OpenAI is coming for Google Search, and it’s working.
“ChatGPT’s volume of search-like queries is now roughly one-fifth the volume of queries of Google,” Adam Fry, ChatGPT’s consumer product lead, told Alex Kantrowitz of the Big Technology newsletter.
It’s good news for OpenAI. For Google? Not so great. It not only shows that people are willing to use AI, but that people were really, really, fed up with Google search results.
We already knew that basic information searching is currently the top use of LLMs like ChatGPT. Pew Research Center released a report this summer that showed that the top use case was to “search for information,” with 42% of American adults surveyed saying they have used AI for that.
(“For tasks at work” was the second most common use, with 38% of employed adults reporting that. Lesser uses include entertainment, getting news, and medical advice.)
Pew Research Center
This survey wasn’t specifically about ChatGPT, but it’s by far the most popular AI chatbot (44% of American adults surveyed told Pew they’ve used ChatGPT, and only 24% have used Gemini, the next most common).
ChatGPT now has 1.2 billion weekly users. We don’t know how many of those users are doing informational searches as opposed to other things like coding or generating images, but it seems likely that someone who uses it to make images or chat would also ask it for information.
Google did not immediately respond to a request for comment on the competition from ChatGPT, but last year Alphabet CEO Sundar Pichai addressed the issue during an interview with Bloomberg’s Emily Chang.
“You’ve had these chatbots scale up to hundreds of millions of users. We’ve grown in queries,” Pichai said at the time. “So I think it feels very far from a zero-sum game to me. To use another parallel, TikTok came in, everybody started using TikTok. YouTube grew and did very, very well in those moments too.”
People started to feel like Google search had gotten worse
Google says that it handles 5 trillion searches per year; it’s dominated search for two decades, and very profitably!
But many people were getting really sick of Google search. So often in the last few years, I’ve heard friends or other people complain how Google search feels broken. Search results were so stuffed with SEO-friendly links that it often felt hard to trust anything or find the results you actually were searching for.
Users had started to add the term “Reddit” at the end of searches in hopes of getting better results — which ended up being a boon in traffic for Reddit (sadly, this didn’t last; Reddit just reported a drop in search referral traffic due to Google’s AI Overviews that it doesn’t expect to get back).
So when ChatGPT came along, suddenly there was a new way to get information and answers without all the SEO fluff — and plenty of people were thrilled. Sure, the information isn’t always reliable. Everyone should know better than to blindly trust facts doled out by AI, which is prone to hallucinations or just plain bungling, like telling people to put glue in their pizza.
Despite the issues with trusting AI for information (and to be fair, it has improved since the glue/pizza fiasco of 2024), it’s clear that tons of people were very, very happy to switch over to ChatGPT instead of Google for their informational searching needs.
I think what we should consider is that says a lot more about how bad Google had gotten than it does about how good ChatGPT is. There’s a compelling case to be made that Google enshittified its flagship product so badly — perhaps from the years and years of having no meaningful competitors — that it was a sitting duck, ready to have its lunch eaten by the first new disruptor to come along and offer people a better way.
Google eventually realized it was in trouble, issuing a “code red” in 2022 over ChatGPT. It might be time to issue another one?
