Alphabet Intelligence Brief

Alphabet is growing faster at $4.14 trillion⁶ actual than it was at half the size: revenue rose 24.2% year over year to $119,796 million in Q2 2026² est, led by Google Cloud at +81.8% with operating margin expanding from 23.7% to 35.6%^2,3 est against a cloud infrastructure market growing 43%⁷ actual — so it is taking share, not riding a tide. The headline profitability is not what it appears: Q2 net income of $112,193 million² actual includes a $99,031 million mark-to-market gain the company attributes primarily to SpaceX and an unnamed private company² actual, against operating income of just $40,770 million² actual. Underneath, the capital model has inverted: buybacks went to zero from $28,306 million a year earlier² actual while the company raised $105.8 billion in six months across common equity, its first-ever preferred stock and debt² est — including a $10.0 billion private placement to Berkshire Hathaway² actual — to fund $80,598 million of half-year capex² actual that cut free cash flow to $4.3 billion^2,3 est. Meanwhile a six-year federal court judgment now compels Alphabet to share its search index with competitors at marginal cost⁵ actual.

Read every Q2 2026 profitability figure through one number. Reported diluted EPS of $9.11² actual beat consensus of $2.90 by 214%⁶ actual, and trailing P/E of 16.97⁶ actual makes Alphabet look cheaper than every megacap peer. Both are artefacts of the $99.0 billion securities gain. Excluding it and tax-effecting at the statutory rate gives roughly $2.75 est — close to what analysts actually modelled. Use the forward P/E of 22.79⁶ actual, not the trailing one.

Financial Snapshot

Revenue (TTM)

$445.9B

est FY25 $402,836m + H1-26 − H1-25; matches vendor total^3,2,6

Market cap

$4.14T

actual at $338.46, 4 Sep 2026⁶

YoY growth (Q2 26)

+24.2%

est accelerating from +15.1% in FY2025^2,3

Operating income (Q2)

$40.8B

actual vs $112.2bn net income — see callout²

Cloud backlog

$513.9B

actual of $519.5bn total, 98.9% Cloud²

Capex (H1 2026)

$80.6B

actual 35.1% of revenue²

Capital raised (H1)

$105.8B

est equity + preferred + debt²

Buybacks (H1 2026)

actual from $28,306m a year earlier²

Quarterly Performance and Segment Trend

$ millions	Q2 2025	Q1 2026	Q2 2026	Q2 YoY
Total revenues	96,428	109,896	119,796	+24.2%
Google Services	82,543	—	94,540	+14.5%
Search & other	54,190	—	63,271	+16.8%
YouTube ads	9,796	—	11,055	+12.9%
Google Network	7,354	—	7,303	−0.7%
Google Cloud	13,624	—	24,768	+81.8%
Other Bets	373	—	382	+2.4%
Income from operations	31,271	39,696	40,770	+30.4%
Google Cloud	2,826	—	8,814	+211.9%
Other income (expense), net	2,662	37,716	97,983	—
of which equity-securities gain	1,286	36,915	99,031	—
Net income	28,196	62,578	112,193	+298%
Diluted EPS ($)	2.31	5.11	9.11	—
Diluted EPS ex-gain ($) est	—	—	~2.75	—

All figures actual from the Q2 2026 10-Q² and the XBRL exhibit⁴. Q1 2026 column is from XBRL where the 10-Q does not present it. Consensus for the quarter reported 28 Oct 2026 is $3.01⁶ actual.

Annual trend

SWOT

$ millions	FY2023 as of 2023-12-31	FY2024 as of 2024-12-31	FY2025
Total revenues	307,394	350,018	402,836
YoY growth	—	+13.9% est	+15.1% est
Google Cloud	33,088	43,229	58,705
Income from operations	84,293	112,390	129,039
Net income	73,795	100,118	132,170
Capital expenditure	32,251	52,535	91,447
Operating cash flow	101,746	125,299	164,713
Free cash flow est	69,495	72,764	73,266

Strengths

Cloud grew 81.8% against a market growing 43% — taking share^2,7 est
$513.9bn Cloud backlog, roughly 20.7 quarters of current revenue² est
No customer above 10% of revenue in 2023, 2024 or 2025³ actual
Gemini app: 4.58 stars across 44.1m ratings, outscoring core Search app^11,12 actual

