On May 14, 2024, Google rolled out AI Overviews in the U.S. Just three months later, the feature expanded to six more countries: the UK, India, Japan, Indonesia, Mexico, and Brazil. Now it’s available in even more places, suggesting that Google feels pretty confident about how well it’s working.
Google is like the main gateway to the internet. But with the way people search evolving, it risks losing that top spot. If Google doesn’t adapt, it could become a huge threat to its future.
Lately, Google has been slowly testing some AI tools in search. They seem to be careful about not:
- Freaking out loyal users (We’re all used to the same old Google look, right?)
- Turning Google Search into just an “Answer Engine” (That’s what Gemini is for).
In the end, Google’s challenge is balancing both current users and future users without scaring off either group.
That takes us to, AI Overview.
After using AI Overviews for a few months, I feel like this might be Google’s first real success at slowing down the shift to answer engines. Have you noticed how Google feels more like a hybrid engine now? It’s still mostly a search engine, but with some answer engine features sprinkled in (though you can’t ask follow-up questions yet).
And honestly, that’s still pretty great. When we’re on the go, most of the time, we just want a quick answer. For deeper research, though, we’ll probably still head over to a full-on answer engine.
Lets take a closer look on whats brewing here.
There are a few things happening with Google’s search results now. First, there’s the AI Overview, which works like a one-way answer engine – kind of like listening to a radio. Then, Google kept the Featured Snippet, which has been a great tool since it launched in 2014. Finally, you still get the usual search results and links underneath.
If you scroll down a bit, you’ll see something else – the related questions section. This part could eventually evolve into a way for users to ask follow-up questions directly to the AI Overview, closing the loop nicely.
While browsing through all these features, I couldn’t help but notice something interesting. Check out the comparison below – Google and Perplexity are flying pretty close to each other. Both of them seem to have a similar feature: “related questions.”
Is Perplexity trying too hard to become Google, or is Google changing for good?
AI Overview Response appears to unique
At first, I thought the AI Overview answers were just pulled from the SERP (Search Engine Results Page) links below. But after running a few test searches, I noticed they’re close but not identical. The “Sources” cited in the AI Overview don’t always match the ranked search results, and they often appear in a different order.
This makes me think the AI Overview and SERP might be running on different engines. Google is probably testing both approaches to see which one users prefer. It wouldn’t be surprising if they’re using this feedback to refine things further.
Thats like in a way AB testing Google’s own search and AI capabilities?
Personally, I’ve found myself reading the AI Overview quite a bit, and I’m pretty satisfied with the responses. They feel more relevant—definitely better than the quality of the current SERP results for quick answers. It works really well for those searches where I just need the answer right away.
For other types of searches, like finding the exact website link for a brand I already know (like Expedia or the Wait But Why blog), the traditional Google SERP still gets the job done perfectly.
Still a fair bit of the road ahead is unclear…
With all that said, this feels like a masterstroke by Google. It strikes a nice balance that could slow down the drift of users to other answer engines. But the big question remains: What’s next?
Will the AI Overview eventually evolve into a feature where we can interact with the response, leading us into a Gemini-like interface?
Another issue is click-through rates. This shift could seriously hurt the traffic for regular websites, especially since AI Overview doesn’t encourage us to click through to verify sources—much like how we tend to use answer engines.
There’s also a big question around Google’s AdWords and AdSense revenue. If users get their answers straight from AI Overviews and leave without interacting with search ads, what happens to Google’s main revenue stream?
It makes you wonder: Is Google betting on traffic now and figuring out the money later?
There’s still a lot to uncover here, and I’d love to know what others think about this. Feel free to share your thoughts in the comments—I’m curious to see how everyone is reading this situation. Thanks for dropping by!
I actually liked the AI overview and subconsciously use it quite often. It still feels incomplete… especially when you want to followup. I think it just improved the search experience when you want a quick answer. But for deep dives, i still prefer openAI or perplexity..
Agree Steve, its an improvement but not addressing the elephant in the room!
Did you notice that the AI Overview is reduced in in its presence greatly? Recently read that Google has reduced the experimental feature from 20% to under 10%? what does that say, is it failing? keen to hear your thoughts..
yeah have noticed that but they are specific to some categories that they may have seen a decline. But will be very interesting to watch where this goes long term. Also, the arrangement of ads and section is changing quite fast…
Saw this in their earnings call (latest one) that they have started seeing positive signals from the way AI overview is being consumed by search users. However the tricky bit will be how are they going to monetise it?