Cloud Margins Double, AI Overviews Monetize at Parity, and the Bear Case Loses Its Best Argument — For Now

Results vs. Consensus

Quality of the Beat

• Revenue: Genuine and broad-based — Search, YouTube ads, subscriptions, and Cloud each grew double digits. FX was a 2-point headwind, so constant-currency growth was 14%, accelerating from Q4 2024's 12%.
• Operating margin: 230bps of expansion is high quality — driven by revenue mix (Search growing faster than higher-TAC Network), moderated comp growth, and operating leverage despite a >$1B Y/Y increase in depreciation. Management explicitly warned depreciation acceleration will pressure margins from here.
• EPS: The $2.81 vs. $2.01 consensus headline overstates the operational beat. $7.7B of the $34.5B in net income was a one-time unrealized gain on an investment in a private company (within the OI&E line). Stripping that out, "core" EPS was closer to ~$2.19 — still a clean beat vs. consensus, but the stat-of-the-quarter "+40% beat" is not all operational.
• YouTube ads: Light vs. consensus, but Q1 has a tough comp and Schindler called out "very solid" growth in Q1 with brand strength from cultural moments (Coachella, March Madness) and reservation-based ads "more than doubling" Y/Y. We don't read this as a deceleration signal yet.

Segment Performance

Google Services — Search Holding Up Under AI Pressure

The single most important data point in the quarter is that Search & other revenue grew 10% to $50.7B with AI Overviews now serving 1.5 billion monthly users across 15+ languages and 40 countries. Schindler reiterated the line that has become the central tenet of the Alphabet bull case post-AI: AI Overviews are monetizing "at approximately the same rate" as classic Search results. Commercial query volume is up. Multimodal queries (Lens, Circle to Search) are up materially — Circle to Search alone is on 250M+ devices with usage +40% Q/Q.

"For AI Overviews, overall, we continue to see monetization at approximately the same rate, which gives us a strong base on which we can innovate even more." — Philipp Schindler, Chief Business Officer

Vertical strength was led by financial services (insurance) and retail, with health and travel "sizable contributors." Schindler also flagged that the de minimis exemption changes will be a "slight headwind" to ads in 2025, primarily from APAC retailers — but this is a known, contained risk.

YouTube continues to compound: 1B+ monthly podcast users, 125M+ Music & Premium subscribers, watch-time growth strongest in Shorts and living room (CTV). Reservation-based ads more than doubled Y/Y.

Assessment: The quarter that was supposed to show AI cannibalization didn't. We need at least two more quarters of the same data before declaring victory, but the bear case loses considerable force when the largest-ever AI feature deployment in Search history is happening with monetization parity. Subscriptions/Platforms/Devices at +19% is also quietly the highest-growth piece of Services.

Google Cloud — The Margin Story Finally Arrives

This is the segment that, in our view, justifies a constructive stance even at peak capex intensity. Cloud revenue grew 28% to $12.3B — solid but the real news is operating margin expanding from 9.4% to 17.8%, with operating income more than doubling to $2.18B. The leverage is showing up exactly where the AI infrastructure bull case said it would: scaling fleet utilization is offsetting (so far) the depreciation lift.

Pichai disclosed that demand still exceeds capacity. Ashkenazi said Cloud exited 2024 with more demand than capacity and the same was true in Q1 — meaning Cloud growth is currently capacity-constrained, not demand-constrained, and capacity comes online weighted to late 2025.

"We exited the year in cloud specifically with more customer demand than we had capacity. And that was the case this quarter as well." — Anat Ashkenazi, CFO

Product velocity remains high: Ironwood (7th-gen TPU, 10x compute uplift, ~2x power efficiency, designed for inference at scale); Vertex AI hosting 200+ models including Gemini 2.5 Pro/Flash, Imagen 3, VO2, and third-party models like Llama 4 and Anthropic; the announced Wiz acquisition for cloud security; the new Agent Development Kit and Agentspace product. Strategically, the company is positioning to be the cloud of choice for AI agents, which is the right pond to be fishing in.

Assessment: Cloud's Q1 margin print is the single most bullish operational data point in the report. If this leverage holds — and Q4 capacity additions extend it — the segment becomes a genuine second profit engine, not just a strategic option. This is the answer to "what does the AI capex actually buy you."

