X-Ray AnalysesCommunication ServicesAlphabet Inc.
G

Alphabet Inc.

NASDAQ: GOOGL·Internet Content & Information·United States
Price at analysis
$357.37
▲ 1.1% · 52-wk range $173–$409
◆ The Buffett Lens
◆ Educational analysis & opinion — not investment advice. Figures as of 30 June 2026. See full disclaimer below.
The Scorecard · one-second read
Moat
8
Management & Capital
8
Financial Strength
10
Growth
8
Valuation
6
◆ Type · CompounderBusiness · Wide moat, contestedDividend · Token (buybacks)
8.0
"A wonderful franchise — contested for the first time in 20 years."
Search scale · YouTube · a top-tier AI lab · Waymo optionality — but the AI-search question and a $185B capex bet are live · fair price
Part I

The business, in plain English

What it does · how the money flows · why it matters

Charlie Munger once said missing this company was one of the great mistakes of our careers — that we sat there "sucking our thumbs" while our own insurer, GEICO, paid Google ten dollars a click for something that cost Google almost nothing. So let me start with humility. Alphabet is a toll-booth on the world's information and advertising. When two billion people go looking for something, Google shows them answers — and, alongside, an advertisement that a business paid to place there. That auction, running billions of times a day at almost no marginal cost, is one of the greatest money-machines ever built.

Founded in 1998 by two Stanford students, Google organised the internet so well that its name became a verb in dozens of languages. Today, as Alphabet, it is far more than search: it owns YouTube (the world's largest video platform), Android and Chrome (which carry its search to billions of devices), a fast-growing and now-profitable Cloud business, a world-class AI lab in DeepMind, and Waymo, the clear leader in self-driving cars. In the year ended December 2025 it booked $403 billion in revenue and $132 billion in profit, and it is worth about $4.32 trillion.

Revenue, 5 years
$183B → $403B
FY2020 → FY2025 · ~17%/yr
Net income, 5 years
$40B → $132B
~3.3×
$1,000 at the 2004 IPO (approx.)
≈ $85,000
an ~84× compound, split-adjusted

"You could see in our own operations how well that Google advertising was working. And we just sat there sucking our thumbs. So we're ashamed. We atone." — Charlie Munger, on missing Google

The reason Munger was ashamed is the reason this is a franchise: an advertising auction has almost no marginal cost. When Google shows one more ad, it costs the company essentially nothing, so nearly every extra dollar of ad revenue falls straight to profit. That is why Google earns operating margins above 30% at a scale of hundreds of billions — and why, for twenty years, it has been closer to a royalty on human curiosity than an ordinary business.

Founded1998 · Mountain View, California (USA) · IPO 2004
Sector / IndustryCommunication Services · Internet Content & Information
CEOSundar Pichai (Alphabet CEO since 2019)
Makes money fromAdvertising (~73%), Cloud (~15%), subscriptions/devices
Revenue (FY2025 · TTM)$403 B · ~$423 B
Market capitalisation~$4.32 T · newly a Dow-30 member
Part II

The history — from a dorm room to the Dow

How a search box became one of the world's great franchises

The road matters, because it shows both a company of extraordinary technical vision and one that has, more than once, been caught flat-footed and recovered — most recently by artificial intelligence.

YearMilestone
1998Larry Page & Sergey Brin found Google — 'PageRank' ranks pages by links, not keywords. A better search is born.
2000–03AdWords + the pay-per-click auction (bid × quality) — the economic engine of the whole company. AdSense extends it to the web.
2004IPO by Dutch auction. Dual-class shares from day one lock in founder control.
2005–08Buys Android (~$50M — a legendary deal) and YouTube (~$1.65B); launches Chrome. The distribution empire is built.
2014Buys DeepMind. With Google Brain, invents the Transformer (2017) — the 'T' in GPT — a decade before the LLM boom.
2015Reorganises into Alphabet; splits the cash machine (Google) from the moonshots (Other Bets). Sundar Pichai runs Google.
2022–23ChatGPT triggers an internal 'code red.' Bard's flubbed demo wipes ~$100B off the market cap in a day.
2024–26Gemini reaches frontier parity; AI Overviews reach ~2B users; Cloud turns profitable; Alphabet joins the Dow.