Weaknesses

H1 free cash flow $4.3bn, down from $73.3bn full-year 2025^2,3 est
Google Network revenue declining: $7,354m to $7,303m² actual
Buybacks halted entirely while raising $105.8bn² actualest
Enterprise reviewers cite charged troubleshooting logs and weak direct support¹⁶ actual

Opportunities

Cloud operating margin expanded 23.7% to 35.6% in six months^3,2 est
Wiz adds multicloud security to Cloud for $29.5bn² actual
Intersect gives owned renewable capacity for data centres² actual
Digital ad market forecast at 68.7% of a >$1tn 2026 total¹⁵ actual

Threats

Court-ordered search index sharing at marginal cost for six years⁵ actual
Two unresolved ad-tech antitrust cases in Texas and Virginia² actual
$5.2bn EU fine paid July 2026; $2.1bn PriceRunner charge accrued² actual
Practitioner backlash on Android verification and Chrome AI installs¹³ actual

Capital Structure and Funding History

Alphabet has no venture funding history. What matters instead is a capital model that inverted in 2026: from returning cash to raising it at scale².

Customer and Employee Sentiment

Date	Event	Amount	Counterparty / terms
H1 2025	Share repurchases	$28,306m	Open market; the prior norm² actual
Q1 2026	Senior unsecured notes, USD	$20,000m	4.80% weighted-average coupon, 15-year average maturity² actual
Q1 2026	Notes, Sterling and Swiss Franc	GBP 5.5bn / CHF 3.1bn	5.31% over 31 years; 1.06% over 10 years² actual
10 Mar 2026	Intersect acquisition	$5,900m	Renewable energy developer; consolidated as a VIE² actual
11 Mar 2026	Wiz acquisition	$29,500m	Cloud security; $8,300m intangibles; into Google Cloud segment² actual
4 Jun 2026	Underwritten public offering	$20.5bn net	29m Class A at $355.1982; 29m Class C at $351.8018² actual
4 Jun 2026	Private placement	$10.0bn net	An affiliate of Berkshire Hathaway Inc.; 14m Class A + 14m Class C² actual
Q2 2026	Mandatory convertible preferred	$19,063m net	6.25% Series A and B; listed GOOGM / GOOGN; first preferred in Alphabet's history² actual
H1 2026	Net debt proceeds	$56,226m	$20.0bn USD plus $31.8bn foreign-currency notes during 2026² actual
H1 2026	Share repurchases	$0	Complete halt² actual
H1 2026	Dividends paid	$5,231m	$0.22 per share declared in Q2, up from $0.21² actual
Total capital raised, H1 2026 est	$105,788m	Common $30,499m + preferred $19,063m + debt $56,226m²

Composite score

7.5/10

est Discounted from a ~8.6 implied by app ratings alone. Store ratings are prompted and inflated; practitioner sentiment on platform control is measurably negative.

Consumer apps

Gemini 4.58/5 — 44,098,663 ratings¹¹ actual
Google 4.04/5 — 29,423,541 ratings¹² actual

Parsed from each page's schema.org block, not a summary.

Enterprise (Google Cloud)

4.2/5

81 reviews, 97% would recommend¹⁶ actual. Recovered on rung 4 after three 403s. Small sample.

Praise themes

GKE and Kubernetes reliability for microservices¹⁶ actual
Cost-effectiveness and pay-as-you-go flexibility¹⁶ actual
Straightforward, quick setup¹⁶ actual
Gemini model quality and cadence; open-weight Gemma releases¹³ est

Complaint themes

Charging to enable deep troubleshooting logs¹⁶ actual
Direct cloud support “not strong”; documentation gaps¹⁶ actual
Android developer verification — 3,050 points, 2,330 comments¹³ actual
Chrome installing a 4GB on-device AI model without consent¹³ actual

Employee sentiment

Practitioner attention is split by topic, not by mood: 204 Hacker News stories about Google above 400 points since January 2025 and 47 about Gemini¹³ actual. The Gemini threads are overwhelmingly about capability; the four highest-scoring Google threads are all complaints about platform control or privacy.

