Google AI Study: Lost Clicks Not Lower Quality

The extra outbound clicks that websites gain when Google’s AI Overviews are removed appear to be just as valuable as the clicks generated when the summaries remain in place. That conclusion comes from an updated, randomized field experiment, which recorded a 39.8% drop in organic clicks when the overviews were shown.

Researchers Saharsh Agarwal and Ananya Sen expanded their working paper with fresh analyses of click quality, treatment switches, and query types. When Search Engine Journal first covered the study in April, the reported reduction stood at 38%.

The revised data shows that outbound clicks per search increase and zero-click searches rise when AI Overviews are active. Since the overviews appeared on roughly 41% of queries, the result is a measurable decline in outbound clicks across all searches.

What The Click Quality Data Shows

For queries where an AI Overview would have been displayed, the authors compared how often users hit the back button to return to the search results, how frequently a visit ended within 10 seconds without interaction, and the total time spent on the destination site.

None of these three metrics showed a statistically significant difference between the groups. In both conditions, about 4 in 10 same-tab clicks led back to the results page, roughly 18% of visits ended within 10 seconds, and time on site was statistically indistinguishable.

Google’s Vice President of Search, Liz Reid, has argued that AI Overviews reduce “bounce clicks” but has not released data to support that claim. This experiment approaches the same question from the opposite direction. If the summaries had absorbed mostly low-value visits, the extra clicks in the no-Overview group would be expected to look worse , but they do not.

The authors write that the result is “at odds with the view that AIOs primarily eliminate low-engagement website visits.”

Disclosure: The revised paper cites two Search Engine Journal articles in its discussion of Google’s claims.

Behavior Reversed When The Groups Switched

After the initial two-week period, the researchers rotated the assignments.

Participants who were served AI Overviews saw their external clicks per search decrease, while those who no longer received them saw their external clicks per search increase. Notably, the zero-click rates mirrored each other.

Informational Queries Carry The Effect

When the authors examined the data by query type, they found that the losses are concentrated in informational searches, while removing AI Overviews increased clicks per search. Navigational and transactional queries showed no measurable change, though those samples were smaller.

AI Overviews were triggered or intended to trigger on 53% of informational queries, compared with 15% of navigational and 6% of transactional ones. A position breakdown adds that the top three ranked results gained the most clicks when a top-of-page overview was removed, with position one nearly doubling.

Why This Matters

Google’s response to concerns about lost traffic has relied on the idea of increased click quality. This experiment tested that claim and found no measurable difference.

Looking Ahead

The authors write that aggregate traffic losses could grow if AI Overviews appear on more queries over time. The paper is a working draft on SSRN and has not yet completed peer review.