Two lessons for an owner. First, the capital allocation is among the best in history: $50 million for Android and $1.65 billion for YouTube may be the two finest acquisitions ever made. Second — and this is the caution — Google was genuinely blindsided by ChatGPT, and its first answer was a public embarrassment. It recovered impressively (Gemini is now a frontier model). But the episode proved that even this fortress can be surprised, and the surprise came at the one place it could least afford: the front door of search.

Part III

The circle of competence

How it works — and how confidently you can know its future

The what is beautifully simple. The ten-year picture is, for the first time in two decades, genuinely uncertain. Here is the machine:

01
Organise the world's information
Search, YouTube, Maps, Android, Chrome.
02
Attract billions of users
~90% of search; a verb in every language.
03
Auction their attention
Advertisers bid, in real time, per click.
04
Serve the ad at ~zero marginal cost
Almost every extra dollar is profit.
05
Reinvest in Cloud, AI & moonshots
Gemini, Waymo, the next platform.
How hard is it to understand?
Understandable, but in flux · 3/5 — you grasp an ad auction as easily as a lemonade stand. What you cannot yet know is whether AI answers monetise as well as ten blue links (Part VI). Simpler than Nvidia (2/5), less certain than Coca-Cola (5/5) — sitting, fittingly, right beside Microsoft.

This is the honest heart of the matter. For twenty years Google's future was as predictable as Coca-Cola's — you could see the search box being there in a decade. Generative AI has, for the first time, put a real question mark over how people will find information and how well that will pay. The business is understandable; its next chapter is not yet written. That is precisely the tension a careful owner must price.

What you must believe to own it
  • AI search monetises roughly like classic search — Google's most important claim; if true, the franchise survives the transition intact.
  • Distribution wins — Chrome, Android and the defaults let Google push Gemini to billions faster than rivals can build reach.
  • The $185B capex era earns its keep — the AI infrastructure bill produces adequate returns on a vastly larger capital base.
Part IV

How it makes money

Revenue by line · geography · the ad auction

Three quarters of Alphabet is still advertising, but the shape is changing — Cloud is now a real second engine, and subscriptions are a fast-growing third. FY2025 revenue by line:

Google Search & other ads~56%
~$224B — the core money-machine, and the AI battleground.
Google Cloud (GCP + Workspace)~15%
~$59B, growing ~48%/yr — now profitable. The bright spot.
Subscriptions, platforms & devices~12%
~$48B — YouTube Premium/TV, Google One, Play, Pixel.
YouTube ads~10%
~$40B — Shorts + connected-TV. A crown jewel.
Google Network ads~7%
~$30B — the shrinking, antitrust-targeted ad line.
Where in the world the money comes from
United States~48%
EMEA ~29% · Asia-Pacific ~17% · Other Americas ~6%~52%
A truly global franchise — nearly half the money comes from outside the US.

How the auction works. Every time Google shows ads, it runs a real-time auction: advertisers bid a maximum per click, and Google ranks them by bid × quality (how relevant and likely-to-be-clicked the ad is), with the winner usually paying just enough to beat the next bidder. Because showing one more ad costs Google almost nothing, this is a business of extraordinary operating leverage — Google Services earns ~42% operating margins, and those fat profits pay for Cloud's build-out, Waymo's losses, and the enormous AI capex. The whole edifice rests on that one, magnificent, near-zero-marginal-cost auction.

Part V

The moat

Wide and deep — but, for the first time, under attack

Google's moat is one of the widest ever built. It is filled from five springs:

Scale & data network effects. More searches → more data → better results and targeting → more users and advertisers → more data. This flywheel has held Google near ~90% of search for twenty years.

A two-sided marketplace. Billions of users on one side, millions of advertisers on the other — neither can easily leave the other.

Distribution. Google pays ~$20 billion a year to be the default search in Apple's Safari, and owns Chrome (~65% of browsers) and Android (~70% of phones). Most people never change a default — so owning the default is worth a fortune.

Brand. "To google" is the verb for the whole category — mindshare no rival can buy.

YouTube & Cloud stickiness. A creator-viewer network no one can rebuild, plus real switching costs in Workspace and Cloud.

The honest caveat. The distribution, data and brand springs are structurally durable. But the product itself — the ten-blue-links page — is, for the first time, under genuine attack from AI answer-engines. So I score the moat an 8, not a 10: still vast, but contested at the front door in a way it has never been before.