Recent Strategic and Regulatory Moves

The Search Remedy, In Its Own Words

Five Key Insights

Date	Event	Detail
5 Dec 2025	Final Judgment entered	US et al. v. Google LLC, Case 1:20-cv-03010-APM, Doc 1462⁵ actual
Jan 2026	Google appeals	Moved to pause certain remedies; court denied the stay as premature² actual
Feb 2026	DOJ and states cross-appeal	Both sides now appealing the same judgment² actual
10 Mar 2026	Intersect acquired	$5.9bn renewable energy developer; accelerates data-centre power² actual
11 Mar 2026	Wiz acquired	$29.5bn; multicloud and AI-driven security into Google Cloud² actual
Apr 2026	Gemma 4 open models released	1,812 points on Hacker News¹³ actual
4 Jun 2026	Equity raise and Berkshire placement	$20.5bn public offering plus $10.0bn to a Berkshire Hathaway affiliate² actual
Q2 2026	First preferred stock issued	$19.1bn net; 6.25% mandatory convertible, GOOGM / GOOGN² actual
Q2 2026	PriceRunner charge	$2.1bn accrued; Stockholm Patent and Market Court; Klarna subsidiary² actual
Jul 2026	EU Android fine paid	$5.2bn fine and interest, previously accrued² actual
Sep 2026	Gemini 3.8 Flash released	1,158 points, 665 comments on Hacker News¹³ actual

Provision	What the judgment actually says
Distribution	Google may not condition Google Play licensing on distribution or preload of Google Search, Chrome or Google Assistant on any US device⁵ actual
Search index	Must be made available to Qualified Competitors at marginal cost: DocIDs, duplicate notation, DocID-to-URL map, first-seen and last-crawled times, spam score, device-type flag⁵ actual
User-side data	Data used to build or operate the GLUE and RankEmbed models, at least twice, non-discriminatory⁵ actual
Carve-out	Does not require disclosure of intellectual property or trade secrets “such as algori[thms]”⁵ actual
Syndication	Five-year licences on financial terms no worse than offered to any other syndication user⁵ actual
Divestiture	None ordered. The word does not appear in the 35-page judgment⁵ actual
Oversight and term	Technical Committee funded by Google, reporting quarterly for three years; judgment expires six years after the Effective Date⁵ actual

Record earnings are mostly a mark-to-market gain. $99,031m of Q2 income came from SpaceX and one private holding² actual
The capital model inverted. Buybacks $0, $105.8bn raised, Berkshire took $10bn of primary stock² actualest
Free cash flow has gone. $73.3bn in FY2025 to $4.3bn in H1 2026 as capex hit 35.1% of revenue^2,3 est
Cloud is now the growth engine. +81.8% with margin 23.7% to 35.6%, versus a market growing 43%^2,7 est
Search revenue is not yet cannibalised. Search & other grew 16.8% despite the AI-disruption narrative² est

Three Recommendations

Three Competitive Threats

Investment Thesis

Bull case

Cloud compounding at 81.8% with 35.6% segment margin and $513.9bn backlog² actualest
Consolidated growth accelerating: 13.9%, 15.1%, now 24.2%^3,2 est
No divestiture ordered; remedies are conduct and data, not structure⁵ actual

Bear case

Free cash flow $4.3bn in H1 while capex ran at 35.1% of revenue² est
Search index and model training data must be shared for six years⁵ actual
Equity and preferred issuance dilutes while buybacks stopped entirely² actual

Base case

Alphabet converts an advertising annuity into an AI infrastructure business at enormous capital cost, and the market prices the transition rather than the reported earnings. The stock sits about 17% below its 52-week high⁶ actual in the same quarter it reported record GAAP profit, which is the clearest evidence investors are already discounting the securities gain. The decisive variable is whether the $513.9bn Cloud backlog converts at the 35.6% margin now being demonstrated² est; if it does, today's capex is justified, and if it slips, Alphabet will have raised $105.8bn and suspended buybacks to fund depreciation.

Stage

Mature megacap in a capital-intensity transition. Third most valuable listed company in the peer set at $4.14tn, behind NVIDIA and Apple⁶ actual. Forward P/E 22.79 against Microsoft 21.20 and Meta 17.64⁶ actual.

