How to keep control of your startup without owning the majority

Mark Zuckerberg owns about 13 % of Meta's equity. He controls a majority of its voting power. No board can remove him, no activist fund can force a strategy on him, and no merger happens without his agreement.

That is neither an accident nor a privilege reserved for trillion dollar companies. It is a stack of legal decisions spread over twenty years, and the first one was made in 2005, when Facebook was worth less than a hundred million.

Equity and power are two different columns

A share carries two distinct things: a slice of the profits and a slice of the votes. Nothing forces those two slices to stay equal.

That is the whole subject, and it is what most founders work out too late. When you look at your cap table after three rounds you read a single column, the equity percentage, and you conclude that dropping below 50 % means losing the company. It does not. Economic dilution is unavoidable and even desirable, it is the price of growth. Dilution of power, on the other hand, is a choice you make round after round.

The four floors of the machine

The story everyone tells reduces to dual-class shares. That is the least interesting part, and the last one to arrive.

1. The board, locked before the equity

In 2005, at the Series A, the real negotiation is not only about valuation. It is about who sits on the board and who appoints which seats. The founders keep control of that.

It is the least spectacular decision in the whole sequence and the most decisive. A founder who controls the board cannot be removed, and a founder who cannot be removed negotiates everything else from a different position.

2. The voting proxies, the mechanism nobody cites

This is the real masterstroke, and it is almost absent from the coverage.

Before the IPO, Facebook's successive investors agree to let Zuckerberg vote their own shares. By the time the IPO prospectus is filed he therefore votes a majority of the rights without personally holding that majority.

Put differently, he did not own control: it was lent to him, line by line, by the very people funding the company. This mechanism needs no exotic share class. It only needs enough leverage at the moment of signing.

3. The two share classes, at the IPO

Then comes the familiar part: Class B shares with ten votes for insiders, Class A shares with one vote for the market.

This does not create control. It freezes it, at the precise moment the equity opens up to tens of thousands of shareholders who can no longer be gathered around a table to renegotiate a proxy.

4. Paying in shares without giving up power

Instagram in 2012, WhatsApp in 2014: two acquisitions settled largely in stock. The economic dilution is considerable. The dilution of power is marginal, because the shares issued are the one-vote ones.

That is the quiet benefit of the structure: it turns shares into acquisition currency without every purchase costing a fraction of control.

One problem, five different answers

Facebook is not an isolated case, and the variations are instructive.

Alphabet pushed the logic further by creating a third, non-voting class, which allows issuance almost indefinitely. Snap crossed the line by listing shares with no voting rights at all, which prompted a response from index providers. Palantir built the most technically elegant structure of the set, with a class whose voting rights recalibrate to hold a floor regardless of how the equity moves.

The limit, and it is a serious one

This is the part founders skip, and it is the most useful. Legal control changes the nature of a negotiation. It does not replace the trust of the people who run the company.

What actually transfers to a European startup

None of the above transfers directly, because the law is not the same. The logic transfers completely.

Before any IPO, the toolbox comes down to four things:

1. The corporate form. In France, the SAS offers very wide statutory freedom. The articles of association are where you set required majorities, the grounds for removing the CEO, and the rights attached to each class of shares.
2. Preferred shares. They let you adjust the financial rights and the voting rights attached to a share class, within the limits the law sets.
3. The shareholders' agreement. Veto rights over structural decisions, board or strategic committee composition, exit clauses. Its main weakness is that it binds only its signatories and gets renegotiated at every round.
4. Actual governance. Who sits on the strategic committee, who appoints, and what majority decides. This is the direct equivalent of the 2005 decision.

For companies heading to a public listing the framework recently changed: the French law of 13 June 2024 on attractiveness opened multiple voting rights for founders at IPO, capped and time limited, in line with European work on the subject.

What to take into your next round

The useful lesson is not to copy a structure designed for a listed American company. It is that power is negotiated early, separately from equity, and on topics that look secondary at the moment you sign them.

Concretely, before your next term sheet, be able to answer three questions:

• Who can remove me, and on exactly what grounds?
• Which decisions require my agreement no matter what?
• What happens if the next round is priced lower than this one?

If you cannot answer, your equity percentage is not the first number to look at. For the full picture of each round and what gets negotiated at each one, our guide to startup funding stages walks through the whole sequence.