Part VI

The AI question — is AI eating search?

The single most important issue in the whole analysis

Everything turns on this. When an AI chatbot simply answers your question, you may never see — or click — the ten blue links and the ads beside them. Does that break Google's money-machine, or does Google simply become the best answer-engine and keep the tolls? Here are both sides, honestly.

The bear caseThe bull case
AI answers replace the links — and the ads beside themGoogle is a top-tier AI lab — Gemini reached frontier parity; it invented the Transformer
Fewer searches per task — searches/user reportedly ~-20%AI Overviews reach ~2B users; management says they monetise at ~parity with search
One AI answer shows far less ad inventory than a results pageCustom TPU chips let Google serve AI cheaper than rivals renting Nvidia
Innovator's dilemma — cannibalising a 90%-margin cash cowUnmatched distribution — Chrome/Android push Gemini to billions instantly

The single most important data point, if it holds, is Google's claim that AI Overviews monetise at roughly the same rate as classic search. If that is true, the franchise survives the transition with its distribution intact — Google simply changes the shape of the answer and keeps collecting the toll. If instead AI answers structurally monetise worse — one paragraph where there used to be a page of high-intent ads — then Google could "win" the technology and still lose margin and revenue for years. That is the irreducible judgment call, and no one — not management, not I — can yet prove it either way. It is the reason this business, uniquely among the great franchises, cannot be bought with total confidence today.

Part VII

The competition

Search & AI · advertising · cloud · autonomy

ArenaWhoWhere Google stands
Search / AIOpenAI/ChatGPT, MS Copilot, Perplexity, Anthropic#1 but contested — Gemini now at parity
AdvertisingMeta (social), Amazon (retail media)Co-leads the duopoly; Amazon nibbling
CloudAWS #1, Azure #2 — Google #3#3 but fastest-growing (~48%), now profitable
Video / attentionTikTok, Netflix, Meta ReelsYouTube dominant, esp. on the TV
AutonomyTesla robotaxi (Cruise exited)Waymo the clear leader — ~$126B value

Read that table the right way. In its core — search and advertising — Google faces the most serious competition in its history, but from a position of overwhelming strength. In cloud, it is only #3, but it is the fastest-growing of the three and has finally turned profitable, with its AI stack (Gemini + TPUs + DeepMind) as the differentiator. And in Waymo it holds a genuinely commanding lead in what could be an enormous future market — an option most owners get almost for free inside the $4T mothership. Few companies fight on so many fronts; fewer still are winning or gaining on nearly all of them.

Part VIII

Management, ownership & control

Who runs it · who really controls it

Able managers — and a governance structure every investor must understand before buying a single share.

S
Sundar Pichai · CEO, Alphabet
CEO of Google since 2015, of Alphabet since 2019. A steady, product-deep operator who navigated the ChatGPT 'code red' from panic to a credible frontier position, and turned Cloud from a cash drain into a profit engine.
A
Anat Ashkenazi · Chief Financial Officer
CFO since 2024 (from Eli Lilly). Steering the largest capital-spending ramp in the company's history alongside the first-ever dividend and continued buybacks.
Skin in the game — and the control lock
Founder voting control
~51%
Page & Brin control ~51% of votes with only ~12% of the equity, via 10-vote Class B shares. Brin alone holds ~5.8%. An unassailable, owner-operator control lock.
Institutions
~65%+
Vanguard ~7% · BlackRock ~7% · Fidelity. Insider trades this period: routine vesting/tax, no signal.

A Buffett-style investor reads the dual-class structure two ways. On the one hand, it lets the founders make very long-horizon, unpopular bets — Waymo, the vast AI capex — without being second-guessed by the market each quarter, and their record earns real trust. On the other, public shareholders have essentially no governance recourse if that judgment ever goes wrong. You are, in plain terms, a passenger — riding with owner-operators who have been superb, but a passenger nonetheless. Know that before you board.

