Alphabet shared its Q2 earnings this week, showing that Google Search & Other revenue increased by 17% year over year to reach $63.27 billion. For all the details, check out our earnings brief, including how the growth rate has slowed from 19% in Q1, marking the first slowdown after four quarters of accelerating growth.
The numbers also leave out part of the story. Alphabet provides a detailed breakdown of Search business earnings in its securities filing, while Google describes what the web gets back in broad assurances. Both perspectives are now part of every quarter, and the distance between them is the story.
What The Quarter Established
The revenue aspect of Search offers everything you’d hope for from a data source. Alphabet provided clear numbers, comparisons year-over-year, and consistent definitions across quarters, making it straightforward to follow the progress.
The trajectory is checkable: Search & Other growth went from 10% in Q1 2025 to 12%, then 15%, 17%, and reaching 19% in Q1 2026, before easing back to 17% this quarter.
In its earnings release, CEO Sundar Pichai connected these results to AI, saying the company’s “popular AI features are driving Search query growth.” Chief Business Officer Philipp Schindler pointed out strengths across different industries, with retail leading the way, followed by finance, technology, media, and entertainment, and credited the Gemini integration within Google’s advertising systems. However, Google didn’t specify the exact contribution of each vertical.
The spending side got the same precise treatment. Roger Montti covered this, including the capital expenditures that led to negative free cash flow for the quarter.
That said, none of this detailed financial data addresses the most important question for businesses: the number of clicks to websites. Neither the Q4 2025 nor Q1 2026 reports we covered offered this information. Google’s broader claims about outbound traffic come from other sources like blog posts, interviews, and social media updates from its executives.
3 Kinds Of Claims About The Web
Google’s public statements about website traffic can be grouped into three categories. By organizing them this way, we see what kind of data each one would need to be testable.
The first group includes usage claims. In April, Pichai mentioned that “queries are at an all-time high” and repeated the query-growth link in this week’s release. In July, Nick Fox, senior vice president of Knowledge and Information, noted that Search reached its highest usage on record during a World Cup match. These claims highlight superlatives without providing specific counts. Having more queries does not mean more clicks leaving Google.
The second group involves volume claims. Fox shared in a July LinkedIn post that AI features in Search send billions of clicks to websites weekly. He also mentioned the previous daily figure: “Actually, as we’ve shared before, we continue to send billions of clicks to the web every day through Search.” However, both numbers only specify “billions,” without exact figures. Since one is daily and the other weekly, it’s unclear how the AI-related click count compares to overall clicks or what portion of Search traffic comes from AI surfaces.
The third group focuses on quality claims. Liz Reid, Google’s head of Search, stated that AI Overviews mainly remove “bounce clicks,” visits where users quickly find a fact and leave. In an August 2025 blog post, Reid mentioned that the total organic click volume has been “relatively stable” year-over-year, with an increase in average click quality. She also explained that Google is directing slightly more “quality clicks” to websites than a year earlier. Reid defines quality clicks as visits where users don’t swiftly return to the results page. The post also challenged third-party reports of sharp traffic drops, attributing them to flawed methods and isolated cases. These statements are directional comparisons without a published baseline or measurement.
Overall, all three categories lack precise counts, consistent definitions over time, splits between traditional results, AI Overviews, and AI Mode, and site-level click reporting for AI features. Our review of Google’s blog posts and executive statements before mid-July, as covered in our Fox analysis, shows no breakdown of click figures by AI features. As of publication, that hasn’t changed. Search Console’s generative AI reports, accessible to some websites, show impressions without clicks, including details by page, country, and device type, but no click-through rates.
Compare this to the revenue reports Alphabet filed this month. One side provides a detailed time series, while the other offers three types of assurance.
What Outside Measurement Can And Cannot Test
While independent data can help fill in some gaps, each source measures different things and doesn’t fully replace Google’s overall data. Since Google has challenged outside measurements, it’s helpful to understand the limitations of each dataset.