Plausible acquirers

None. At $4.14tn with insiders holding 54.3% of voting power¹⁴ actual, acquisition is impossible. The realistic corporate-action set is court-driven separation, and the December 2025 judgment expressly declined to order any⁵ actual.

Unit Economics

Market Sizing

Measure	Value	Confidence and stated inputs
Operating margin	34.0%	actual TTM; Q2 2026 was $40,770m on $119,796m^6,2
Gross margin	60.9%	actual TTM⁶
Cloud segment margin	35.6%	est $8,814m / $24,768m, Q2 2026; was 23.7% in FY2025^2,3
Capex intensity	35.1%	est H1 2026 capex $80,598m / revenue $229,692m; FY2025 was 22.7%^2,3
Free cash flow margin	1.9%	est H1 2026 FCF $4,261m / revenue $229,692m; FY2025 was 18.2%^2,3
Revenue per employee	~$2.34m	est TTM revenue $445,866m / 190,820 employees. Mixes a TTM numerator with a 2025-12-31 headcount^6,3
Backlog coverage	~20.7 quarters	est $513,900m Cloud backlog / $24,768m quarterly Cloud revenue. Backlog is contracted commitment, not guaranteed revenue²
LTV, CAC, payback	—	unavailable Not computable and not meaningful. No churn, no product-level subscriber counts, no acquisition cost disclosed; most users are unpaid and monetised by advertising³
Burn and runway	—	unavailable Not applicable. $242.5bn cash and investments against $120.8bn debt⁶

Layer	Figure	Reasoning and reliability
Cloud infrastructure market	$143.4bn/qtr	actual Synergy Research, Q2 2026, +43% YoY, 11th successive quarter of acceleration. Definition stated: IaaS, PaaS and hosted private cloud⁷
Cloud SAM annualised	~$574bn	est $143.4bn × 4. Straight annualisation of one quarter in a fast-growing market, so likely conservative⁷
Google Cloud share	15%	actual versus AWS 28% and Microsoft 20%⁷
Global advertising market	>$1tn in 2026	actual dentsu, +5.1%; digital +6.7% and 68.7% of total; 56 markets¹⁵
Advertising market share	—	unavailable Deliberately withheld. dentsu does not decompose the market by company, so any share figure would have no named peer set. See gaps¹⁵
SOM realised	$445.9bn	est TTM revenue: $294,691m advertising (FY2025) plus $58,705m Cloud plus subscriptions and devices^6,3

Peer-set sanity check. The one market-share figure retained — Google Cloud at 15% — is measured against a denominator whose named peer set includes AWS at 28% and Microsoft at 20%, plus Oracle and neocloud providers⁷. Two unambiguous competitors, a research firm that publishes its market definition, and a figure that reconciles to Alphabet's own segment revenue once Workspace is excluded. The advertising share was computable from two retrieved sources and is not published, because its denominator has no named peer set.

Competitive Map and Moat

Company	Market cap	Revenue	Fwd P/E	Overlap with Alphabet
NVIDIA	$5,562bn	$303.0bn	14.84	Supplier and constraint on the AI build⁶ actual
Apple	$4,670bn	$466.8bn	33.42	Search distribution counterparty; now unbundled by court order^6,5 actual
Alphabet	$4,139bn	$445.9bn	22.79	—
Microsoft	$3,711bn	$331.8bn	21.20	Cloud #2 at 20% share; direct AI rival^6,7 actual
Amazon	$2,788bn	$775.7bn	24.86	Cloud #1 at 28%; retail-media ad competitor^6,7 actual
Meta	$1,571bn	$228.2bn	17.64	Digital advertising competitor⁶ actual
Oracle	$457bn	$67.4bn	14.48	Named in the cloud peer set below the big three^6,7 actual

Proprietary data — weakening by court order

The search index was the deepest moat. For six years Alphabet must supply it to Qualified Competitors at marginal cost, along with the user-side data behind GLUE and RankEmbed⁵ actual. Algorithms are carved out, so the moat is narrowed rather than removed — but it is narrowed by an authority Alphabet cannot out-compete.

Switching costs — strengthening

$513.9bn of contracted Cloud backlog, just over half recognisable within 24 months² actual, is roughly 20.7 quarters of current Cloud revenue est. Wiz adds multicloud security as a further retention layer² actual.