Other Bets / Waymo — Quietly Material

Other Bets revenue declined Y/Y to $450M (lapping a 2024 milestone payment) with operating loss of $1.23B. The standout disclosure is Waymo: 250,000+ paid passenger trips per week, up 5x year-over-year. Paid service launched in Silicon Valley this quarter; Atlanta paid launch via Uber later this summer; D.C. and Miami announced for 2026; airport access and freeway driving in active development.

"Waymo is now safely serving over a quarter of a million paid passenger trips each week. That's up 5x from a year ago." — Sundar Pichai, CEO

Assessment: Waymo is no longer a science project hidden in the Other Bets cost line — it is the only commercial-scale robotaxi service in the U.S. The 5x Y/Y growth, multi-city expansion, and Pichai's first-ever earnings-call answer to a Waymo question (he flagged this himself) suggest management is preparing to monetize the option value soon. We don't underwrite Waymo upside in our base case, but it is no longer a rounding error in the optionality stack.

Key KPIs

Key Topics & Management Commentary

Overall Management Tone: Confident and substantive. Pichai opened with a level of AI infrastructure detail (2M miles of fiber, 33 subsea cables, Ironwood TPU specs, model leaderboard data) that signals a posture of "we are not playing catch-up, we are setting the pace." Ashkenazi was disciplined on the depreciation drag but visibly comfortable with the margin profile. No defensive posture on AI cannibalization yet — that question simply was not pressed hard, which is itself a tell about where the consensus sits going into the print.

The AI-vs-Search Cannibalization Thesis

The overhang on Alphabet for the past 18 months has been a simple thesis: as users get conversational AI answers (in ChatGPT, in Perplexity, or in AI Overviews themselves), they will click less, advertisers will pay less per query, and the cash cow withers. Q1 is not the final answer to that thesis, but it is meaningful evidence against it. AI Overviews is now serving 1.5 billion monthly users and Search & other still grew 10% with monetization at parity. Pichai's framing is that AI is expanding what Search can answer rather than substituting away from it — and 5T+ annual queries with rising commercial query volume is consistent with that.

"With the launch of AI Overviews, the volume of commercial queries has increased. Q1 marked our largest expansion to date for AI Overviews, both in terms of launching to new users and providing responses for more questions." — Philipp Schindler, Chief Business Officer

Assessment: This thesis cannot be settled in one quarter. But the prior (a deceleration as AI Overviews scaled) didn't materialize, and the new product (AI Mode) is generating queries 2x longer than traditional search — which is at minimum engagement-positive. We treat this as a thesis-supportive print and adjust our prior on cannibalization probability downward.

The Capex Cycle and the Depreciation Wave

$75B is the announced 2025 capex number; Q1 was $17.2B. Ashkenazi flagged that the depreciation growth rate, already 31% Y/Y in Q1, will accelerate through 2025. This is the structural P&L pressure that bears point to: even if revenue growth holds, EPS could be capped by the back-end of the depreciation curve.

"The first quarter, we saw 31% year-on-year growth in depreciation from the increase in technical infrastructure assets placed in service. Given the increase in CapEx investments over the past few years, we expect the growth rate in depreciation to accelerate throughout 2025." — Anat Ashkenazi, CFO

Assessment: The depreciation wave is the single biggest near-term P&L risk. But Q1 shows that operating leverage in Cloud, mix shift toward lower-TAC Search, and headcount discipline can absorb it for now. The risk is later quarters — the back half of 2025 — when capacity comes online and depreciation hits without a matching revenue ramp.

Gemini Ecosystem & Model Position

Pichai's framing of Gemini 2.5 as "the best model in the industry" is now backed by leaderboard data — top of Chatbot Arena by a significant margin. AI Studio + Gemini API active users are up 200% YTD. All 15 of Alphabet's products with 500M+ users now use Gemini models. Gemma open models have been downloaded 140M+ times. The Gemini app at 35M DAUs (per disclosed trial data) trails ChatGPT meaningfully — Pichai didn't dispute that, but pointed to the AI Overviews user base of 1.5B as the more relevant distribution figure.

