How Alphabet (Google) Pays

Technical ladder — L3 through L9

Management ladder — L5 through L8

Executive and board — named officers, chief executive and directors

Disclosed executive layer

Track divergence

Hierarchy qualifications and legacy structures

• Alphabet's securities filings contain no global level catalogue. The L-numbers are observable through employee salary submissions and recruiter material, not through a company publication.
• The mapping is strongest in software engineering and progressively weaker in product, design, sales, marketing and corporate functions, where equivalent scope can be placed differently by organisation.
• Management levels reuse the same L-numbers rather than running a separate M-series. An L6 can be a staff engineer or an engineering manager depending on the organisation, which is why this report shows the two rows side by side.
• There is no consistently published EVP tier between Senior Vice President and the chief executive at Google. Officers should be represented by their disclosed title, not by an assumed corporate ladder rung.
• The 2012 Stock Plan is a closed legacy equity plan and is not evidence of a second employee hierarchy. Acquired companies and Other Bets can retain their own conventions during integration, and Alphabet discloses no crosswalk.
• The non-employee director row is a governance package rather than an employment band, and is excluded from the range chart along with the two executive rows.

Peer-level mapping

Office and market catalogue — calibration factors versus Mountain View

Alphabet reports total global headcount but does not publish reliable employee counts by city, so this catalogue does not rank offices by size. Mountain View, Sunnyvale, San Francisco, New York, London, Dublin, Bangalore, Hyderabad, Zurich, Sydney, Singapore and Tel Aviv are all strategically important, but no audited ranking exists. Auckland and Dubai appear in the source bundles as low-confidence models derived from the Australian and Singapore benchmarks and are excluded here rather than published as observations.

Mountain View anchor medians — the basis of every modeled cell

• The Mountain View column is the anchor. L3 through L9 medians come from Levels.fyi United States software-engineering records and the four manager rows come from the same dataset's engineering-management snapshot, each multiplied by the 1.08 Mountain View city factor.
• L3 through L6 are directly observed component medians. L7 through L9 are observed total-compensation medians with an analytical base, bonus and equity split, because the public sample stops reporting components at that height.
• Bangalore L6 is a direct observation at 9.42 million rupees base, 10.85 million rupees annualized stock and 1.96 million rupees bonus for a 22.23 million rupee package, roughly 34 percent of the United States level median.
• India total compensation runs at about 19 percent of the Mountain View anchor at L3 and rises towards 40 percent at L9, because senior Indian packages carry a far larger equity share. Base pay compresses harder than total, starting at about 16 percent.
• Several high-cost markets move the other way. Sydney, Zurich and Tel Aviv sit closest to the anchor at L3 and fall away at L6, because United States senior equity grows faster than anywhere else in the dataset.
• The named-executive and chief-executive rows are global proxy disclosures converted for comparison only. They are not a local pay range in any selected city, and the geographic factor applied to them should be ignored.

Executive incentive targets — percent of salary

Named-executive outcomes

• A simple point-in-time ratio of outstanding rights to outstanding rights plus remaining reserve is about 34.5 percent. That is not an official utilisation rate, because plan accounting can recycle forfeitures and the June 2026 reserve increase came after the December snapshot.
• The 534 million remaining reserve and the 200 million June 2026 authorisation should not be added together and called a 734 million unused pool. Grants, vests and forfeitures continued in between, and the current unused balance is not publicly disclosed.
• Unrecognised compensation cost for unvested restricted stock units was about $42.9 billion at 31 December 2025, to be recognised over a weighted-average 2.6 years.
• Weighted-average grant-date fair values tell the story of the share price rather than of grant sizing: $188.82 for units granted during 2025, $159.75 for units still unvested at the year end and $142.33 for units forfeited.

Equity plan capacity and grant activity

Vesting — common reported employee schedule

The 10-K says only that restricted stock units generally vest over four years. The 38, 32, 20 and 10 percent front-load above is the commonly reported new-hire cadence in market data, usually monthly and usually with no cliff, and it is the single most important mechanical fact in a Google offer. It means the fourth year of an initial grant is worth barely a quarter of the first, so an unrefreshed package falls off sharply just as the employee becomes most valuable. Executive annual grants are award-specific and time-based, performance units settle on a three-year cliff after certification, and non-employee director units vest one forty-eighth monthly across 48 months.

Eligibility by hierarchy level

• Equity is not confined to senior staff. Public samples show an annualized stock component at L3 in the United States, India and Australia, which is the clearest evidence that Google Stock Units reach the entry level.
• There is still no published minimum eligible band, no standard grant matrix by level or country and no guarantee that any individual employee receives a grant. Alphabet's proxy states that future employee awards are discretionary and not presently determinable.
• Observed annualized equity at the Mountain View anchor runs about $36,700 at L3, $79,500 at L4, $173,100 at L5, $343,500 at L6 and $1.34 million at L9. In India the same levels run about 1.31 million rupees at L3 and 10.87 million rupees at L6, and in Sydney about A$29,900 at L3 and A$135,000 at L6.
• Annualized equity is not a grant face value. It mixes vesting from an initial front-loaded grant with any refreshers, so a single year's number cannot be read back as the size of one award.
• Intern and apprentice equity is offer-specific and should not be assumed. No public minimum band threshold exists for short-term programmes.

