How Founders Exit With Nothing

In 2018, Paddy Power Betfair acquired FanDuel for $465 million. It was the kind of number that makes an acquisition look self-evidently successful. The founders were owed zero.

FanDuel had been valued at roughly $1.2 billion the year before, when a proposed merger with DraftKings collapsed under regulatory pressure.

The founders received nothing. So did the employees.

This was zero without a euphemism. The two principal investors held liquidation preferences entitling them to the first $559 million of any sale. The sale produced $465 million.¹² Cofounders Nigel Eccles, Lesley Eccles, Tom Griffiths and Chris Stafford, along with more than a hundred former employees, later sued Shamrock Capital Advisors and KKR. The suit alleges that the investors colluded to undervalue the company ahead of the merger.³⁴

This outcome follows a common shape: a respectable sale below a heavy preference stack. The preference protects investors from the downside. Almost nobody models who absorbs it at a mid-range exit. The fundraising deck has already assigned the future a much nicer number.

Investors buy a different, better class of stock. The preferred label sounds ceremonial. The payment priority is not.

Preferred stock gets priority boarding in the exit waterfall. It carries a contractual right to a defined payment before common stock receives anything. Founders and employees hold common. Preferred goes first, common goes last, and sometimes the proceeds run out between the two.

A 1x non-participating preference on $50 million raised sends the first $50 million of proceeds to investors. Everything above that gets split by ownership. This is the founder-friendly version. In venture documents, restraint is allowed its own marketing category.

Four mechanisms make the stack more dangerous:

• Participating preferred takes the preference and then participates pro rata in the remainder. Investors get paid twice from the same proceeds.
• A multiple turns a 1.5x or 2x preference on $50 million into a $75 million or $100 million claim before common sees a cent.
• Stacking layers each round on the last, with later rounds usually senior. Series D gets paid before Series A, which gets paid before you.
• Boards facing a bad exit often create management incentive plans off the top. That pool is paid ahead of everyone, including common.

Run your numbers through the dilution calculator. Then run them again with a sale at half your last valuation. The first run belongs in the deck. The second belongs in the decision.

Fiduciary duty is not a guaranteed backstop against a zero payout to common. Delaware has already considered the matter.

In July 2005, SDL plc acquired Trados for $60 million. Under Trados's charter, the merger triggered a liquidation and entitled preferred holders to $57.9 million. A management incentive plan took the first $7.8 million. Preferred took $52.2 million. Common stockholders received nothing.

The common holders sued. In 2013, after trial, the Delaware Court of Chancery held that the sale was entirely fair to the common stockholders and that the directors had not breached their fiduciary duties.⁵⁶⁷ Delaware called the result entirely fair, a phrase with the bedside manner of a repossession notice.

The court did criticise the board's process, and the case is now standard reading on the conflicts that arise when preferred-designated directors approve a sale.⁸ The outcome stands. A board can approve a transaction that pays common zero, and a court can find that entirely fair. Your fiduciary duty protections are thinner than you think.

Good Technology is the case to read if you want to understand how this feels to the people it happens to. The preferred financing price and employee tax bill told different stories. The IRS was not accepting the optimistic one.

The company peaked above a $1.1 billion valuation in early 2014 and filed to go public that May. It pulled the IPO. In March 2015, the board turned down an $825 million cash acquisition offer, confident of a billion-dollar public listing. Six months later, Good Technology sold to BlackBerry for $425 million.⁹¹⁰

When the shareholder documents circulated, employees learned that their common stock was valued at 44 cents per share. A year earlier, the internal mark had been $4.32. The preferred stock held by the venture investors was worth more than $3 per share, roughly seven times the common.⁹

Some employees had already exercised options and paid tax on the shares at the peak valuation. In some cases, they borrowed money or drained their savings to cover the bill.⁹ They paid cash taxes on paper gains that then evaporated. Reporting at the time noted staff anger severe enough that someone put a conference room window out.¹¹

The 409A price can govern an employee's tax bill. The preferred price supplies the company valuation repeated outside the finance department. Those numbers move independently. Tax calculated from the first had to be paid in cash; the value behind it disappeared. Only one side of that asymmetry could reach an employee's bank account.

Microsoft bought Skype for $8.5 billion in 2011. Several former employees discovered their equity was worth exactly zero.