Assessment: The model gap is closing or closed. The distribution gap (Gemini app vs. ChatGPT consumer) remains but matters less than bears have made it — Alphabet's distribution moats are Search, Android, Workspace, YouTube, and Cloud, and Gemini is being deployed across all of them.

Wiz Acquisition

Pichai reiterated the announced intent to acquire cloud security firm Wiz. The deal is positioned as multi-cloud security infrastructure that incentivizes more multi-cloud computing — strategically smart given that Wiz's value is precisely that it works across AWS, Azure, and GCP, so Alphabet would acquire scale across competitors' clouds. Regulatory risk on the deal is real but not in this quarter's signal set.

Capital Return Reset

$70B additional buyback authorization on top of the existing program; quarterly dividend up 5% to $0.21. Q1 repurchases were $15.1B; trailing pace suggests $55-60B annualized buyback, plus ~$10B dividend, against $74.9B TTM FCF. Even with peak capex, Alphabet is returning roughly all of free cash flow to shareholders.

Assessment: The capital return signal is bullish. Companies that genuinely doubt their long-term cash generation don't add $70B to a buyback during peak investment.

Guidance & Outlook

Alphabet does not provide formal numerical guidance, but management's color was specific:

• 2025 capex: Reaffirmed at approximately $75B.
• Q2 / 2025 ads: Lapping 2024 financial-services strength will be a Y/Y comp headwind. De minimis exemption changes a slight ads headwind from APAC retailers.
• Cloud: Tight demand/supply through the year; revenue growth rates may be uneven quarter-to-quarter depending on capacity deployment, with relatively higher capacity coming online toward the end of 2025.
• Depreciation: Will accelerate through the year from the 31% Y/Y Q1 rate.
• Headcount: Modest growth expected in key investment areas; Q1 SBC is structurally lower than later quarters due to award timing.

Implied Q-over-Q ramp: With Q1 at $90.2B and a tougher Q2 ad comp, plus FX dynamics, mid-teens revenue growth on a constant-currency basis through 2025 looks achievable but Q2 may show the lowest growth rate of the year before reaccelerating into Q4 as Cloud capacity comes online.

Analyst Q&A Highlights

AI & Search Monetization

• Brian Nowak (Morgan Stanley): Asked about commercial query trends and the AI products driving them. Schindler pointed to AI Overviews as the primary driver, with commercial query volume up since launch and Q1 the largest expansion to date.
• Doug Anmuth (JPMorgan): Pressed on AI Overviews monetization mechanics — click-through rates, conversion. Schindler held the line on "monetization at approximately the same rate" and declined to disclose specifics.
  Assessment: The non-disclosure on CTR/conversion is itself a tell — likely some shift in click behavior even if revenue per query is flat. Worth pressing in subsequent quarters.
• Ken Gawrelski (Wells Fargo): Asked whether AI Overviews, AI Mode, and Gemini app would converge to one product or remain distinct. Pichai's framing: Search and Gemini will remain distinct surfaces; AI Mode is "the tip of the tree" pushing forward, with learnings flowing into AI Overviews for the broader user base.

Capital Allocation & Margin Resilience

• Mark Mahaney (Evercore): Noted multi-year-high margins in both Services and Cloud and asked whether enough levers remain to offset rising depreciation. Ashkenazi said productivity is "continuous, not episodic," but acknowledged depreciation pressure will become "more difficult" to fully offset later in the year.
  Assessment: The honest answer. Margins likely peak in H1 2025.
• Eric Sheridan (Goldman): Asked about flexibility in 2025 investment levels if macro deteriorates. Ashkenazi reaffirmed $75B capex and framed investment as serving long-term growth, but noted continuous focus on efficiency.

Waymo

• Mark Mahaney (Evercore): Asked whether Alphabet is committing to a long-term business model (licensing vs. operating fleet) for Waymo. Pichai declined to commit, framed the answer around "building the best driver" and a flexible mix of partnerships, OEM deals, and potential personal-ownership optionality.
  Notable that Pichai flagged this as the first-ever Waymo question on an earnings call — management may be loosening up to discuss commercial trajectory more in coming quarters.