Indicative annual grant value by band — Mountain View

Annualized planning value (±25%), not the face value of every new-hire grant. Share equivalents use the $343.34 reference price at 25 August 2026, Class C GOOG and ignore plan valuation rules and PSU performance. Highlighted rows are disclosed grant-date stock awards.

Named executive target equity

Two disclosure quirks matter when reading these rows. First, Pichai's 2025 total contains no stock award, so it understates the scale of his opportunity by roughly an order of magnitude against the $210 million 2026 target. Second, Alphabet states that named executive benefits do not differ materially from broad employee programmes and that it maintains no separate executive pension or supplemental retirement arrangement, which is unusual for a company of this size. The 2026 proxy records a company contribution of up to $11,750 in 2025 to a named executive's retirement plan under the standard matching arrangement.

Board compensation framework

Actual 2025 totals were $437,396 for Frances Arnold and for Martin Chávez, and $462,396 for Roger Ferguson, whose total reflects the Audit Committee chair retainer. Sergey Brin's row shows $1 of other compensation and nothing else, which is the clearest illustration that employee directors are not paid for board service.

Not publicly disclosed

Modeled promotion planning timeline

The one substantive change in the sources is a design change rather than a budget change. For 2026 outcomes Google said it would give more employees the opportunity to reach Outstanding Impact and improve bonus and equity rewards for the highest performers, while slightly reducing multipliers for Significant and Moderate Impact to keep the change budget-neutral. Historical recruiter material describes merit increases effective around 1 January and promotion increases around 1 May or 1 November, and equity refresh planning discussed near the year end, but no current global calendar is published and this should be treated as process context rather than a 2026 rule.

Dataset summary

City-level H-1B wage history

All-years aggregates. P25/P75 are rounded reconstruction values marked modeled; medians and record counts are the stronger fields.

Selected title / worksite records

Evidence classification

Explicit “Not publicly disclosed” index

Known gaps and diligence before relying on a cell

1. 1The two bundles use different foreign-exchange snapshots, 24 August and 25 August 2026. This report standardises on the 24 August snapshot because it is the only one covering all thirteen currencies used here; the difference between them is under 0.4 percent on every rate.
2. 2The bundles disagree on L7 United States total compensation, at $931,000 and $954,000. This report uses $931,000 because it comes with a published component split and a stated observed range of $710,000 to $1.22 million.
3. 3Ruth Porat's 2025 Summary Compensation Table row is reported to the nearest $10,000 in one bundle and explicitly declined in the other, which recommends reading it directly from the proxy. Her row here carries the rounded figures and should be treated as one significant figure less precise than the other four.
4. 4Neither bundle reports a United States L2 intern or apprentice pay range, and both state that intern equity is offer-specific. No entry row below L3 is published here rather than estimating one.
5. 5L10 and L11 employee compensation is not published in either bundle with enough observations to state a median. This report prints no row for either level, which is why the band table jumps from Director of Engineering to the disclosed named-executive average.
6. 6Non-United States L7 to L9 medians do not exist in either bundle. The senior-end geographic factors are extrapolated from the observed L3 and L6 country ratios with damping, and the highest reported India and Australia L7 observations, at $350.6K and $343.2K, are single high submissions rather than medians.
7. 7City calibration factors come from one bundle's model. The other bundle declines to publish city-level numbers at all and uses country proxies, so every non-benchmark city in this report is a grade less reliable than its country anchor.
8. 8The H-1B extracts disagree materially on FY2026, at 2,988 applications on one site and 7,486 on another, and on the FY2026 median, at $193,500 against about $186,802. Both figures are preserved rather than averaged.
9. 9Alphabet discloses geographic revenue in its filings, but neither bundle captured the split, so no regional revenue panel is published here rather than inventing one.
10. 10No companywide 2026 salary increase percentage, pay freeze, salary cut, campus offer revision or off-cycle correction was located in either bundle. The 2025 restructuring of roughly 200 global business roles and the 2024 shift of some roles to growth hubs are workforce events, not compensation actions.

FX rates used — 1 USD equals, snapshot 2026-08-24

Single-date conversion layer for comparability; it ignores payroll-date FX, tax, purchasing power, benefits valuation and hedging. Compensation intelligence, not legal, tax, investment, immigration or employment advice.

Source register — 16 sources

Growth resumed at about 4 percent in 2025 after two years of restructuring, and it happened while stock-based compensation rose 19 percent, so the cost per employee moved faster than the headcount. Alphabet publishes no city-level headcount and no attrition rate, so the geographic signal in the filings is limited to the growth hubs named in press reporting: Bangalore, Chicago, Atlanta, Dublin and Mexico City.