Part IX

The numbers

A gusher of cash · and a capex bill that hides it

MetricValueRead
Revenue growth (5-yr CAGR)~17%▲ fast for the size
Net income (FY2020 → FY2025)$40B → $132B▲ ~3.3×
Gross margin~60%▲ high
Operating margin~32%▲ Services alone ~42%
Return on equity (ROE)39%▲ excellent
Return on invested capital (ROIC)19%▲ well above cost of capital
Balance sheetNet cash · Altman-Z 13.9▲ fortress
Capex / revenue~26%▲ $91B FY25 → ~$185B guided FY26
Free cash flow / share vs EPS$5.33 vs $13.24◆ FCF crushed by the AI build

The core is a money-gusher — 39% on equity, near-zero debt, tens of billions in profit a quarter. But note the same asterisk as Microsoft, only bigger: free cash flow ($5.33/share) is a fraction of reported earnings ($13.24), because Alphabet is pouring an astonishing ~$185 billion into AI data centres this year — and Mr. Pichai says even that "won't be enough." Almost none of that is maintenance; it is a colossal growth bet. Whether it earns an adequate return on so vast a capital base is the second great unknown here — and it is why the free-cash-flow figures, and the models built on them, are so badly distorted right now (Part X).

Part X

Valuation

Cheap on earnings, 'expensive' on cash — and why the DCF lies

YardstickTodayForwardRead
P/E — reported earnings~27x~25x (FY26) · ~21x (FY28)fair for the quality & growth
P / Free cash flow67xdistorted — capex-crushed
EV / EBITDA20xreasonable
Price / Sales10.2xfull but not extreme
PEG (trailing)0.59growth-adjusted, attractive
Why the models violently disagree
Reported EPS
$13.24
→ P/E ~27×
Owner earnings ◆
the honest middle
profit less maintenance capex — well above FCF, well below EPS
Free cash flow / share
$5.33
→ P/FCF ~67× (capex bulge)

Here is a genuinely instructive mess. A naive discounted-cash-flow model spits out an intrinsic value of about $132 — implying the stock is 63% overvalued. Ignore it. That model treats today's $185 billion of growth capex as a permanent expense, crushing free cash flow and the value built on it — it is artificially, absurdly low. Wall Street's analysts, looking through the capex to earnings, land near $413 (+16%). The truth sits between, and closer to the analysts: on ~27× earnings — even ~25× next year's — for a wide-moat franchise still growing in the mid-teens, with Cloud inflecting and Waymo thrown in, Google is fairly-to-attractively priced. This is the mirror image of the free-cash-flow trap: the cheapest way to misjudge Alphabet today is to trust its free cash flow.

PRICE vs. VALUE — why the two models are $280 apart
Naive DCF ~$132
Price $357
Analysts $413
◀ DCF (capex-distorted)Analysts ▶
The ~$280 gap is the story: the DCF (~$132) is artificially low because it capitalises the AI-capex bulge as permanent; the analysts (~$413, +16%) look through to earnings. On ~27× earnings the franchise is fairly valued — the margin of safety widens meaningfully toward $300. → Interactive DCF model
Part XI

Antitrust & other risks

The courts · the AI question · the capex bet

🤖 AI answer-engines eroding search monetisation⚖ DOJ search-monopoly case (data-sharing; appeals)⚖ DOJ ad-tech case (remedies pending)🇪🇺 EU fines (~€11B+ cumulative)💸 ~$185B AI capex — returns unproven🗳 Dual-class control — no shareholder recourse

The two biggest risks are the two open questions: does AI-answer search monetise like classic search (Part VI), and will the $185B capex earn its keep (Part IX). On antitrust, the news is better than the headlines: in the DOJ search case a judge ruled Google an illegal monopolist (2024) but in September 2025 declined to break off Chrome or Android — ordering instead an end to exclusive defaults and some data-sharing (both under appeal into 2026). A separate ad-tech case and ~€11B of EU fines sting the cash flow but do not threaten the franchise. The honest summary: the existential legal risk has receded; the AI and capex questions have not.

PART XII · To our shareholders
The Letter

Let me begin with a confession, because it frames everything that follows. Charlie and I missed this company — badly — and we knew better. Our own insurer was paying Google handsomely for advertising that cost Google almost nothing, and we sat on our hands while a great franchise compounded in plain sight. I have carried that lesson for twenty years, and I bring it to the page today so that I do not make the opposite error — dismissing a wonderful business because it now wears a fashionable worry.

