Founder Voting Power vs Share Ownership at 9 Public Companies
All Nine Matched Founders Had More Voting Power Than Their Disclosed Share-Count Ownership
By Lee BaileyPublished September 11, 2026Sources reviewed through August 26, 2026Version 1.09 matched companies · 1 source-gap exclusion
Key findings
All nine matched founders had more voting power than disclosed share-count ownership
The selected cohort shows why founder control cannot be read from share ownership alone. Four matched founders held majority voting power, while the underlying share-count percentages ranged from 2.53% to 35.22%. Alphabet is intentionally excluded from the matched percentage comparison because its current proxy omits non-voting Class C from the ownership table.
9matched founder/company observations
Each preserves the same voting-power endpoint used in the founder voting-control study.
9matched rows where voting power is higher than share ownership
Nine of nine.
4matched founders above 50% total voting power
Dropbox, Meta, DoorDash and Snap.
Voting power and share-count ownership are different measurements. This study does not subtract or divide them into a “control premium,” ratio ranking, governance score, market value or net-worth estimate.
Disclosed share-count ownership across the nine matched rows
The bars use the issuer's matched ownership basis, not total voting power. Most rows reproduce SEC beneficial-ownership tables; Shopify uses its principal-shareholder owned/controlled/directed table, while Zoom directly reports an approximate capital-stock percentage.
Share-count ownership percentage
Issuer-disclosure percentages; see the table below for voting power and denominator notes.
Drew Houston · Dropbox35.22%
Mark Zuckerberg · Meta Platforms13.49%
Brian Armstrong · Coinbase Global12.45%
Brian Chesky · Airbnb11.05%
Evan Spiegel · Snap9.33%
Jack Dorsey · Block8.20%
Eric Yuan · Zoom Communications7.10%
Tobi Lütke · Shopify6.11%
Tony Xu · DoorDash2.53%
A longer bar means a larger share-count ownership percentage under the issuer's stated disclosure basis. It does not measure control quality, economic value or net worth.
Matched voting-power and share-count observations
Each row keeps the two percentages side by side without manufacturing a difference score or ratio. The ownership basis and calculation note explain exactly what went into the denominator.
Class A + Class B common stock; Class C had no shares outstanding
Houston's 9,427,838 Class A plus 74,616,889 Class B beneficially owned shares are divided by the proxy's 163,263,829 Class A plus 75,374,718 Class B shares outstanding. The Class A restricted-stock-award shares in Houston's row are already reflected in the issuer's outstanding-share basis rather than added again.
Voting mechanism: Class B carries 10 votes per share versus one for Class A.
Zuckerberg's 639,347 Class A plus 341,823,978 Class B beneficially owned shares are divided by 2,196,045,588 Class A plus 342,377,716 Class B shares outstanding at the proxy record date.
Voting mechanism: Class B carries 10 votes per share versus one for Class A.
Class A + Class B common stock; Class C had no shares outstanding
Xu's 1,583 Class A plus 11,099,103 Class B beneficially owned shares are divided by the issuer's 412,560,024 Class A plus 24,459,494 Class B base, plus 1,184,139 Class B options exercisable within 60 days that the proxy deems outstanding specifically for Xu's percentage calculation.
Voting mechanism: The separate 14,576,170 Class B shares subject to irrevocable voting proxies from co-founders are included in Xu's 55.5% voting authority but are not added to Xu's beneficial-share numerator.
Spiegel's 30,099,361 Class A, 5,862,410 Class B and 123,683,019 Class C beneficially owned shares are divided by the same three classes outstanding at year-end. Spiegel's footnote does not add a 60-day option or RSU adjustment.
Voting mechanism: Class A is non-voting, Class B carries one vote per share and Class C carries 10 votes per share.
Armstrong's 8,167,972 Class A plus 25,639,618 Class B beneficially owned shares are divided by 222,377,223 Class A plus 41,033,891 Class B base shares, plus 8,167,446 Class A options exercisable within 60 days that the proxy deems outstanding specifically for Armstrong's ownership percentage.
Voting mechanism: Class B carries 20 votes per share. The separate independent-trustee entities established by Armstrong are not added to his individual beneficial-share or voting-power row.
Dorsey's 1,000,000 Class A plus 47,844,566 Class B beneficially owned shares are divided by 535,370,474 Class A plus 59,991,242 Class B shares outstanding. His disclosed row contains no 60-day equity adjustment that needs to be added to the denominator.
Voting mechanism: Class B carries 10 votes per share and converts to Class A on a share-for-share basis.
Class A subordinate voting shares + Class B restricted voting shares + one Founder Share
The circular's principal-shareholder table reports 1,548,000 Class A shares, 77,750,132 Class B shares and one Founder Share owned, controlled or directed by Lütke. Those figures are divided by the same classes outstanding on April 21. P8K does not substitute the separate 60-day SEC executive-ownership table because that would mix ownership bases with the 40.02% principal-shareholder observation.
Voting mechanism: The Founder Share has a variable vote count and represented 1.78% of aggregate voting power at the observation date.
Outstanding capital stock as described by the issuer
Zoom directly reports that Yuan and affiliates held approximately 7.1% of outstanding capital stock and controlled approximately 37.7% of voting power as of July 31, 2026. P8K preserves the issuer-reported percentage instead of reverse-engineering a more precise share count.
