Silicon Valley is a place where people like to talk about failure. Posters at Facebook’s headquarters urge employees to “fail fast, fail often,” and books, such as one from Fab investor Ben Horowitz, “The Hard Thing About Hard Things,” canonize the subject. The countless entrepreneurs on Medium droning on about their shortcomings also preach that ultimately, to win big, you have to take risks. But failure isn’t always a quick stop on the way to success. Sometimes it’s final.

“A lot of people say they’re humbled by success,” Goldberg said. “I’ve experienced a real sort of humility from the colossal failure that was Fab. I’m still working my way, personally, through it.” He said he thinks often of his investors, who lost their money, and his employees, who dedicated themselves to the vision. “I let them down as a leader, and that’s not fun. I feel awful for it.”

Goldberg and his best friend, Bradford Shellhammer, launched Fab in 2011 and sold $20 million worth of merchandise in their first six months. The early problems were good ones to have. Sometimes, demand outstripped supply, and a lot of manufacturers were small and inexperienced. A baker offering rainbow “unicorn poop cookies” assumed nobody would buy them but ended up selling out. Fab scrambled to rent her an industrial kitchen, and employees in aprons arrived to help her bake. Some presents people ordered didn’t arrive in time for Christmas 2011, so Fab paid for similar things on Amazon.com, just so there would be something under the tree (with the original gift still en route).

Fab sped up shipping, but it wasn’t cheap. The company bought warehouses to prepurchase and store inventory. Meanwhile, Goldberg and his investors were encouraged by how enthusiastic people seemed to be about the products, so they started spending on marketing to expand more. They had a $30 million marketing budget in 2013, complete with a TV commercial over the holidays. The customers flooded in, but not in the way Goldberg expected. They would see an ad and make a purchase, and then never come back.

“We kept getting new customers, but we started to see that the profit margins were zero to negative,” said Howard Morgan, an investor at First Round Capital who was on Fab’s board. “The only way the model made money was if people bought multiple items. If you could lose money on each customer, you’re not going to get money out of the whole, and it took us too long to understand that’s what was happening.”

It’s a common story among startups, especially in e-commerce, said Jeremy Levine, a partner at Bessemer Venture Partners. When capital is easy for a hot startup to obtain, investors are so desperate to get in on the deal that they don’t take the time to understand how much the business is spending to acquire each new customer. Meanwhile, entrepreneurs often assume that as long as they can keep raising money, they can fix their business models down the line.

“If the math doesn’t work, eventually the capital dries up,” Levine said. “And sometimes it dries up very quickly.”

Since Fab, there’s been a deeper focus among investors on understanding the unit economics of startup businesses—how much the company makes from every item sold, according to Kirthy Kalyanam, who teaches marketing at Santa Clara University. Fab was badly positioned because it was selling a product people don’t need regularly, so it costs a lot to acquire and retain customers through marketing. Couple that with how much it costs to store and ship products, as well as the capital needed to support small-scale designers, and Fab’s model was bad even when the business looked successful.

“Some of these people invest thinking that the founder has some magic, that he can think about it in a way that nobody understands,” Kalayanam said, noting that Fab had some high-profile investors—Andreessen Horowitz, Menlo Ventures. “Why did it take so long for investors to figure it out? That’s the effect of a charismatic founder.”

In 2012, Goldberg was convinced that it was time to expand into Europe. Copycat sites had cropped up there, and he was eager to beat them while they were still small, instead of having to acquire them expensively later. He acquired one in Germany and another in London. He had the London employees move to Germany and visited them frequently. The chaos was mounting.