Google acknowledged the AI Overviews fiasco that unfolded over the past few weeks after Google I/O, and as a result, it is now drastically reducing the probability of an AI Overview appearing as a response to the user’s search query. This is significant, considering AI Overviews has played a major role in Google’s AI product portfolio.
A report from Wired which cites data from Brightedge suggests that the average appearance of AI Overviews in search results has dropped to just 11 percent of queries, down from around 27 percent when the tool launched publicly in mid-May. Compared to back in December 2023, when AI overviews was an opt-in experience called Search Generative Experience, the visibility of these has dropped down by a whopping 84 percent.
Industry-wise, Google’s AI Overviews continue to appear most often in Healthcare where 63% of the times you’ll see an AI overview, down from 76% in January. AI Overviews now appear in less than 1% of searches for Restaurants (down from 36% in January) and Travel (down from 30% in January).
In a statement to Wired, Google said, “We’re continuing to refine when and how we show AI Overviews so they’re as useful as possible, including a number of technical updates in the past week to improve response quality.” However, Google didn’t corroborate the stats shared by Brightedge or gave its own.
Google said in a blog post earlier last week that the odd results in AI overviews were due to what’s called a “data void” or “information gap,” where there’s a limited amount of high quality content about a topic. Because of these data voids, users were served with results such as when they were told eat one rock a day when they asked Google “How many rocks should I eat?.” Google has identified that it needs to improve the results in areas including where it has to interpret nonsensical queries and satirical content.
This wasn’t just the issue with Google’s Gemini-powered AI overviews, as even content creators and publishers have criticised the company for this specific feature.