Founder Control

Founder Control

Doximity is controlled by an insider group holding 80.3% of the votes on 31.3% of the economics (the founder alone 76.3% on 28.9%), and in a February-March 2026 exchange with the SEC that same company conceded that the active-user engagement metrics its executives cite do not correlate highly with subscription revenue — and will be dropped from earnings calls — and that its systems cannot identify or allocate revenue by module [1]. The engagement-metric and module-revenue concessions are set out in full in Disclosure Quality; this chapter's concern is what they mean set beside the control wedge — the same insider block outside holders cannot outvote is also the company telling its regulator that the operating story cannot be independently checked.

The stakes are structural. Outside holders cannot outvote the insider block for roughly five more years: the super-voting Class B converts automatically to one-share-one-vote by June 25, 2031, and until then a two-thirds supermajority and an 80% insider vote leave minority holders no lever to force a change. Set against that, the two primary non-financial inputs to the growth and AI story — the engagement KPIs management cites and module-level revenue — are both now conceded non-verifiable, so a minority holder can neither outvote management nor independently check the operating story it tells. Three facts cut the other way in the same breath: the SEC exchange was a routine disclosure-improvement review closed with no enforcement (closeout March 26, 2026), the dual-class structure auto-sunsets in 2031 rather than running indefinitely, and the founder made zero open-market sales through the roughly 70% drawdown.

For a reader whose exclusion list bars weak governance, this chapter is the gate the rest of the case has to clear.

The wedge between ownership and votes

Source: 2025 Proxy Statement, beneficial-ownership table (as of June 11, 2025) [2].

Doximity carries two classes of stock. Class A, the listed shares, carries one vote; Class B, held by insiders, carries ten [3]. That ratio drives the ownership-versus-control gap. Tangney's 28.9% economic interest is real money — roughly $1.1 billion at the mid-2026 price, even after the shares fell about 70% — but it buys him 76.3% of the vote [4].

Source: 2025 Proxy Statement, beneficial-ownership table [5].

The three large index managers — FMR, Vanguard, BlackRock — together own roughly a quarter of the economics and cast under 6% of the votes [6]. Outside holders supply most of the capital and hold almost none of the control. As of March 31, 2026 the Class B block controlled approximately 79% of the combined voting power [7], consistent with 132.2 million Class A and 50.9 million Class B shares outstanding [8].

Control has concentrated, not loosened

The more revealing fact is the direction of travel. At the June 2021 IPO, executive officers, directors and their affiliates owned 52.4% of the shares and 59.3% of the votes [9]. Four years later their economic stake had fallen to 31.3% while their share of the votes had risen to 80.3% [10]. Ownership and control moved in opposite directions.

Sources: IPO Prospectus, concentration of ownership [11]; 2025 Proxy Statement, beneficial-ownership table [12].

The mechanism is written into the charter. When a Class B holder transfers shares, they convert to Class A and lose the extra votes [13]. Early venture backers — Emergence Capital, T. Rowe Price, InterWest and others — have steadily sold and converted since 2021, and every conversion lifts the relative weight of the votes that remain. Doximity's own filing states the effect plainly: conversion "will have the effect, over time, of increasing the relative voting power of those holders of Class B common stock who retain their shares" [14]. The buyback works the same way: the completed $500 million program retired only Class A shares [15], shrinking the low-vote float against a fixed high-vote block. Both forces push control toward the founder.

The entrenchment toolkit

Super-voting stock is reinforced by a standard set of anti-takeover provisions. The board is divided into three staggered classes, so only one-third stands for election each year and no single meeting can replace it [16]. Amending key charter or bylaw provisions requires two-thirds of the voting power — a threshold only the Class B block can clear or block [17].

Sources: FY2026 Annual Report, anti-takeover provisions [18] and Section 203 [19]; 2025 Proxy Statement, classified board [20] and board leadership structure [21].

Tangney chairs the board he leads as CEO, and the proxy discloses no lead independent director; the company defends the combined role as giving "a single, clear chain of command" [22]. For an outside holder, the practical consequence of the full toolkit is the same: there is no mechanism to force a change of control, a board refresh, or a sale at a premium against the founder's wishes.

What cuts the other way

Three facts keep this short of the misalignment pattern that usually earns a governance red flag.

First, the structure expires. Every Class B share converts automatically to a single class of one-vote stock on the earlier of ten years from the June 25, 2021 prospectus — i.e. June 25, 2031 — or the day holders of two-thirds of Class B elect to convert [23]. Super-voting control is bounded, not permanent; it has roughly five years left.

Second, the board around the founder is independent. Five of six directors are independent, and the audit, compensation, and nominating committees are composed entirely of independent members [24], Corporate Governance — classified board — p.15"). With majority voting control, Doximity could claim NYSE "controlled company" exemptions and drop that independence; it does not.

Third, and most telling for alignment, the founder is not selling. Across roughly 150 insider transactions filed since mid-2025, Tangney's only dispositions were shares withheld to cover tax on vesting RSUs; he has sold nothing on the open market. Total insider open-market sales over the twelve months to June 2026 came to about $9.8 million — routine, pre-scheduled 10b5-1 diversification by non-founder directors and the CFO, against a founder stake worth over a billion dollars.

Source: SEC Form 4 filings, June 2025–June 2026, as reported.

The founder also does what this reader looks for in a founder: he personally opens and answers questions on every quarterly earnings call, most recently the May 2026 fiscal fourth-quarter call [25].

Reading the red flag

The evidence points to founder control that is entrenched but aligned, not the value-destroying pattern the governance screen is built to catch. The alignment case rests on three legs: a founder economic stake of roughly $1.1 billion, no open-market founder selling through a 70% drawdown, and a majority-independent board the company is not required to keep. The entrenchment is real but bounded — control collapses to one-share-one-vote in 2031.

The strongest fact against a clean read is that outside holders have no lever until then. A classified board plus a two-thirds supermajority plus 80% insider votes means minority shareholders cannot force a board change, a strategy shift, or a premium sale for the next five years, and the mechanics of conversion and buybacks are tightening the founder's grip in the meantime. Compensation sits alongside that: Tangney's fiscal 2025 pay was $17.2 million, almost all equity, about 75 times the median employee [26] [27]. Nearly all of that award is newly issued stock, granted to the person who already controls the company.

What would move this from amber to red: material open-market selling by Tangney, any move to weaken or extend the 2031 sunset, erosion of committee independence, or related-party dealing with the controlling holder. None of those is present today. What would move it to green is mechanical — the arrival of one-share-one-vote in 2031 — provided the alignment holds until then.