Voting mechanism: Class B carries 10 votes per share versus one for Class A.
Class A + Class B + Class H common stock; Class C had no shares outstanding
Chesky's 4,210,326 Class A plus 62,492,000 Class B beneficially owned shares are divided by 418,897,799 Class A, 175,627,453 Class B and 9,200,000 Class H shares outstanding. The proxy reports no Chesky 60-day option/RSU shares requiring a person-specific denominator adjustment.
Voting mechanism: Class B carries 20 votes per share, Class A one vote, and Class H no votes. Class H participates ratably in liquidation and converts one-for-one to Class A when transferred outside Airbnb subsidiaries.
Why the denominator is not always headline shares outstanding
Coinbase: 8.17 million founder-specific option shares
Armstrong's 8,167,972 Class A plus 25,639,618 Class B beneficially owned shares are divided by 222,377,223 Class A plus 41,033,891 Class B base shares, plus 8,167,446 Class A options exercisable within 60 days that the proxy deems outstanding specifically for Armstrong's ownership percentage.
DoorDash: 1.18 million founder-specific option shares
Xu's 1,583 Class A plus 11,099,103 Class B beneficially owned shares are divided by the issuer's 412,560,024 Class A plus 24,459,494 Class B base, plus 1,184,139 Class B options exercisable within 60 days that the proxy deems outstanding specifically for Xu's percentage calculation.
What this comparison supports—and what it does not
Supported conclusions
All 9 matched rows have higher total voting power than their corresponding share-count ownership percentage.
4 matched founders exceed 50% of total company voting power.
DoorDash demonstrates that voting authority supplied by an irrevocable proxy can be much broader than beneficial ownership.
Issuer-specific SEC 60-day ownership rules can require a person-specific denominator adjustment.
Unsupported inferences
The selected cohort is not representative of all public-company founders.
The two percentages should not be subtracted or divided into a universal “control premium” or “control multiple.”
Share-count ownership is not a market-value or net-worth percentage.
Alphabet's missing matched percentage is an unresolved source-coverage state, not zero ownership.
This study does not reconstruct insider transaction history or infer purchases, sales, gifts, conversions or transfers.
Keep each founder's total-voting-power percentage and observation date unchanged from that study.
Publish a matched share-count ownership percentage only when the same issuer disclosure supplies all relevant common-share classes or directly reports the companywide capital-stock percentage.
For SEC beneficial-ownership rows, sum the disclosed class numerators and reproduce the issuer's denominator rules, including person-specific 60-day option/RSU shares when applicable.
Do not fill an omitted class by reconstructing Forms 3/4/5. Alphabet therefore remains a source-coverage exclusion in this edition.
Preserve issuer-specific ownership standards and never convert the paired percentages into a synthetic ratio, control premium, governance score or valuation.
Latest included source filing date: August 26, 2026.
Limitations
This is a selected matched-disclosure subset of the frozen P8H v1 cohort, not a statistically representative sample of founders or public companies.
The share-count metric follows each issuer's disclosed ownership basis. Most rows use SEC beneficial-ownership rules; Shopify uses its principal-shareholder owned/controlled/directed table; Zoom directly reports an approximate capital-stock percentage.
SEC beneficial-ownership rules can deem options or other securities exercisable within 60 days outstanding for the specific person. DoorDash and Coinbase therefore require person-specific denominator adjustments rather than a naive numerator divided by headline shares outstanding.
Voting power and share-count ownership are different measurements. P8K deliberately does not subtract or divide them into a synthetic control premium, control multiple, governance score, market-value measure, or net-worth estimate.
DoorDash's 55.5% voting authority includes Class B shares subject to irrevocable proxies from co-founders; those proxied shares are not treated as Tony Xu beneficial ownership.
Alphabet is excluded from the matched percentage comparison because its current proxy omits non-voting Class C from the ownership table. P8K does not fill that source gap with reconstructed transaction history.
Observation dates differ by company. The study preserves each issuer's measurement date rather than pretending the cohort is synchronous.
Primary SEC sources
Every matched row is anchored in an issuer disclosure that contains both the voting endpoint and its ownership/share basis, with Airbnb's securities description supplying additional Class H economic-rights context.
JSON preserves the nine matched rows plus the explicit Alphabet source-coverage exclusion. CSV includes all ten parent-cohort companies with a status field so missing coverage cannot be read as zero.
When citing a paired row, keep the ownership basis and observation date attached. In particular, do not describe DoorDash's 55.5% voting authority as Tony Xu beneficially owning 55.5% of DoorDash.
Lee Bailey. “Founder Voting Power vs Share Ownership at 9 Public Companies: All Nine Matched Founders Had More Voting Power Than Their Disclosed Share-Count Ownership.” Grizzly Bulls, September 11, 2026. Version 1.0. Data snapshot August 26, 2026. https://grizzlybulls.com/research/founder-voting-vs-ownership
Alphabet is a source-coverage exclusion because the matched proxy omits Class C from its ownership table. The missing matched ownership percentage is not a zero and should not be filled from unreviewed transaction reconstruction.