Skype's equity documents contained a repurchase provision. On termination of employment for any reason, a departing employee had to offer vested common shares back to the company at the original exercise price, and the company could buy them.¹²¹³ Yee Lee, a former engineer and product manager who left voluntarily after about a year, was among those who had vested stock bought back at grant price while the company sold for billions.¹²

Silver Lake and Skype were called evil in the press. The outrage lasted a fortnight. The repurchase clause had a longer runway. The contract held.

Read your repurchase terms and post-termination exercise window. Treat anything labelled "call right" or "forfeiture" as an instruction. Assume that whatever it permits will eventually be done to you by someone you have never met.

Set aside the catastrophic cases. Assume things go fine in the ordinary startup sense. You raise a Series A, run the company for seven years, and exit somewhere respectable. The announcement gets adjectives. Your bank account gets arithmetic.

What did you earn on the way?

Kruze Consulting publishes salary benchmarks drawn from actual payroll data across hundreds of venture-backed startups. Its figures do not come from self-reported surveys. Seed-stage CEOs averaged $153,000 in 2026. Series A CEOs averaged $203,000.¹⁴¹⁵

Meanwhile, per Levels.fyi, median total compensation for a senior software engineer is roughly $432,700 at Google L5 and $468,400 at Meta E5.¹⁶¹⁷

A Series A founder-CEO earns roughly 43% of what a senior engineer at Meta makes, a difference of about $265K each year. The senior engineer also has liquid RSUs vesting quarterly, no personal guarantees on a lease, no liquidation preference sitting above their compensation, and the ability to quit on two weeks' notice.

Over seven years, that gap is somewhere around $1.8 million of foregone cash. The engineer's equity is liquid. Yours is illiquid. Your entire economic case rests on the exit clearing the preference stack by enough to make your common stock worth more than $1.8 million after tax.

For a founder holding 15% after dilution, the sale needs to clear the stack by roughly $15 million before you have broken even against the job you did not take. That is the threshold for catching up, before founder mythology applies its own valuation.

The spectacular zero makes the case studies. The routine version is a respectable payout that, divided across seven years, works out to less than the engineering job the founder declined.

A sale price ranks acquisitions cleanly. Founder outcomes are less cooperative.

Josh Pigford sold Baremetrics to Xenon Partners in November 2020 for $4 million in cash. He walked away with $3.7 million, and $300,000 went to his ten-person team.¹⁸¹⁹

A $4 million exit paid the founder more than FanDuel's $465 million exit paid its founders.

Structure produced that inversion. Baremetrics had raised $800,000 from General Catalyst and Bessemer. Both investors agreed to write off their investment instead of claiming their preference.¹⁸ The stack was small, and the investors waived it, so almost the entire purchase price flowed to common.

The $4 million sale paid its founder more because very little sat ahead of common and both investors waived their claims. The smaller deal produced the larger founder outcome.

Venture money can still be the right instrument. Most term sheet negotiations give valuation centre-stage billing because it is the number everyone can repeat before reading the rest. Give three or four specific clauses more attention than that number, because valuation cannot rescue common stock from a preference stack.

• Insist on 1x non-participating preferred. Participating preferred is double-dipping. If a lead insists on it, ask what discount to valuation they would accept to drop it, then take the lower valuation.
• Model the whole stack. Include every round, then run the waterfall at 0.5x, 1x and 2x of your current valuation. If common goes to zero at 1x, you have already lost before signing.
• Later-round seniority reaches backwards. A Series C that stacks senior to everything below it changes the outcome of every earlier round retroactively.
• Read every word of the repurchase and forfeiture language, especially anything triggered by termination "for any reason." See Skype.
• Put management carve-outs in place before you need one. A carve-out negotiated at Series A, while everyone is friendly, is a very different document from one negotiated during a distressed sale after the board's interests have diverged from yours.
• Track your 409A against the preferred price. The gap is your tax exposure on exercise. It is most dangerous at the moment the company looks most successful.
• Model the exit that companies usually reach. Most sell somewhere between $30 million and $300 million, well above zero and well below a billion. That is precisely the band where preference stacks eat everything. The announcement will still know which adjectives to use.

The paperwork does not need a villain. Every mechanism here was disclosed and signed with counsel present. Whatever the outcome of their suit about FanDuel's valuation, the preference stack paid the founders what the contracts allocated to common on a $465 million sale: nothing.

Before signing the next financing, give the waterfall three sale prices. Make one low enough to embarrass the fundraising deck. That is the number most likely to make the documents introduce themselves.