Because a wonderful business is exactly what this is. Alphabet is a toll-booth on the world's information and advertising, with a search franchise that has held ninety percent share for two decades, the largest video platform ever built, a browser and a phone operating system that carry its products to billions, a cloud business that has finally turned profitable and grows near fifty percent a year, one of the two or three best AI laboratories on Earth, and — almost as a free option — the clear leader in self-driving cars. It earns thirty-nine percent on equity, carries a fortress balance sheet, and is run by owner-operators with a capital-allocation record — Android for fifty million dollars, YouTube for a billion and a half — that belongs in a textbook. On quality, this is a franchise of the very first rank.

And unlike the chip-maker we looked at recently, I can broadly understand where it earns its keep: an advertising auction is no harder to grasp than a lemonade stand, and it is one of the finest business models ever devised. Nor is it priced for the heavens — at about twenty-seven times earnings, even twenty-five times next year's, this trades like an ordinary blue chip, not a mania. I would warn you, in passing, to ignore the discounted-cash-flow model that screams the stock is worth a third of its price: it is fooled by the enormous, temporary bulge of artificial-intelligence spending, treating a growth bet as if it were a permanent cost. On the earnings a franchise like this reliably produces, the price is fair — and toward three hundred dollars, genuinely attractive.

So why do I stop short of pounding the table? Because for the first time in twenty years, the front door of the castle is contested. When a person asks a machine a question and simply receives an answer, they may never see the ten blue links — or the advertisements beside them — that have paid for all of this. Google's own claim is that its AI answers monetise about as well as its old results page; if that is true, the franchise sails through, and this letter will read as far too cautious. But if one synthesised paragraph structurally earns less than a page of high-intent ads, then Google may win the technology and still lose margin for years — the classic dilemma of cannibalising your own cash cow. I cannot yet prove which way it breaks, and neither can anyone else. Add a hundred-and-eighty-five-billion-dollar annual capital bill of unproven return, and a control structure that makes you a passenger rather than a director, and you have real, unresolved questions sitting beneath a magnificent business.

What, then, would I do? I would own it — watchfully. This is not the chip-maker in the "too hard" pile; it is a franchise I understand, at a fair price, with two honest question marks I can monitor quarter by quarter — search revenue per query, and the returns on that capex. I would take a position here and let it compound, size it for the uncertainty rather than bet the farm, and add with real conviction on any weakness toward three hundred dollars, where the margin of safety opens and Mr. Market hands me the doubt at a discount. Having once sat sucking my thumb while this very company ran away from me, I do not intend to make the same mistake twice — but nor will I pretend the question at its front door isn't real.

With humility, and having learned my lesson,— The Dividend Line Desk
The Bull Case
A toll-booth franchise — ~90% search share for two decades, YouTube, unmatched distribution (Chrome/Android), and a brand that became a verb.
Fairly priced for the quality — ~27× earnings (PEG 0.59) for a mid-teens grower, with Cloud inflecting to profit and Waymo as a near-free option.
A top-tier AI hand — Gemini at frontier parity, custom TPUs, DeepMind, and the distribution to push AI to billions instantly. It may well win the transition.
The Bear Case
The core is contested — AI answer-engines could disintermediate search; if AI monetises worse than links, Google wins the tech and loses margin.
A $185B capex bet of unproven return crushes free cash flow today and rests on a vastly larger capital base earning adequate returns tomorrow.
You're a passenger — founders control ~51% of votes with ~12% of the equity; plus ongoing antitrust (data-sharing remedies, ad-tech case, EU fines).
Own,
Watchfully
A wonderful franchise I understand, at a fair price — not the 'too hard' pile. Take a position and let it compound, sized for the AI question; add with conviction on weakness toward ~$300, where the margin of safety opens.
Want to be alerted if GOOGL dips toward the buy zone? Add the $300 price trigger to your Watchlist.
The Buffett Lens · Dividend Line Research · As of 30 Jun 2026 · Price $357.37
Disclaimer: This analysis is educational opinion, not personalised financial advice or a recommendation to buy or sell. Figures reflect 30 Jun 2026 and may be out of date; items marked “approx.” are estimates. Some legal matters (antitrust remedies, appeals) were unresolved at the time of writing. Do your own research and, where appropriate, consult a licensed professional before making any investment decision.
Dividend Line · X-Ray Analyses — written in the house methodology