Somehow, this keeps happening.

A company goes public with a dual class share structure – 10 votes per share for insiders, 1 vote per share for the public, something like that. 

But the company pays its employees in stock, so it issues more 1-vote shares.  Then maybe the company wants to make stock acquisitions – and it issues still more 1-vote shares.  The insiders want to monetize some of their stock, so they convert their 10-vote shares to 1-vote shares and sell them.

Eventually, there is a risk that the insiders’ 10-vote shares will no longer represent a majority.

The board could, I suppose, issue more 10-vote shares to the founders, but even if the charter permits that, it creates difficult questions.  How much should the founders pay for that extra control?  What’s a good price for it? 

When this first happened, the company was Google, and their solution was to amend the charter to create a new class of no-vote shares that could be issued for acquisitions and so forth without diluting the founders’ control.  (It was also an interesting end-run around the exchange listing rules that prohibit disparately reducing or restricting the voting power of traded shares, because, if anything, issuing new no-vote shares enhanced the voting power of traded shares, relative to their economic stake in the company.)

But it prompted a lawsuit in Delaware, with shareholders arguing that the board and the founders breached their fiduciary duties to shareholders by creating a whole new class of stock to perpetuate the founders’ control.  That lawsuit settled for additional limits on how the no-vote shares would be issued, but in the end, Google got its no-vote Class C shares, which now trade alongside the 1-vote Class A shares.

Mark Zuckerberg was so impressed with Google’s plan that he tried the same thing at Facebook.  To stave off a shareholder lawsuit, he agreed to negotiate with an independent special committee – and the shareholders turned up evidence that independent committee member Marc Andreessen was texting Zuckerberg a play by play of committee deliberations and coaching him on how to negotiate.  Ultimately, Zuckerberg dropped the idea.

Next there was Snap.  Snap went public with a tri-class structure.  The public received no vote shares, while the founders had 10 votes per share, and the founders and other insiders had additional 1-vote shares.  The charter specified that when the founders died, or relinquished most of their 10-vote shares, all shares would convert to 1 vote per share.  But the founders wanted to give away stock to charity while maintaining control!  So they proposed charter amendments that changed the conditions under which the share class structure would collapse, such as, among other things, requiring a sell down of both the 10-vote shares and the 1-vote shares.  Shareholders sued, alleging that the charter amendments had the effect of changing the rights of the public no-vote shares, and therefore could not be effected without a separate vote of the public shareholders under DGCL 242. That case also settled when Snap withdrew the proposed amendments. 

There may be other similar examples but the latest entry in the genre concerns Cloudflare, which is now the subject of at least 3 pending complaints that will presumably be consolidated into a single case.  Fourth verse same as the first: The company went public with a dual class structure, whereby the founders had 10 votes per share and the public had 1 vote per share; the charter provided that when the founders’ holdings of 10-vote shares dropped below a certain threshold, everything became 1-vote per share; the founders eventually got antsy about that bargain and proposed a recapitalization.  After the founders got independent special committee approval, the board proposed charter amendments that create a new class of no-vote shares, and a new class of golden shares with no economic rights but that have 9 votes per share.  All shareholders – public and insiders alike – get what is essentially a dividend of 1 no-vote share per outstanding share.  Then, the founders exchange each 10-vote share for one 1-vote share plus 1 golden share.  Meaning, the founders end up with golden shares to maintain control, and monetizable 1-vote and no-vote shares, though if the founders sell enough of their 1-vote stake, the share structure collapses into a single class.  There are also certain conditions involving the founders remaining with the company, and so forth.

Obviously, the public shareholders are claiming this is a breach of fiduciary duty and a transfer of control rights from the public to the founders, but here’s the thing.

We all remember (boy do we remember) the dust-up in Delaware about revising DGCL 144 to make it easier for controllers cleanse conflict transactions, right?

Well, new DGCL 144 has a carveout: its cleansing procedures do not

Limit judicial review for purposes of injunctive relief of provisions or devices designed or intended to deter, delay, or preclude a change of control or other transaction involving the corporation or a change in the composition of the board of directors;

The most obvious application of that carveout was for activist situations, but the plaintiffs are arguing that because Cloudflare is also a case about a provision or device “designed or intended to deter, delay, or preclude a change of control,” the new cleansing procedures do not apply, and we are back in MFW land.

For those who want to keep track of all this, the case numbers are 2026-0772, 2026-0763, and 2026-0734, and it looks like VC Laster caught them.

In any event, I assume founders have now learned their lessons and these scenarios will eventually fade from view; SpaceX, for example, is going public with an authorized class of no-vote shares in its charter, which presumably will be used for acquisitions and whatnot, so that Musk’s voting power does not have to be diluted with new issuances of 1-vote shares.