Gumroad: The "Failed Billion-Dollar Company" Is a Thriving $20M-a-Year Business

In 2019, Gumroad founder Sahil Lavingia published an essay titled “Reflecting on My Failure to Build a Billion-Dollar Company.” It’s the full account of a company that raised VC money, laid off 75% of its staff, and had its investors sell their shares back to it for a dollar. But that Gumroad hasn’t collapsed. It survives today as a highly profitable business bringing in roughly $20M a year (about ¥3B), with zero full-time employees. This is a case where “failure” and “thriving company” describe the exact same business, and it’s material for examining the yardstick of success itself.

Numbers Behind the Collapse and the Rebirth

A 52,000-Visitor Launch Day, and the Gravity of 20% a Month

Lavingia was Pinterest’s second employee. He left in 2011 without waiting for his equity to vest and, at 19, founded Gumroad, a payments platform letting creators sell digital content directly with a single link. The prototype was built over a weekend, and when it went up on Hacker News on Monday morning, 52,000 people showed up on day one. Backed by a roster of well-known names (Max Levchin, Chris Sacca, Naval Ravikant, Accel, First Round) and led by a $7M round from KPCB, the company raised over $8M in total, and its goal was, as a matter of course, to become a billion-dollar company.

That goal demanded growth of over 20% a month. Gumroad’s GMV (gross merchandise value) kept climbing, but it never hit that bar, and growth stalled in November 2014. By January 2015, the bank balance had fallen below 18 months of runway, and the path to further funding closed.

The 75% Layoff and Three “Zombie” Years

In 2015, Lavingia let go of 15 of the company’s 20 employees, including close friends. TechCrunch reported “Layoffs Hit Gumroad,” and some of the platform’s marquee creators left. The remaining “skeleton crew” of five gradually shrank further, until eventually it was just Lavingia alone.

The rebuild from this low point came not as a bold single stroke but as a squeeze on the P&L.

In a single year, operating expenses fell to one-eleventh their previous level, while revenue doubled. The moment Gumroad stopped all “spending for growth,” it simply became a healthy business.

What Selling Back Shares for a Dollar Meant

On November 27, 2017, lead investor KPCB sold its shares back to Gumroad for just $1. The reason given was a departing partner and cleaning up administrative overhead, and it shrank the liquidation preference from $16.5M to $2.5M. Other investors followed. By the VC yardstick, this reads as a “death sentence for a portfolio company,” but from the company’s side, it meant freedom from the obligation to chase a billion dollars. From then on, Gumroad reset its management metrics away from growth rate and toward profit and the amount paid out to creators. Cumulative payouts to creators have reached $178M, and starting in 2018, the company began publishing monthly financials and donating 8% of profit ($23,775) to charity.

An Organizational Design With Zero Full-Time Employees

Today’s Gumroad has no traditional employment. Everyone is a contractor working 20–35 hours a week, paid in cash plus equity. Meetings barely exist. Operations run almost entirely on asynchronous, document-based work. Lavingia’s explanation is simple: “If you hire right, you’re surprised by how little process you need.” What he says he was aiming for was “a company we ourselves would want to work at.”

If Tony Dinh’s four hours a day and Levels’ zero-employee approach represent how solo developers work, Gumroad is a rare case that implemented that same philosophy at the scale of dozens of people and ¥3B a year in revenue. It’s proof that designing a company that doesn’t assume full-time employment isn’t a trade-off against scale.

The Price Paid: This Case’s Dark Side

This is also a case where the cost is too large to simply consume as a feel-good story. Lavingia describes the years after the layoffs as feeling “trapped”. He cut off contact with his mother and friends and left San Francisco. The same person who’d joked in 2013 that he wanted to “succeed quickly so I could write about failure” spent years crushed under the actual weight of that failure. On the business side too, the layoffs drove out key creators, the retreat to a smaller, stable footing wasn’t without wounds.

The analysis Lavingia ultimately arrived at isn’t a celebration of hustle, either. Growth, he concluded, is determined almost entirely by the growth rate of the market itself. Product quality and shipping speed were secondary factors. His clear-eyed conclusion is that the creator economy simply wasn’t big enough in 2011 to fit a VC’s time horizon, and that was the real cause of the “failure.”

The Billion-Dollar Yardstick Belonged to KPCB’s Side of the Table

This case isn’t an argument for “turn down VC money.” Without the $8M+ raised and the early name recognition it bought, today’s Gumroad wouldn’t have its foundation either. The question we want to press is who owned the goal. A billion-dollar yardstick is an external demand coming from an investor’s own structure (fund size and liquidation preferences), and the moment you internalize that as your own definition of success, a profitable ¥3B-a-year company becomes a “failure.” The day KPCB handed its shares back for $1, what happened on the company’s side was also the return of that yardstick. In the context of Japanese side businesses and small businesses, even though the funding scale differs by orders of magnitude, the structure of scoring your own business by an external yardstick (a peer’s boast about revenue, a social-media success story) has the same shape.

One more point: the fact that fully publicizing failure became Gumroad’s single biggest marketing asset makes it the largest-scale version of Bannerbear’s transparency strategy. It only becomes an asset, though, for a business with accurate numbers, such as monthly financials, that can withstand being made public.

Also Worth Reading

• PhotoAI — the prototype of the zero-employee philosophy
• Tweet Hunter — a counterexample that struggled “after the sale”