Network effects — intact but contested

Search & other still grew 16.8% year over year² est, which is the direct counter-evidence to the AI-cannibalisation thesis. Alphabet discloses no query volumes or search share, so revenue is the only observable.

Brand — concentrated in one word

The Google LinkedIn page has 42,366,833 followers against 259,272 for Alphabet Inc.^9,10 actual. Gemini's app now carries 44.1m ratings at 4.58 stars, more and higher than the core Google app^11,12 actual.

Team Scorecard

ESG Scorecard

Entity Relationships

Closing Thesis

Dimension	Detail	Assessment
Sundar Pichai	CEO of both Alphabet and Google³ actual	Named key-person risk in the 10-K
Anat Ashkenazi	SVP and CFO; signed the Q2 2026 10-Q; $31.0m annualised target compensation^2,14 actual	Executed a $105.8bn multi-instrument raise
Named executives	Philipp Schindler $43.0m; Ruth M. Porat $30.0m; Kent Walker $30.0m target comp¹⁴ actual	PSUs vest on TSR versus S&P 100, 2025-2027
Board	10 nominees, 70% independent¹⁴ actual	Conventional for a megacap
Voting control	Officers and directors as a group hold 92.2% of Class B and 54.3% of total voting power¹⁴ actual	Insider majority control
Largest outside holder	BlackRock: 6.1% of Class A, 2.5% of voting power¹⁴ actual	No outside holder can influence outcomes
Headcount	190,820 at 2025-12-31³ actual	Revenue per employee ~$2.34m est
Attrition	Not disclosed³ unavailable	Blind 4.2/5 overall is the only proxy⁸

Pillar	Evidence	Grade
Environmental	Company states “AI's energy and water demands have made efforts to reduce our emissions more complex and challenging across every level”³ actual. Response was to buy Intersect, a renewable developer, for $5.9bn² actual	Mixed — candid disclosure, capital-backed response
Social	Practitioner backlash on Android developer verification (3,050 points) and a Chrome on-device AI install (1,755 points)¹³ actual. Employee sentiment strong at 4.2/5⁸ actual	Mixed
Governance	Insiders hold 54.3% of voting power¹⁴ actual; Class C carries no vote¹⁴ actual. Antitrust exposure on three fronts including a six-year judgment^5,2 actual	Weak — dual class plus live antitrust

Entity	Relationship	Evidence
SpaceX	Equity holding; primary driver of the Q2 2026 gain	Named in the 10-Q as a source of the $99.0bn unrealised gain² actual
An unnamed private company	Second driver of the same gain	Referred to only as “a private company”; identity not disclosed² unavailable
Berkshire Hathaway	Primary equity investor from June 2026	$10.0bn private placement of 14m Class A and 14m Class C² actual
Wiz	Acquired subsidiary within Google Cloud	$29.5bn, completed 11 Mar 2026² actual
Intersect	Acquired renewable energy developer, consolidated as a VIE	$5.9bn, completed 10 Mar 2026² actual
DOJ and 14 state plaintiffs	Antitrust plaintiffs in the search case	Arkansas, California, Georgia, Florida, Indiana, Kentucky, Louisiana, Michigan, Missouri, Mississippi, Montana, South Carolina, Texas, Wisconsin⁵ actual
European Commission	Regulator; Android and AdSense actions	$5.2bn paid July 2026; EUR 1.5bn AdSense fine annulled, EC appeal pending² actual
PriceRunner (Klarna)	Private damages claimant	$2.1bn accrued after a Stockholm court decision² actual
BlackRock	Largest disclosed outside shareholder	356,934,964 Class A shares, 2.5% of voting power¹⁴ actual
Larry Page	Co-founder; signed the FY2025 10-K as a director	10-K signature page³ actual

