Google: AI Mode Queries Run 3x Longer as Search Turns Exploratory

Queries in Google's AI Mode average three times longer than traditional searches, according to figures Google's Ginny Marvin shared on October 28. The episode of Ads Decoded, Google's ads podcast, also reported that more than one in six AI Mode searches in the U.S. is multimodal.

Marvin, Google Ads Product Liaison, appeared alongside Shivani Mohan, Search Vice President of Data Science and User Experience Research. The discussion covered how AI-powered search changes user behavior and what advertisers should weigh for 2027 planning.

What Did Google Confirm?

In a LinkedIn post, Marvin outlined three takeaways:

• Discovery is moving up the funnel. People use conversational search for brainstorming and discovery well before they reach a purchase decision.
• Curiosity is expanding. AI Mode queries run 3x longer than traditional searches, and one in six searches is multimodal. Marvin's post does not specify baseline definitions or the measurement window.
• Users want human perspectives. Podcasts, video reviews, and firsthand experience remain part of how people research decisions.

The official episode notes add detail. People ask 3x longer questions, go back and forth on follow-ups, and search with voice and images. That gives advertisers more signal than the initial search phrase alone.

How Solid Are the Figures?

Google's October 28 advertising announcement cites the same 3x figure and links to a U.S. AI Mode report published in May 2026. The statistic compares average query length. It does not show how differences in users or search tasks affect the comparison.

Longer questions give advertisers more information about user needs. The figure alone does not demonstrate stronger buying intent or better conversion rates.

The report's multimodal claim is narrower than Marvin's wording. It says more than one in six AI Mode searches in the U.S. is multimodal. That does not apply to all Google searches. Both figures are confirmed Google statements, not independent tool measurements, but they lack published methodology.

Why Do Users Still Want Human Content?

Mohan wrote on LinkedIn that users:

…appreciate the streamlined clarity of AI summaries, yet they actively crave the rich detail, lived experience, and earned authority of the open web.

She named forums, podcasts, and videos as destinations for those perspectives. For advertisers, that means customers want more than a product description. They want to know how a product performs, where it falls short, and why someone chose it.

Mohan also referenced billions of clicks to the open web, without a timeframe or comparison. The figure does not show whether traffic is growing or how clicks distribute across sites. Her comments describe demand for firsthand experience, not a guarantee that more videos will bring more Google traffic.

The episode notes also suggest discovery and evaluation can happen inside one conversation. A user can explore a problem, then ask about options and tradeoffs as they learn.

What Should Paid Search Teams Do for 2027?

On October 29, two commenters on Marvin's post asked how the behaviors translate to campaign management. Md Jihad asked how small businesses should adapt ad copy and landing pages for exploratory intent. Ana Trujillo Martin asked which signals retail teams should watch, including how to interpret impressions and click-through rate.

Neither the posts nor the episode notes provide specific campaign settings. The discussion does support four planning checks:

• Find pre-product needs. Review customer questions, sales conversations, and search terms for recurring problems. Test ad messaging that addresses those needs.
• Enable comparison. Audit landing pages for answers on price, compatibility, setup, and limitations. Make them easy to find before a quote request or purchase.
• Connect content to decisions. Link reviews, demonstrations, and customer stories from landing pages. Give visitors a clear next step once a concern is answered.
• Measure downstream impact. Where measurement allows, check whether visitors to comparison or educational pages return and convert. Impression, CTR, or engagement shifts alone do not show business results.

What Comes Next?

The podcast and posts leave advertiser questions open, including how these behaviors vary by industry and buying stage. Google has described longer queries and earlier discovery but has not supplied conversion benchmarks. Advertisers will need to test which customer questions their content leaves unanswered and whether closing those gaps moves users toward purchase.