Ever since Google decided to embrace AI search, rather than fight it, a question has lingered about how it will monetize the new paradigm. Among other things, there’s an ad inventory shortage in moving from several links to one answer. Google has partly answered that question with multiturn advertising.
Recently rolled out, these ads are meant to cover every turn of the conversation in AI-mode dialogues. Since users are asking several successive questions, it makes sense that each one contains a contextually-relevant ad. And it redeems some of the ad inventory Google enjoyed in traditional SERPs.
Oversimplifying that for the purposes of demonstration, traditional SERPs contained an average fill rate of 3 ads. That goes for general search queries, as well as local – given the classic 3-pack. So Google’s target in matching or exceeding its traditional ad inventory is about 3 ads per search query.
Multiturn ads get Google closer to that goal. The rough math: one search with 3 ads is equivalent to 3 turns of an AI-mode dialogue, each with one ad. Of course, real-world conditions are more complex, but this simplistic math starts to give an idea of what Google is going for with the rollout of multiturn ads.
The timing is also right given that Google’s Q3 earnings signaled quality declines for AI-mode ads. Though they’re still in development and will likely evolve in their targeting and overall performance, Google reports that AI Mode ads currently show a lower monetization rate than traditional search.
Both Ends of the Candle
Sticking with that last point, it was surprising when Google reported that ads in AI search are underperforming versus traditional search ads. We say that because AI Mode gives Google much more surface area by which to infer user intent. It can therefore target ads better, at least in theory.
To unpack that a bit, the greater surface area that AI Mode stimulates specifically refers to word count. Whereas traditional search involved a few keywords, AI Mode is all about full sentences. That means users in AI Mode tend to explain exactly what they want, reducing inference and guesswork for Google.
As a side note, we were recently remarking during a guest spot on SOCi’s monthly SEO Juice webinar that it’s amazing how quickly Google has untrained the world on 25 years of keyword search behavior. Though it still has a ways to go, a significant portion of users habitually input natural-language queries.
The bottom line is that AI mode has theoretical advantages in the quality of ad targeting it’s capable of. Though Q3 earnings indicate that this hasn’t been fully realized, the potential is still there, and we’re confident that Google will refine and tune things to yield more effective and performant results.
Altogether, Google is burning both ends of the candle considering AI Mode’s evolution on qualitative and quantitative levels. The latter brings us back to multiturn advertising, and that will join several efforts, such as bringing AI Mode into apps, and other monetization-boosting moves we can expect rapidly.
Header image credit: Kelly Sikkema on Unsplash