Alphabet is an advertising annuity being spent, deliberately and at speed, to buy a position in AI infrastructure. The evidence for the annuity is intact — Search and other still grew 16.8% year over year² and no customer exceeds 10% of revenue³ — and the evidence for the spending is unambiguous: capex at 35.1% of revenue², free cash flow down to $4.3 billion, buybacks at zero, and $105.8 billion raised in six months across common stock, the company's first preferred issue and debt². For this to work, one thing must become true: the $513.9 billion of contracted Cloud backlog² must convert at something like the 35.6% segment margin now being demonstrated², because that is the only disclosed line capable of earning a return on this scale of capital. The mechanism is already visible — Cloud growing 81.8% against a market growing 43%^2,7 is share capture, not drift — and the timeframe is roughly eight to twelve quarters, since just over half the backlog is expected to convert within 24 months². Two things could break it. The first is the six-year judgment compelling Alphabet to hand its search index and model-training data to competitors at marginal cost⁵, which attacks the annuity funding the build. The second is subtler: with GAAP earnings now dominated by a $99 billion mark-to-market gain from holdings including SpaceX², the reported numbers have temporarily stopped describing the business, and a market that misreads them in either direction will misprice the transition.

Contested Figures and Arbitration Outcomes

Q2 2026 earnings per share high Q2 2026 · USD per diluted share · period confirmed

9.11 — Reported GAAP diluted EPS [2]
2.9 — Analyst consensus estimate [6]
2.75 — Derived ex-securities-gain estimate [2]

Arbitration: RESOLVED THREE WAYS, AND IT IS A MEASUREMENT-DEFINITION ISSUE RATHER THAN A SOURCE DISAGREEMENT. Reported GAAP diluted EPS is $9.11, quoted directly from the 10-Q. Analyst consensus was $2.90. My own independent estimate excluding the equity-securities gain, tax-effected at the 21% statutory rate, is approximately $2.75. The consensus and my estimate agree closely with each other and both differ from GAAP for a single disclosed reason: a $99,031m mark-to-market gain the company attributes primarily to SpaceX and an unnamed private company. The GAAP figure is correct and so is the consensus; they measure different things. Any headline describing a '214% beat' is reporting an accounting artifact.

Google Cloud market share vs reported segment revenue medium Q2 2026 · USD / percent · period confirmed

24768000000 — Google Cloud segment revenue, Q2 2026 [2]
15% of $143.4bn — Synergy Research cloud infrastructure share [7]

Arbitration: RECONCILED — the apparent mismatch is a scope difference, not a disagreement. Synergy puts Google at 15% of a $143.4bn quarterly cloud infrastructure market, implying roughly $21.5bn, while Alphabet reports Google Cloud segment revenue of $24,768m for the same quarter. The gap is explained by Synergy's stated market definition, which covers IaaS, PaaS and hosted private cloud only, whereas Alphabet's Google Cloud segment also includes Google Workspace and other subscription software. The two figures are consistent once the definitions are aligned, and the reader should not treat the segment revenue as an infrastructure-market number.

Employee headcount low 2025-2026 · people · period uncertain

190820 — Audited 10-K disclosure [3]
198933 — Market-data vendor, undated [6]
296731 — LinkedIn self-associated profiles [9]

Arbitration: NOT A CONTRADICTION — three different measures at three different dates. The 10-K figure of 190,820 is audited and precisely dated to 2025-12-31 and is the only one used in this brief. The market-data vendor reports 198,933 with no stated date, which is plausible as a later reading. LinkedIn's 296,731 is a count of profiles whose owners self-label Google as employer, which includes former employees and contractors and is not a headcount at all. Reporting any of the three as interchangeable would be wrong.

Trailing versus forward price/earnings high September 2026 · multiple · period confirmed

16.97 — Trailing P/E [6]
22.79 — Forward P/E [6]

Arbitration: BOTH CORRECT, AND THE TRAILING FIGURE IS ACTIVELY MISLEADING. Trailing P/E of 16.97 divides price by earnings that include roughly $135.9bn of H1 equity-securities gains; forward P/E of 22.79 divides by expected earnings that do not. A reader who takes 16.97 as evidence that Alphabet is cheap relative to its megacap peers would be comparing a distorted denominator. This brief uses the forward multiple throughout and shows the trailing figure only with this caveat attached.

