M.G. Siegler • #google search • May 8, 2025

Trying to Solve the Apple Safari Google Search Riddle

Apple says search queries are falling, Google says they're not...

Given that Google's stock fell 7.5% yesterday on the testimony of Apple exec Eddy Cue at the remedies portion of Google's search antitrust trial, the company probably had to respond. And so here it is, in full. Google responding to one very specific, particularly damning data point from Cue:

Here's our statement on this morning’s press reports about Search traffic.

We continue to see overall query growth in Search. That includes an increase in total queries coming from Apple’s devices and platforms. More generally, as we enhance Search with new features, people are seeing that Google Search is more useful for more of their queries — and they’re accessing it for new things and in new ways, whether from browsers or the Google app, using their voice or Google Lens. We’re excited to continue this innovation and look forward to sharing more at Google I/O.

While there is no transcript of what Cue actually said, reading dozens of reports on the matter would seem to paint a pretty clear picture that he noted that search queries fell in the Safari browser for the first time ever last month. "That has never happened in 22 years," is his direct quote many publications are citing.

So how do we square that with Google's response above? In particular, the notion that: "We continue to see overall query growth in Search. That includes an increase in total queries coming from Apple’s devices and platforms."

As I noted yesterday, it seemed entirely possible that Google's search query health was fine overall – something the company keeps insisting, which already isn't a great sign if you have to keep reaffirming that – while Apple's was falling. Apple could be some sort of outlier, or, more likely, perhaps a harbinger of what's to come, thanks to their younger and more affluent user base. As big as the iPhone is, it's still undoubtedly just a slice of Google's overall query volume. Perhaps a sizable one, but also perhaps not sizable enough to shift the entire needle negative. But again, that's why the second part of Google's statement here is key: "That includes an increase in total queries coming from Apple’s devices and platforms."

So is Google saying Eddy Cue is lying? Is Eddy Cue lying? Under oath?!¹ I think there's a way, perhaps a few ways, to square this circle.

Again, it's hard to know for sure without the direct transcript, but assuming Cue did simply say that search query volume in Safari fell for the first time last month, you'll note that he's not specifically talking about Google there. But it's basically implied. Because Google is the default – a big part of what this whole trial is about! – it seems safe to assume that nearly all of the overall queries are Google.

Except in one place: China.

Google famously doesn't operate in China and so Chinese users, unless they've broken through the Great Firewall, are not using Google for their queries. They're using Baidu, the default search partner there.

So what portion of the iPhone user base is in China? Let's turn to... ChatGPT for that. As of early 2025, the number seemed to be 17%. And if we assume roughly 1.4B iPhones in use overall, that's about 240M iPhone users in China.

Because Cue (apparently) didn't specify Google searches falling, it's definitely possible that Chinese users are driving an overall query volume drop. So perhaps it's less bad news for Google and far worse news for any of the China-specific search engines. We obviously can't know that for sure from Cue's comments. And it would seem a bit odd to have one country's usage be so different from the rest of the world so as to skew the numbers, but again, it's possible. Perhaps Chinese users are ahead of the curve when it comes to using AI for most queries. That certainly seems possible. But again, at such a volume to move all the data?

One more wrinkle here is that Apple famously/infamously hasn't rolled out Apple Intelligence in China yet. Because they needed a local AI partner – i.e. not OpenAI – there was a delay beyond the delays in the rest of the world in rolling out any AI features (they should be coming soon). Still, the iOS AI integration is probably not what moves the needle here anyway, and Chinese users can obviously download AI apps right now to use. And perhaps they are at a rate far above users in the rest of the world. In fact, maybe they are downloading and thus using AI apps more than they would otherwise because there’s no built-in Apple option? It's possible!

Another possibility: Cue may have just been referring to queries on the iPhone in particular. Again, there's no indication that's the case, but given how big the device is relative to other Apple products, and how vital it is to this case, it's possible he decided to focus in on that product for his comments (without specifying he was doing so).

And perhaps that's why Google, in their statement, very specifically noted "an increase in total queries coming from Apple’s devices and platforms". Maybe there's some wiggle room here in terms of query volume on the iPhone versus the Mac and iPad and every other Apple device. All of which also have Safari.

Perhaps the most likely scenario: the semantics one. That is, while Cue noted that queries within Safari were falling for the first time, and Google noted that queries across Apple devices and platforms were rising, what if the discrepancy is simply in people using other mechanisms to do Google queries on Apple devices?

This resonates with me because this is what I actually do. Most Google queries I perform on the iPhone and iPad are through Firefox Focus, a simple, private browser app made by Mozilla. I prefer it simply because I find it far faster to perform a query (mainly because that's basically all you can do with it), and to a lesser extent because I don't want nor need most of my Google searches saved forever, tied to my account.

Meanwhile, on macOS, I predominantly do Google searches not through Safari, but through my browser of choice: Arc. And while I'm undoubtedly an outlier in this regard, certainly many, if not most, macOS users are running Chrome as their main web browser. You know, the other key element in all of this. As such, any Google searches are going through there, not Safari.

What if it's Safari that's dipping overall and not just the query volume going through it – i.e. the query volume is dipping because the overall usage is dipping? That doesn't seem to be what Cue is suggesting, but again, you could argue it, semantically.

Meanwhile, Google's own statement seems to be suggesting that perhaps people are using other ways outside of Safari to do search, "the Google app, using their voice or Google Lens". Apple, of course, also has a deal to use Lens technology as a part of the Visual Intelligence aspect of Apple Intelligence. But presumably it's not a huge volume driver right now. Are any of these others?

It's worth noting that Cue himself immediately followed his comment about the query volume drop by noting that he believed AI services were eating into it. But without more detail there, it could be just like, well, you know, his opinion, man.

Regardless, the wrinkle here would be less about the drop an more about the change in trend. The fact that this was the first drop ever, again, in 22 years, according to Cue. So either something changed in a major way with user behavior/usage of search within Safari or overall app preferences, including, perhaps, using more AI apps for search, changed last month for some reason. Either way, it seemingly says something!

Again, it's hard to know for sure without more color from Cue. But these are three potential ways to square that circle. And perhaps it's even a mixture of the them.

Still, the high level point remains intact. The market is freaking out because it believes that Google Search is going to be disrupted by AI and Cue's comments were the first real data point indicating it was starting to happen. It may not yet be happening in the way that his comments were interpreted, but that doesn't mean it's not going to happen. The real question is if Google can be the ones to disrupt themselves here. They're clearly working like hell to do just that with Gemini.

A product which, incidentally, should be baked into the iPhone soon.

Disclosure: I worked at Google for 11 years as a partner at their venture fund, GV. Obviously, my thoughts are my own on these matters.