Consensus forecasts total revenue up 11% year over year and GAAP EPS up 13%. This compares to revenue growth of 12% in March 2025. Revenue has beaten expectations in 15 of the past 20 quarters, and EPS beat in 16 of 20, reflecting management’s tendency to outperform.

Search remains the workhorse and should account for about 55% of revenue in the quarter and 75% of profits. The Street expects Google Search & Other ad revenue to be up 9% YoY in June. That would be a slight deceleration from 10% in the March quarter and below the 13% Search growth for full year 2024. For the back half of calendar 2025, analysts project Search growth to remain around 9% YoY, suggesting analysts don’t expect a measurable impact from ChatGPT this year, and Google’s search business is expected to remain stable for the back half of this year.

Outside of Search, YouTube will likely grow at 10% YoY, in line with the 10.3% in March.

The cloud segment continues to be a growth standout (28% growth vs. Search at 9%), and profitability is improving. In March, Google Cloud revenue was up 28% YoY, ahead of AWS (+17% YoY) and close to Azure (+30% YoY). For June, consensus expects $13.1B in Cloud revenue (26% YoY). Key metrics:

• March 2025 revenue: $12.3B, up 28% YoY
• June 2025 consensus: $13.1B, implying 26.6% YoY growth. Cloud is now a $50B+ annual run-rate business.
  A year ago in June, 2024 revenue was $10.35B, up 29% YoY.
• Operating income was positive in March, with 18% margins.
• Last quarter, management indicated that AI demand exceeds current infrastructure capacity, prompting $75B in expected 2025 capex.
• If margins continue expanding alongside mid-20s% growth, Cloud can be a $10B+ annual profit contributor by 2026.

Waymo

Waymo continues to gain traction in recent months, which could add upside (or at least positive sentiment) to Google’s valuation. Three months ago, the company highlighted that Waymo is now providing 200,000 paid rides per week, a 20× increase in two years. In July, Waymo announced it expanded its autonomous service area in Austin to 90 square miles, effectively doubling its coverage and “responding” to Tesla’s robotaxi expansion. It’s also launching new cities: commercial ride services in Atlanta (in partnership with Uber) and Washington, D.C. are expected to begin in the coming months, and a Miami launch is expected in early 2026. Waymo now operates in Phoenix, San Francisco, and Los Angeles with 1,500+ autonomous vehicles in its fleet (Tesla is at about 50). Lastly, in July, Waymo hit 100 million miles driven with no human driver, doubling that metric in six months.

Waymo’s progress begs the question: what does it add to Google’s overall valuation? I believe Waymo could be worth $350–$850B by 2030, if spun off, which would—assuming none of it is priced into GOOG shares today—add 15–35% to Google’s overall valuation. My sense is the Street is valuing the business closer to $100B today (the last round was October of 2024 at $45B), which suggests Waymo can still move shares of GOOG higher by 10–30%.

Here’s my note from last fall on the Waymo valuation topic.

I’ve been critical that Google is not moving fast enough when it comes to AI. After reviewing the last earnings call and May’s Google I/O event, it’s clear that the company has been racing to infuse AI into Search to defend its dominance.

Google I/O 2025 (May)

• Sundar highlighted the “relentless” pace of AI progress, noting Google is “shipping faster than ever.” That means the company is no longer saving breakthroughs for the annual I/O event, but rolling them out as soon as they’re ready.
• “AI Overviews are driving over 10% growth in the types of queries that show them… it’s one of the most successful launches in Search in the past decade.” — Pichai. I believe the “growth” he is referring to is in engagement. AI Overviews, which are AI-generated summaries, launched in May 2024, so the company now has a solid year of data for comparisons.
• There are over 1.5B users per month now seeing these AI Overviews, which compares to about 4.5B monthly Search users, suggesting AI Overviews is still early.
• “For those who want an end-to-end AI Search experience, we’re introducing an all-new AI Mode. It’s a total reimagining of Search… all of this is available as a new tab right in Search.”
• “We want to get our best models into your hands (developers)… so we’re shipping faster than ever.” This shows they understand speed is important.
• After I/O, Google indicated they will be bringing ads to AI Mode. My experience with AI Mode has not yielded any ads. The key question is how many ads will find their way into AI Mode, and will the presence of ads diminish the consumer experience enough that they’ll go to Grok or ChatGPT?

March 2025 earnings call
The bottom-line message was AI is a tailwind for Search, not a headwind.“In Search, we saw continued double-digit revenue growth (it was 10%). AI Overviews is going very well.” — Pichai.

• “Q1 marked our largest expansion to date for AI Overviews, both in terms of launching to new users and providing responses to more questions.” — CBO Philipp Schindler. The message is that Google is confident in AI Overviews to increase engagement, as evidenced by its expanding availability.