Global ad spend total for 2026 low 2026 forecast · USD · period confirmed

surpassing $1 trillion / 68.7% digital — dentsu page as opened [15]
$1.06 trillion / 69% — Search-result summary only — quarantined, not cited (no source — quarantined)

Arbitration: EXTRACTION DISCIPLINE, NOT A CONTRADICTION. A search-result summary rendered the dentsu forecast as '$1.06 trillion' and digital's share as '69%'. On opening the dentsu page, the text says 'surpassing 1 trillion US dollars for the first time' and '68.7 percent of total investment'. The more precise-looking figures existed only in the search summary and are quarantined; the page's own wording is what this brief reports. This is exactly the failure mode the quarantine rule exists to catch.

What We Could Not Establish

Every hole is listed. For a company of this disclosure quality the notable result is that all five critical fields were filled — revenue, valuation, growth, funding and customer concentration all come from audited filings. What remains is second-order, and mostly structural.

X/Twitter following and engagement blocked

Question that would close it: How many followers does @Google have, and what is engagement per post?

Ladder rungs attempted:

x.com/Google — HTTP 200 returned, but the served HTML is the SPA shell: only the bio ('Here to help.') is present and the follower count is JS-rendered and absent
Wayback Machine — a snapshot exists for 2026-08-01 but is a 5.8KB stub with no follower data
SocialBlade — HTTP 403 Forbidden
Press citing a figure — searched; only SEO ranking listicles surfaced, none carrying an @Google number, and all excluded as content_farm

Closable. Closable with an authenticated X session or a paid social-analytics API. LinkedIn was recovered instead, on both the Google and Alphabet pages.

Employee attrition structural

Question that would close it: What is Alphabet's annual and regretted attrition rate, and how has it moved?

Ladder rungs attempted:

10-K FY2025 — point-in-time headcount (190,820) only, no turnover metric
10-Q Q2 2026 — no headcount or attrition disclosure
Blind — sentiment scores recovered but no attrition data

Not closable by any amount of searching. Structural: US registrants are not required to disclose attrition and Alphabet does not. Blind sentiment (4.2/5 overall, management 3.5/5) is the closest available proxy and measures something different.

Recent marketing campaigns and cost structural

Question that would close it: What campaigns has Alphabet run in the last 12 months and at what cost?

Ladder rungs attempted:

10-K and 10-Q — sales and marketing is reported as a single combined expense line with no advertising or campaign breakout
No individual campaign is named in any filing retrieved

Not closable by any amount of searching. Structural under current disclosure. Absence of finding, not proof no campaigns exist.

Share of the digital advertising market not_found

Question that would close it: What share of global digital advertising does Alphabet hold, and by whose measure with what peer set?

Best available: Inputs exist but the share is deliberately not published

Ladder rungs attempted:

dentsu global ad spend forecast — SUCCEEDED for market size and digital share (68.7% of a market surpassing $1 trillion in 2026)
No retrieved source breaks the digital advertising market down by company, so no named peer set exists for the denominator

Closable. Deliberately withheld under the peer-set rule. Alphabet's FY2025 advertising revenue of $294,691m divided into the dentsu total would produce a plausible-looking number with no named competitors behind it. Closable with a company-level source such as an eMarketer or Magna share series.

Enterprise customer satisfaction at scale partially resolved blocked

Question that would close it: How do Google Cloud customers rate it across a large sample, versus AWS and Azure?

Best available: 4.2/5 from 81 PeerSpot reviews, 97% would recommend

Ladder rungs attempted:

G2 — HTTP 403 Forbidden
Gartner Peer Insights — HTTP 403 Forbidden
TrustRadius — HTTP 403 Forbidden
PeerSpot — SUCCEEDED: 4.2/5 from 81 reviews with pros and cons

Closable. Closed on the fourth rung but on a small sample (81 reviews). A like-for-like comparison against AWS and Azure ratings was not attempted and would need the same platform for all three to be meaningful.

Identity of the unnamed private company in the Q2 gain structural

Question that would close it: Which private holding, alongside SpaceX, drove the $99.0bn equity-securities gain, and what is each stake worth?

Best available: SpaceX named; second holding unnamed

Ladder rungs attempted:

10-Q Q2 2026 — names SpaceX and refers only to 'a private company'
10-Q equity-securities note — gives the gain split by measurement category but not by holding
10-K FY2025 — no individual holding is sized

Not closable by any amount of searching. Structural: issuers are not required to name individual portfolio holdings, and Alphabet names only one. Speculating about the second holding's identity would be exactly the kind of plausible guess this brief refuses to make.