Click-through rate (CTR) on results pages with AI Overviews provides the clearest insight. Seer Interactive’s analysis found that organic CTR for queries with AI Overviews dropped from 1.76% in 2024 to just 0.61% in 2025. Similarly, Pew Research Center’s data shows users clicked results 8% of the time when AI Overviews appeared, compared to 15% without them. Both sets of data are specific to the queries and panels they examined. Seer notes AI Overview queries tend to be informational, which naturally have lower CTRs.
Looking at position-level CTR across all queries offers a different perspective. Advanced Web Ranking’s Q1 data indicates that CTRs improved for the top five desktop positions but declined for the top mobile position. However, since this data doesn’t specifically isolate AI Overview queries, it can’t definitively support or challenge the bounce-clicks explanation.
Another perspective is assistant referrals. SE Ranking’s data shows that referral traffic from Gemini increased after the Gemini 3 releases. Keep in mind, this measures a different product than clicks from AI Overviews or AI Mode within Search.
Finally, there’s downstream behavior. A Similarweb report found that brands recommended by ChatGPT answers were 2.5 times more likely to receive a site visit within a week. Of that traffic, 55.9% came from branded searches. This suggests AI recommendations can lead to visits through indirect routes, although it doesn’t directly test Google’s claims about clicks from AI features within Search.
What Each Analysis Can Measure
Each analysis captures a different part of the traffic picture. None reproduces Google’s aggregate outbound-click claims.
Together, these outside data points show that while there is some click pressure on specific surfaces that Google’s assurances don’t fully address, there are also indirect benefits elsewhere. This makes it challenging to draw a straightforward conclusion about a loss. Ultimately, no single dataset can capture the full picture. Google’s claims refer to overall clicks across Search and AI features, but every outside measurement focuses on a specific panel, keyword set, or referral path. The two can’t be reconciled with what’s publicly available.
Why This Matters For Search Professionals
The Q2 numbers don’t settle the click debate, and you shouldn’t expect any earnings report to. They show us something a bit narrower: Search revenue can go up without revealing whether organic visits increased or decreased. The quarter added another piece of evidence to one side of that question but doesn’t add clarity to the other.
That’s the practical challenge with the overall claims. Billions of weekly clicks provide important market context, but they don’t tell you if your pages gained or lost visitors because both your site’s performance and the overall market can be true at the same time. A single figure that describes everyone doesn’t really describe anyone.
Your own data is the closest thing to a checkable click number for your site. Tools like Search Console and analytics give you a clear picture of your traffic, even if they can’t always pinpoint which AI feature affected it. If you notice impressions are increasing while clicks are dropping on informational queries, that pattern is consistent with what Seer found with AI Overview queries. Broader CTR data remains mixed across devices and query types.
For publishers, the key indicators to watch are how referral shares from search change over time, and how branded versus non-branded traffic compares. The pattern from Similarweb suggests that AI recommendations might lead to branded search visits rather than direct referrals, so seeing a decline in referrals alongside growth in branded searches can tell a different story than a simple drop.
Be sure to keep paid search analysis separate from the organic side. Google’s expanding ad coverage in AI experiences is mainly about monetization, and it’s a separate track from organic traffic. An increase in AI ad inventory is not evidence that organic clicks went up or down.
When clients ask if Google’s reassurances should influence their strategies, a good answer is that the reassurances can’t be tested, but your own data can. Decisions should be based on the verifiable information you have.
Looking Ahead
This quarter’s gap isn’t between Google’s claims and independent data. It’s between two different disclosure standards within the same company, with only one set of numbers that everyone can verify.
To bridge this gap, we need some specific details: click and CTR data in Search Console’s generative AI reports, a published methodology behind the aggregate claims, and detailed breakdowns by surface. As of now, these haven’t been shared.
Still, Alphabet’s upcoming quarterly report will offer a snapshot of revenue, regardless of these missing details. The earnings reports from the past three quarters haven’t included AI-feature traffic data, and Google hasn’t suggested it will include this information in the future.
More Resources:
Featured Image: Blue Titan/Shutterstock