Management was asked about generative AI’s impact on the ad business. The tone was reassuring: Pichai and team suggested that AI integration makes Google Search more valuable to users and advertisers, not less. They pointed out that Search revenue growth continues to be strong despite worries about generative AI platforms like ChatGPT. Advertisers can potentially run more effective campaigns with AI-driven search experiences. In short, Google argued that AI is augmenting Search (bringing in new queries and improving relevancy), rather than cannibalizing it.

Google’s new “AI Mode” was announced on May 20th, and within six weeks rolled out to the US and India, which by my measure is the fastest new product launch for the company. AI Mode is basically search on ChatGPT. So far, Pichai says user feedback has been “early positive.” Accessing AI Mode currently is out of the typical search flow for most Google users, given it requires visiting the Google Search site or app and tapping the “Generative AI” toggle/tab (it’s not yet the default in Chrome’s omnibox or Android’s search bar). This is important because a majority of Google searches (I estimate 80%) originate from the navigation bar inside of a browser rather than users navigating to Google’s homepage. That means AI Mode’s reach is limited today, likely because AI Mode does not monetize at the same rate as Search.

My experience: AI Mode turns a search into a conversational Q&A, much like ChatGPT. For example, a query such as “I want to buy a car” yields an AI bullet point list of factors to consider (budget, fuel efficiency, reviews, etc.) and allows follow-up questions in context. Similarly, asking “cheapest home insurance?” produces an AI summary of top insurers and tips to lower premiums. Notably, I found traditional blue-link ads to be either nonexistent or much less prominent in these AI responses. Initially, I saw no obvious ad at all, just the AI’s advice (in my tests, the AI Mode results were mostly informational with minimal direct advertisement).

This gets us back to the concern over the past six months: does Google have to choose between a great user experience and ad revenue? Google was clear in late May—they’re already experimenting with ads in AI Mode, with Search and Shopping ads being injected into the AI answers for commercial queries. For instance, the AI answer about “buying a car” could include a labeled ad for an auto dealership or loan offer, blended into the chat response.

It’s early days, but Google clearly intends to carry over its ad business model into the AI interface, and hearing that on a call will be a welcome reiteration for investors.

My question is: how effective will these AI-native ads be? Will users scroll and click them as often as they did with classic search ads? Or will the AI suffice, reducing the need to click anything? This remains the critical area to watch.

From a strategic perspective, AI Mode could be Google’s masterstroke to solve the AI disruption dilemma: it leverages Google’s massive data advantage (index + usage) to provide answers like ChatGPT, while still keeping users in the Google ecosystem. If Google can monetize AI answers nearly as well as the 10 ads on a classic search page, it will have maintained the best of both worlds. On the other hand, if users flock to AI Mode but Google struggles to serve or monetize ads there, it will pressure long-term revenue.

OpenAI’s ChatGPT continues to rapidly improve and expand, underscoring why Google is sprinting on AI. ChatGPT’s usage growth has been unprecedented: as of February 2025, it had 400 million weekly active users, and by mid-2025 OpenAI’s CEO hinted it may be approaching 800 million to 1 billion WAUs globally.

For context, that’s about 10% of the world using ChatGPT weekly. The pace of ChatGPT’s updates has also been fast, releasing new features in quick succession. For example, in a one-month span from mid-June to mid-July, they rolled out “Record Mode” for meetings (June 18), chat search connectors for cloud drives (June 24), and in mid-July, voice upgrades and app features on consecutive days. This real-time upgrade cycle, measured in days, not months, highlights how fast AI products are evolving.

OpenAI also disclosed that ChatGPT handles over 2.5 billion prompts per day. For comparison, I estimate Google handles about 12–15 billion searches daily (3 billion daily users with 4–5 searches each). I’ve seen estimates ranging from 6 billion to 25 billion per day. Either way, ChatGPT is growing fast but still small in the context of Google’s search dominance. Of those ChatGPT prompts, only 13% (about 330 million per day) come from the U.S., with the rest spread worldwide. India is a major user base. This global traction means OpenAI is not just a U.S. novelty,  it’s becoming an international tool.

My take: ChatGPT’s momentum illustrates the competitive stakes. Google’s search moat is still intact (it retains 83% share of global search volume), but ChatGPT’s growth shows users will embrace new products. Google is essentially choosing to disrupt itself before others do.

Grok
Early in July, xAI released Grok 4, calling it “the most intelligent model in the world,” with search integration and 100× more training data than Grok 2. It achieved top benchmarks, including 88% on GPQA Diamond. So the intelligence of the model is impressive. As for usage, my guess is that weekly active users are around 20 million, far below the estimated 800 million that GPT will soon reach. That said, API demand was overwhelming, leading xAI to increase rate limits on July 15.

Disclaimer