead will be as monopolistic going forward, and that is a massive opportunity for challengers.

Consumers Still See Ads, But Where?

Some folks are theorizing that AI agents will significantly shrink the ad market (e.g. Rob Leathern). For example, if an AI agent shops on your behalf and finds content on your behalf, the surface area and duration of your interactions with the open web, including retail sites, declines. But net net, I don't think the size of the ad market drops as a result of this. I think the surface area and duration of your interactions remains the same — it just goes to different places (see the previous section). I might not browse retail sites in the same way. But I'll still consume content, and that content will still be monetized via ads.

Clicks to Conversations

I don't think AI agents will entirely replace the traditional process through which humans discover products or services via ads — that process will evolve. The traditional modality of ads can essentially be thought of as the provision of information: "Here's a product that might be relevant to you; click to learn more." In an agentic world, there will still be the need for paid information signals to kickstart conversations and make people aware of products. But that is the difference: we move from a series of clicks (on ads and landing pages) to a process that starts with a click and quickly evolves into a conversation (with an agent). The ad is the point of departure. The agent takes it from there, and that's a much more three-dimensional, interactive product discovery and transaction process. Michael Rubenstein is pioneering this at his new company, Firsthand.

The Birth of a New Content Ecosystem

The last major technology paradigm shift occurred about 15 years ago with the addition of mobile to a world previously dominated by desktop. This shift led to an entire ecosystem of mobile apps that needed its own ad infrastructure and networks. AppLovin is the bellwether of the mobile adtech category, and its market cap, having soared last year, is now twice that of TTD’s. My understanding from talking to folks working in this space such as OpenAds co-founder and CTO Michael Bishop is that generative apps / content / platforms, like mobile apps, need their own ads infrastructure (and it’s intuitive that they will have their own native ad networks). So, it’s not a huge leap to imagine that there will be many adtech companies, and possibly a behemoth, to support this ecosystem. Perhaps this AI-native corollary to AppLovin and TTD is the next $100 billion adtech company, and perhaps its founders are just getting started.