CompetitorsWide moat

Alphabet (Google) (GOOGL) — moat facet

Alphabet's search monopoly has never been beaten head-on — it has been eroded at the edges by companies that redefined where a search begins.

Alphabet is the rare company that competes with everyone and loses to almost no one, which makes the shape of each rivalry more interesting than the scoreboard. Four are worth understanding, and they attack from four different places.

Four rivals, four numbersOpenAI~800M weekly ChatGPT users (Oct 2025)MicrosoftBing: 4.55% of global search (Aug 2026)Amazon53% of U.S. product searches began there (2020)Apple~$20B a year for the iPhone default (2022)Google's own share of global search: 91.02% (StatCounter, August 2026)
None of the four beat Google at search; each took something else — the question, the cloud, the shopper, or the payment.

OpenAI is the one that changed the question. It did not take Google's search share so much as demonstrate that a different interface could answer the things people used search for — ChatGPT reached roughly 800 million weekly users1 while Gemini passed a billion monthly2, and the fuller treatment of what that does to the search business is this company's own root threat. Microsoft competes across nearly everything except the thing that matters most: Azure against Google Cloud, Copilot against Gemini, Office against Workspace — a broad, grinding rivalry in which search itself has never been genuinely contested. Amazon took the single most valuable slice of query volume years ago, and quietly: roughly half of American product searches now begin on Amazon rather than Google3, which is where buying intent — and the highest-priced advertising — actually lives.

The fourth is not a rivalry at all in the ordinary sense. Apple controls the most valuable audience on earth and does not compete in search; Google pays it roughly $20 billion a year to remain the iPhone default, an arrangement a federal court ruled anticompetitive4, and from January 2026 also supplies the Gemini models behind Siri5. Alphabet's relationship with the company that could most damage it consists mainly of writing cheques.

Read together, the pattern is that Alphabet's search monopoly has never been beaten head-on — it has been eroded at the edges by companies that redefined where a search begins. The pages that follow take them in that order.

Moat trajectory: Narrowing

The core franchise still wins every conventional fight — Microsoft has taken almost nothing in fifteen years — but the unconventional attacks are landing. OpenAI proved a different interface can serve search intent, Amazon already holds the most valuable queries, and the Apple relationship now depends on a court. Dominant, but attacked from more directions than at any point in two decades.

The number that tests this moat
Third-party estimate
Global search share through it all
~90%, essentially unmoved

Fifteen years of Microsoft, a decade of Amazon taking product queries and two years of ChatGPT have not moved the share number — which is why the moat is rated wide and the threats are rated seriously anyway. Watch monetisation per query rather than share: the erosion here shows up in what a search is worth, not in how many happen.

Source: StatCounter search-share data ↗
Dig deeper
DIG DEEPER 01
OpenAI: The Rival That Changed the Question
OpenAI has the one thing Alphabet structurally lacks: no advertising business to protect while building the best answer engine.
DIG DEEPER 02
Microsoft: Everywhere Except Search
Fifteen years of Bing is the strongest evidence the search moat holds against conventional attack — and says nothing about unconventional ones.
DIG DEEPER 03
Amazon: Where Product Search Already Went
Amazon never had to beat Google at search — only to intercept the half of queries that people actually pay to answer.
DIG DEEPER 04
Apple: The Competitor Google Pays
Twenty billion a year is insurance, not distribution: it buys certainty that Apple never builds the alternative it is perfectly capable of building.
References
  1. Third-party estimateChatGPT reached roughly 800 million weekly users.
    OpenAI DevDay disclosure (Sam Altman) — ChatGPT reached 800 million weekly active users — Oct 2025 · publ. Oct 6, 2025 · source ↗
  2. ReportedGemini passed a billion monthly users, the fastest any Google product has reached that mark.
    Press reporting of Alphabet's announcement — the Gemini app passed 1 billion monthly active users (Aug 2026), the fastest Google product ever to the mark; ~90M in Oct 2024, ~350M in Mar 2025, ~650M by Q3 2025 — Oct 2024 - Aug 2026 · publ. Aug 11, 2026 · source ↗
  3. Third-party estimateAround half of U.S. product searches begin on Amazon rather than Google.
    eMarketer / industry surveys — around half of U.S. product searches begin directly on Amazon — Recent surveys · publ. 2024-2026 · source ↗
  4. ReportedGoogle pays Apple roughly $20B a year for default search placement — ruled anticompetitive.
    United States v. Google LLC (D.D.C., Judge Mehta) — DOJ case page: Aug 2024 liability ruling (default-search payments, about $20B a year largely to Apple, found to be unlawful monopoly maintenance); final judgment and memorandum opinion Dec 5, 2025; the United States' response and opening brief on cross-appeal, July 28, 2026 — Liability ruling Aug 2024; remedies 2024–2026 · publ. 2024–2026 · source ↗
  5. ReportedFrom January 2026 Google supplies the Gemini models behind Apple's rebuilt Siri.
    Apple–Google partnership announcement (Jan 12, 2026) — Gemini models to power the rebuilt Siri; reported ~$1B/yr; custom ~1.2T-parameter model served from Apple's Private Cloud Compute; first features iOS 26.4 (spring 2026), full rollout with iOS 27 (fall 2026) — Announced Jan 2026 · publ. Jan 12, 2026 · source ↗
Sources
Generated September 16, 2026