What They're NOT Saying

1. AI Overviews CTR/conversion specifics: Repeatedly declined to disclose. "Monetization at approximately the same rate" leaves room for click-through-rate compression offset by query mix or ad density. Investors should assume the disclosed framing is the most flattering version.
2. Search ad pricing trajectory: No specific commentary on cost-per-click trends. With AI Overviews at the top of more SERPs, blended pricing dynamics have likely shifted — but the company is not disclosing them.
3. Gemini consumer DAUs: Pichai didn't quote a specific Gemini app DAU number — only the indirect reference to AI Overviews' 1.5B reach. The 35M DAU figure disclosed in court filings is the number, and it materially trails ChatGPT.
4. DOJ Search remedies: No mention of the ongoing DOJ remedies phase post-monopoly ruling. Chrome divestment is the worst-case scenario being floated; management chose not to address it on the call.
5. Cloud customer concentration / contract durations: No commentary on RPO / backlog / multi-year contract mix. With Cloud growth tied to capacity deployment timing, this is something the Street will increasingly want quantified.
6. Margin guide for 2025: Ashkenazi acknowledged the depreciation pressure but did not commit to a margin trajectory. The most likely interpretation: Q2-Q4 margins compress modestly from the Q1 33.9% peak.

Market Reaction

• After-hours move (April 24, 2025): +5%+ on the print.
• Analyst posture: Predominantly constructive — the read on the print is "AI Overviews didn't break Search, Cloud margins doubled, capital return was raised" — with capex/depreciation framed as the core risk to monitor.

The stock reaction is a pure relief rally. Going into the print, the market was positioned for evidence of AI cannibalization in Search; the print provided the opposite. Combined with Cloud margin expansion and the buyback authorization, the response is rational.

Street Perspective

Debate: Is AI Overviews monetizing at parity, or is the framing obscuring CTR compression?

Bull view: Schindler's "approximately the same rate" with disclosed commercial query growth is direct CFO-level testimony against the cannibalization thesis. With 1.5B MAU, the sample size is no longer small.

Bear view: "Approximately the same rate" of monetization at the surface level can mask falling click-through rates if commercial query volume is rising faster — i.e., revenue per user is stable while revenue per query is falling. The non-disclosure on CTR/conversion is suspicious.

Our take: Bulls have the better near-term argument. CTR compression that's offset by query expansion is, economically, still a win for Alphabet — and the 10% Search growth speaks for itself. Worth re-litigating quarterly.

Debate: Does Cloud margin expansion sustain through the back half?

Bull view: Q1's 17.8% margin (vs. 9.4% Y/Y) shows the operating leverage thesis is real. As capacity scales and depreciation per dollar of revenue normalizes, margins should keep climbing toward the 25-30% range over multiple years.

Bear view: Q1 benefited from favorable mix (high-margin AI Solutions, Workspace ARPU) and pre-deployment capacity utilization. Once Q4's capacity adds come online with depreciation but lagging revenue, margins compress.

Our take: Both are partially right. We expect Cloud margins to be lumpy quarter-to-quarter but to trend higher over the next 2-3 years if AI demand continues. Q1 is unlikely to be the peak.

Debate: Is the capex peak coming, or are we in early innings?

Bull view: $75B is the right number for the AI moment. Demand exceeds capacity in Cloud; the alternative is ceding share to AWS/Azure during the most important platform shift in 20 years.

Bear view: Hyperscaler capex inflates depreciation faster than revenue can scale, particularly if AI Overviews' "monetization at parity" doesn't fully hold. The OPEN question is what 2026 capex looks like — if it's $90B+, the depreciation curve gets ugly.

Our take: Bears are right that 2026 capex is the bigger story than 2025 — and management was conspicuously silent on 2026. The risk asymmetry tilts negative if 2026 capex steps up another 15-20%.

Model Update Needed

Valuation impact: Modestly positive. Cloud margin path is the bigger swing factor than revenue; if 17%+ Cloud margins prove durable, segment value rises meaningfully. Offset by depreciation drag and 2026 capex uncertainty.

Thesis Scorecard Post-Earnings

Overall: Thesis strengthened. The single most important variable in the bull case (Search monetization through AI) printed in line with the bull narrative; the second-most-important variable (Cloud leverage) printed materially better than expected.

Action: Buy. We initiate Outperform on Alphabet at this print.