← Back to All Research
Commission a Report
| Metric | Value |
|---|---|
| FY2025 Revenue | $402.8B (+15.1% YoY) |
| Q4 2025 Cloud Revenue | $17.7B (+48% YoY) |
| Q4 2025 Search Revenue | $63.1B (+17% YoY) |
| YouTube FY2025 Revenue | $60B+ |
| Q4 2025 EPS (GAAP) | $2.82 (+31% YoY) |
| FY2026 CapEx Guidance | $175–185B |
| Forward P/E (FY2026E) | ~28x |
| MAG7 Peer Median Fwd P/E | ~33x |
| GOOGL Discount to Peers | ~–15% |
Alphabet is the default AI infrastructure layer for the largest addressable market in technology. PRZC Research's proprietary AI market segmentation framework identifies three distinct competitive strata:
Google dominates the largest stratum by volume and wins on cost. Apple's selection of Google Gemini to power Apple Intelligence across 2 billion iOS devices — confirmed January 2026, estimated value $1B/year or $5B total — is the most decisive third-party endorsement in AI competitive history. The world's most valuable company chose Google as its commodity AI infrastructure layer.
The structural engine is Google's TPU silicon. Committed-use TPU v6e at $0.39/chip-hour versus NVIDIA B200 equivalents at ~$5.50/GPU-hour is a 4–5x cost advantage that compounds at the world's largest AI inference scale.
PRZC Research leads with the bear case (CBOM principle).
Google Search (~$198B estimated FY2025 revenue, ~49% of total) depends on advertiser clicks. AI Overviews answer queries without requiring click-through.
Measured CTR impact:
Revenue-at-risk estimate:
Key mitigant: Ads within AI Overviews rose from ~3% of results in January 2025 to ~40% by November 2025. Google is monetising the cannibalization mechanism directly. Q4 2025 Search revenue grew +17% YoY — strongest quarterly rate in years — despite AI Overviews at scale.
Revenue model conclusion: Even if AI Overviews suppress Search growth by 5 pp annually, Cloud revenue offset ($80–120B incremental over 3 years) exceeds the Search revenue loss ($30–50B) by a factor of 2–3x.
NVIDIA data centre gross margins: ~75–80%. Every competitor using NVIDIA hardware embeds this margin in their cost structure. Google does not.
| Metric | Value | Source |
|---|---|---|
| TPU v6e committed price | $0.39/chip-hour | Google Cloud posted pricing |
| NVIDIA B200 equivalent | ~$5.50/GPU-hour | Cloud reseller surveys |
| Performance-per-dollar advantage | 4.7x | Google benchmarks, corroborated |
| Ironwood (TPUv7) TCO vs. GB200 | –44% | SemiAnalysis TCO model |
| Midjourney migration savings | –65% | Public disclosure |
| SF startup (128 H100 → TPU v6e) | $340K → $89K/mo | Ainewshub case study |
Gemini Flash Lite pricing: $0.08 input / $0.30 output per million tokens — already below the sustainable cost floor for GPU-based competitors. OpenAI's equivalent tiers (~$0.15–0.40 input) require GPU economics with higher marginal costs.
FY2026 CapEx guidance: $175–185B (vs. $91.4B in FY2025). Ironwood ramp will widen the cost advantage further as capacity builds.
| Quarter | Cloud Revenue | YoY Growth | Cloud Op. Margin |
|---|---|---|---|
| Q1 2024 | $9.57B | +28.4% | ~9.4% |
| Q2 2024 | $10.35B | +28.8% | 11.3% |
| Q3 2024 | $11.35B | +34.9% | 17.1% |
| Q4 2024 | $11.96B | +30.0% | ~17.5% |
| Q1 2025 | $12.26B | +28.1% | ~18.0% |
| Q2 2025 | $13.60B | +31.4% | 20.7% |
| Q3 2025 | $15.20B | +33.9% | 23.7% |
| Q4 2025 | $17.70B | +48.0% | ~24% |
AI-specific demand signals:
Path to $100B run rate: At current trajectory, Google Cloud reaches $100B ARR within 12–18 months. At 25% operating margin: $25B annual operating profit from Cloud alone.
