Google Search: AI Boosts Queries Without Increasing Revenue

AI-driven queries are progressing rapidly, and Google is multiplying the associated advertising formats. However, the growth of Search revenue is slowing down, paid clicks are increasing at a slower pace, and the actual impact of AI ads remains difficult for advertisers to measure. Between experiments and massive investments, the profitability of conversational search has yet to be demonstrated.

Unclear Advertising Performance and Mixed Signals

In the United States, visits generated by ChatGPT ads brought in 64% less revenue per session for retailers compared to those from Google search ads, at $0.89 versus $2.45. This study focuses on merchant sales without measuring platform revenues, and it does not allow for conclusions about the performance of Google’s AI ads. For now, advertisers cannot specifically target or isolate the performance of Google’s AI inventory, complicating their evaluation. The stable overall results could be explained by still low AI traffic not having a significant impact or by performance similar to existing inventory; Andy Taylor believes that AI volume remains limited. Uncertainty persists regarding the share of additional revenue generated by new AI queries compared to potential cannibalization of better-monetized searches. Conversational exchanges encourage users to share more information, but it remains to be determined which segments of these dialogues are relevant and how to connect them to suitable ads.

Variable Purchase Intentions and Cautious Comparison Reading

According to Andy Taylor (Tinuiti), a product search like "Nike" still displays classic shopping and text formats, as it reflects an advanced purchase intention. In contrast, asking an AI agent which running shoes to choose corresponds to a preliminary research phase. Taylor emphasizes that performance differences between channels heavily depend on purchase intent. A better return on sales may thus reflect an audience closer to making a purchase rather than a superior effectiveness of an advertising format.

Google’s AI Strategy: Expanding Ads and New Entry Points

Google is currently expanding the display of ads in AI summaries and AI mode, with international deployment and testing across multiple conversation turns in the United States. The company has also mentioned the possible arrival of ads in its standalone Gemini app, reusing formats designed for AI mode, without a set timeline. According to Philipp Schindler, most AI mode queries would be additional to existing searches and often occur earlier in the purchasing journey, which Google sees as opportunities to sell more ads in the long run. The search engine is already integrating generative summaries and a conversational mode based on the query.

Slowing Search Growth Despite Increasing Queries

According to Alphabet, the number of search queries is increasing at an unprecedented rate over the past five years. In contrast, revenue growth related to search was limited to 12% in the third quarter, down from about 17% in the previous quarter, and the increase in paid clicks dropped from 13% to 8%. Google claims that AI is boosting search activity, but this momentum has not translated into accelerated revenue growth.

Factors Cited by Google and Debates on AI Monetization

Philipp Schindler attributes the slowdown primarily to currency effects and, to a lesser extent, to changes in advertiser spending in certain sectors. He acknowledges that AI mode queries are currently monetizing less effectively than traditional searches, while expressing confidence and mentioning promising advertising tests. Uncertainty remains regarding the exact origin of this performance: gradual advertising deployment or structural limits of advertising in AI search. Nate Elliott (eMarketer) believes that AI represents the first major disruption to Google’s core business in decades, changing the way people search and click. Andy Taylor and Nate Elliott remind us that a decline in clicks would be a problem for Google to solve.

Costs, Market, and Learning Horizon of Formats

Google is forecasting about $200 billion in infrastructure spending this year, primarily to support AI. Since the earnings announcement, Alphabet's stock has fallen by about 14%. Nate Elliott points out that finding an effective advertising format can take years, as search engines have transitioned from banners to sponsored links, and social networks have experimented with various formats. The formats being tested today may not be the ones ultimately adopted. The stakes are rising with the increasing costs associated with AI, while Search revenues continue to grow at double-digit rates and the profitability of the new search approach has yet to be demonstrated. On the advertiser side, Tinuiti's clients increased their paid search investments by 14% in the third quarter without compromising their advertising return objectives. Andy Taylor believes that Google’s historical strength gives it time, citing the mobile precedent, which was initially less profitable before the shopping experience on phones improved. For Nate Elliott, no one has yet fully solved the advertising equation for AI search.