Google’s core Search business delivered 11% y/y revenue growth in June compared to 10% growth in March. Query volumes continue to grow 4% y/y indicating that user engagement with Search has not been eroded by AI chatbots as I had feared. The message from management was similar to the March quarter, that new generative AI features are driving incremental usage. AI Overviews are driving 10% more queries globally in the categories where they appear, evidence that it is having a positive impact on engagement. That said, the long-term question remains; how does the company ramp monetization in AI Overviews and AI Mode with a user experience that will likely move away from the classic blue sponsored links.

Sundar noted on the earnings call that AI is unlocking new kinds of questions people can ask Google. My sense is he is right. Google is the first stop for information for around 3B people daily, and as consumers learn about AI, it’s logical that they find their way to Google to explore. That’s been the genius of AI Overviews, essentially training everyday users on the value of AI results.

Taking a step back, Google’s Search business remains the workhorse of Alphabet, accounting for approximately 55% of revenue and 75% of profits, and showed stable growth in June despite six months of generative AI hype in the marketplace.

We learned something from Eddy Cue’s May comments. In early May, Apple’s Eddy Cue testified about a dip in Google searches on Safari due to users trying AI chatbots. Google disputed that claim, stating it continues to see overall query growth even on Apple devices. The June numbers back Google’s stance. Any isolated traffic softness in Safari was more than offset by growth elsewhere, such as the Google app, Chrome, and Android. This episode underscores how sticky search is, and if consumers shift from accessing it in one place (Safari), they find another place (Chrome).

On the call we picked up a few helpful datapoints on AI Mode, the most aggressive Search reinvention in 20 years. The company noted that there are now over 100 million monthly users. That compares to my estimate that there are 5 billion monthly Search users. That means AI Mode has 2% of the usage compared to Search. A small number, but in fairness to Google it started less than two months ago. As a point of reference, I was expecting the monthly number to be 50 million in the June quarter, so it’s moving along faster than what I would have expected. It’s available in the US and India, which account for 25% of the world’s population excluding China (where Google is not available). Adjusting for the availability, you can infer that AI Mode usage is tracking more like 8% of total Search usage, which underscores the impressive start. On top of that, AI Mode is not easy to access given it’s only available on Google.com, sidestepping the vast majority (80% by my estimate) of search traffic that comes through the navigation bar in browser. Turning the switch on Chrome and Safari navigation bar will have a massive impact on AI engagement. That likely won’t happen for a couple of years if ever because the company will want to make sure AI Mode is monetizing at an equal rate to the traditional landing page before making the switch.

Google is already testing ads in AI Mode, a trend I have not seen in my usage of AI Mode. While details are limited and no data was shared on click-through rates, the strategic intent is clear; Google is transplanting its existing business model into AI Mode.

This all comes down to a basic question: Can AI native ads be as effective? We won’t learn much on that front in 2025 given the ad load in AI Mode will be light, and will likely have to wait until early next year to get a better read. My bottom line is there is a lot at stake as they ramp ads in AI Mode, and the risk of injuring the golden goose is still on the table for the long-term.

June Cloud revenue was up 32% y/y compared to 28% growth in March. Google Cloud’s operating income improved to $2.8 billion, a 20.7% margin. This marks the third straight quarter of profitability for Cloud. If this trajectory continues, Google Cloud could generate over $10 billion, or 8% of expected 2026 operating profit.

Among a series of Cloud datapoints, the one that jumped out to me was new Cloud customer additions were up nearly 28% quarter-over-quarter. The second most valuable datapoint: Gemini AI models drive a 35x increase in usage compared to non-AI customers. The takeaway is AI is having a profound positive impact on new customer adds and increased Cloud usage.

One leading indicator that the growth will continue at a similar pace is the Cloud backlog in June was up 38% y/y, which is slightly ahead of the reported 32% growth. Backlog growing faster than reported revenue is a trend that allows investors to sleep well at night.

Capex guidance for the back half of 2025 increased by 13%. For the year it’s expected to grow at 61% y/y compared to expectations before they reported June of 43% growth. Bottom line: We’re still early in the AI buildout.

Google continues to aggressively invest in its infrastructure to meet the internal demand for compute and build out Cloud. The company raised expected 2025 capital expenditures to approximately $85 billion, up from $75 billion.

Of that $10 billion annual increase, $4.5B came from more spending than expected in the June quarter, and $5.5B from anticipated spending increase in September and December. This basically means that the back half of the year will see capex 13% higher than what they told us three months ago.

There are two takeaways from this update. Despite investor concern that we are one quarter away from the AI infrastructure showing a meaningful step down in growth, the opposite is happening. The growth continues to be higher for longer. I believe this trend will be echoed next week with comments from Microsoft, Meta, and Amazon. Second, we are still early in AI. Much earlier than what most investors think. When competent companies like Google are meaningfully doubling down on the paradigm shift, that’s a tell to investors that something meaningful is on the horizon. Keep in mind, these companies have the capital to will AI chatbots into general intelligence, which will be an unlock for AI applications across the board.

I would sum up Waymo’s progress over the past three months as steady. The most relevant datapoint was its vehicles hit 100 million miles driven autonomously, up from 50 million six months ago. That’s a sign that usage is ramping. That said, they did not update the February 2025 milestone of 200k rides per week. My sense is that they are likely at 275k per week, and waiting for a round milestone, 300k or 400k per week to give the next update. During the quarter the segment started service in Atlanta and doubled its service area in Austin, and expanded its Phoenix, Los Angeles, and San Francisco territories by a lesser margin. Waymo now has over 1,500 vehicles. Testing is underway in New York and Philadelphia.

The big picture: Waymo is growing like a weed, but it’s still tiny compared to the US ride sharing market. By my estimate, Waymo accounts for (assuming 275k rides per week) about 3% (just over 4B rides in the US) of the US ride hailing market. To get to 50% share, I believe they will have to add 100k to 200k more vehicles. The basic math suggests if they can gain 3% share today with 1,500 vehicles, then they will be able to service 50% of the market with 25k vehicles. In reality the way the coverage map works, it will require exponential more vehicles to scale given the growth above the 30% share will require service outside of dense urban regions.

So what does it mean for Waymo and Google’s valuation? Waymo raised external funding in late 2024 at a valuation over $45 billion. I believe the market is likely assigning around $100 billion value to Waymo within GOOG stock. I believe if they gain 30% of the US ride hailing market it would add 15% to Alphabet’s overall valuation. If they gain 70% (like Uber), it would add an additional 35% to the company and be worth $850B by 2030.