| Channel | Scale | Gemini Role |
|---|---|---|
| Android | 3.0B devices | Default assistant |
| Apple Intelligence | 2.2B devices | Contracted backend (Jan 2026) |
| Google Search | 1B+ daily users | AI Overviews (Gemini-powered) |
| Chrome | 3.4B users | Default sidebar/omnibox |
| Google Workspace | 3B+ users | Gemini for Workspace AI |
| YouTube | 2.7B monthly users | Summaries, search, recommendations |
| Google Maps | 1B+ users | Conversational AI planning |
| Total (non-deduplicated) | 5B+ | Default AI for >half the world |
| Metric | FY2022 | FY2023 | FY2024 | FY2025 |
|---|---|---|---|---|
| Total Revenue ($B) | 282.8 | 307.4 | 350.0 | 402.8 |
| YoY Growth | +6.1% | +8.7% | +13.8% | +15.1% |
| Google Cloud Revenue ($B) | 26.3 | 33.1 | 43.2 | 58.7 |
| Cloud YoY Growth | +37.1% | +25.9% | +30.5% | +35.9% |
| Operating Margin | 26.5% | 27.4% | 32.1% | 32.0% |
| Net Income (~$B) | 60.0 | 73.8 | 94.1 | 132.2 (est.) |
| CapEx ($B) | 31.5 | 32.3 | 52.5 | 91.4 |
| Company | Mkt Cap | Fwd P/E | Rev Growth | PRZC Rating |
|---|---|---|---|---|
| Alphabet (GOOGL) | $3.65T | 28x | +15% | STRONG BUY |
| Microsoft (MSFT) | ~$3.1T | 31x | +13% | Buy |
| Apple (AAPL) | ~$3.3T | 32x | +5% | Hold |
| NVIDIA (NVDA) | ~$2.8T | 38x | +114% | Buy |
| Amazon (AMZN) | ~$2.3T | 36x | +11% | Buy |
| Meta (META) | ~$1.6T | 25x | +21% | Buy |
| MAG7 ex-GOOGL median | — | 33x | — | — |
| GOOGL discount | — | –15% | — | — |
GOOGL trades at a ~15% discount to MAG7 median despite deeper structural AI positioning than Apple (low growth, no native AI moat) and Microsoft (enterprise exposure but paying NVIDIA rates for inference at scale).
| Scenario | Probability | 12-Mo Target | Key Assumptions |
|---|---|---|---|
| Bear | 20% | $220 | Search decelerates to 5% growth; Cloud slows to 25%; CapEx compresses FCF; antitrust risk; 22x fwd P/E |
| Base | 55% | $380 | Search holds 12–15%; Cloud at 35–40%; Apple ramps; multiple re-rates to 32x as Search fear abates |
| Bull | 25% | $480 | Cloud inflects >50%; Gemini 3.x sets frontier benchmarks; TPU advantage recognised; 38x fwd P/E |
Probability-weighted target: $375. PRZC Research price target: $420 (12-month). Current price: $302. Implied upside to target: +39%.
PRZC Research Conclusion: At 28x forward earnings — a ~15% discount to MAG7 peers — the market is pricing Search cannibalization fear without crediting Cloud acceleration, Apple deal economics, TPU cost structure, or compounding distribution advantage. As Cloud margins expand and AI Overview monetisation matures, the multiple should converge toward the MAG7 median. Rating: STRONG BUY. 12-month price target: $420.
PRZC Research | Investment Analysis Division | research@przc.re | przc.re
This report is for informational purposes only and does not constitute investment advice. © 2026 PRZC Research — Perez Capital. All rights reserved.
Commission Research
Commission a bespoke PRZC report on any company, sector, or market question. Full CBOM methodology. Delivered within 5–7 business days. £500 per report.
Commission a Report — £500Bulk credits available from £350/report · Browse all free research