Quantified 2026 capex guidance not_found

Question that would close it: What specific dollar capex does Alphabet guide to for FY2026?

Best available: Qualitative guidance only: 'significantly increase' versus 2025

Ladder rungs attempted:

10-K FY2025 — qualitative guidance only, no dollar figure
10-Q Q2 2026 — no quantified full-year guidance found in the retrieved text

Closable. Alphabet gives quantified capex guidance on earnings calls, which were not retrieved. Closable from an earnings call transcript or the Q3 filing. H1 2026 actual of $80,598m is the anchor used instead.

Gemini user or subscriber counts structural

Question that would close it: How many monthly users and paying subscribers does Gemini have?

Best available: Proxy only: 44.1m Google Play ratings on the Gemini app

Ladder rungs attempted:

10-K and 10-Q — no user or subscriber counts disclosed for Gemini or any product
Google Play — SUCCEEDED for app rating and rating count, which is a distribution proxy, not a user count

Not closable by any amount of searching. Structural: Alphabet discloses no product-level user metrics in its filings. The 44,098,663 Gemini app ratings evidence scale of adoption but are not a user count.

Sources

Every URL below was personally retrieved in this session. No source is cited that was not opened. Blocked and deliberately excluded sources are listed so the reader can see what was attempted and why it failed or was rejected.

Methodology and its limits. This brief rests on 103 recorded facts drawn from 16 retrieved sources, with 7 further sources attempted and blocked or deliberately excluded. Of those facts, 77.7% are tagged actual — the value comes from a document that was retrieved — while 16.5% are derived by arithmetic from quoted figures and carry their inputs, and 5.8% are explicitly unavailable. No fact rests on an unsourced lead. 77.7% of all facts carry a verbatim quote or a directly-read primary extract.

What those percentages do and do not claim. They say a stated share of facts traces to a document that was opened and that a verbatim string was captured from it. They do not claim the source is right, nor that a figure means what a reader assumes. The clearest example in this document is Q2 2026 net income of $112,193m: perfectly sourced, perfectly quoted, and actively misleading as a measure of operating performance, because $99,031m of it is a mark-to-market gain on holdings including SpaceX. 3 facts are marked stale and none appears in the executive summary, financial snapshot, KPI figures or investment thesis; each appears only in the annual trend table with its year shown inline. 0 of the five critical fields came back empty. 0 extraction failures occurred: no quote failed to reproduce on re-fetch.

The tooling limit that shapes all of this. The general-purpose fetch tool does not return web pages — it converts a page to markdown and has a small, fast model answer a prompt against it, so an ordinary retrieval yields a summariser's account rather than the document, which is why the heaviest evidence here was downloaded in full and parsed locally (the FY2025 10-K, the Q2 2026 10-Q, the DEF 14A proxy, the complete XBRL company-facts dataset, the 35-page court Final Judgment PDF, and the Google Play schema.org blocks), and why a figure that appeared only in a search summary — a $1.06 trillion ad-market total the opened page did not support — was quarantined rather than cited.

Generated 8 September 2026. Not investment advice. Prepared from public sources only.

Alphabet

Executive Summary

Financial Snapshot

Quarterly Performance and Segment Trend

Annual trend

SWOT

Strengths

Weaknesses

Opportunities

Threats

Capital Structure and Funding History

Customer and Employee Sentiment

Composite score

Consumer apps

Enterprise (Google Cloud)

Praise themes

Complaint themes

Employee sentiment

Overall

Work-life balance

Management

Recent Strategic and Regulatory Moves

The Search Remedy, In Its Own Words

Five Key Insights

Three Recommendations

Three Competitive Threats

Investment Thesis

Bull case

Bear case

Base case

Stage

Plausible acquirers

Unit Economics

Market Sizing

Competitive Map and Moat

Proprietary data — weakening by court order

Switching costs — strengthening

Network effects — intact but contested

Brand — concentrated in one word

Team Scorecard

ESG Scorecard

Entity Relationships

Closing Thesis

Contested Figures and Arbitration Outcomes

What We Could Not Establish

Sources