‘Behind the Startup’ looks at venture capital and inequality

Well into his new book, “Behind the Startup: How Venture Capital Shapes Work, Innovation, and Inequality,” sociology professor Benjamin Shestakofsky describes what happens at a “strategic breakpoint” for a startup that connects buyers and sellers of local services. He says the San Francisco-based founders successfully brought users to the platform but realized that, to raise another round of venture capital (VC) funding, they needed to show profits. They decided to shift from a monthly flat fee for sellers advertising services to a charge for each contact with a potential buyer.

This was potentially a crisis, Shestakofsky says, because the sellers would feel manipulated and ripped-off, and some might have to pay 10 times more. The company had its 10 customer support agents in Las Vegas—all low-wage, female, and working remotely—call the users and gently explain the change.

“Team members were battered with insults and verbal abuse for eight hours a day throughout the transition period, and none escaped without being brought to tears,” Shestakofsky writes. But they kept enough users onboard for revenue to go up, and the startup soon raised another round of VC funding. It was eventually valued at more than $3 billion.

Shestakofsky wrote his new book based on 19 months of participant-observation research at a startup he pseudonymously calls AllDone. Beyond having a remarkably long period of access, Shestakofsky was in the unusual position of rising from intern to middle manager while working for the company.

“I want people who read this book to understand that if we want the outcomes of technological change to be more equitable, we need to focus less on the design of the technology itself and more on these funding structures,” Shestakosfky says. He believes most software engineers want to create things that are good for the world but are “very constrained when they work inside startups that need to hit the next round of funding in order for them to be able to survive.”

He was based in the San Francisco office of AllDone, which connects sellers and buyers of hundreds of local services, such as plumbing, music lessons, and wedding photography. The work also took Shestakosfky to Las Vegas and to the Philippines, where AllDone had about 200 workers, all independent contractors. He says that, while all 20 full-time employees in San Francisco had some degree of equity in the company, this was only true of a very small number across the other locations.

“There are all these people who are making these really important contributions to the company’s success, and they are not able to share in that success because they’re not getting equity; they’re not even employees of the company,” he says. This matters because wealth comes more from assets than income these days, meaning “ownership of assets plays an increasingly important role in economic inequality.”

An assistant professor in the School of Arts & Sciences, Shestakofsky came to Penn in 2018. He says that he spends a class period on technology and society when teaching Intro to Sociology, and he teaches about the hidden labor behind algorithmic systems in an undergraduate seminar on the future of work.

He says he wants students to understand that “our relationship with technology is socially constructed. Yes, we do make choices as individuals, but our choices are embedded in broader structures that create different sorts of opportunities and constraints for us.”

From grad student to middle manager

After graduating from Wesleyan University in 2005, Shestakofsky joined a college friend’s startup, an online college guide featuring user-generated content from students. His strongest memory was that the founder landed an article in The New York Times Magazine, but, when the article came out, the startup’s website immediately crashed.

“I got interested from that experience in that gap between rhetoric and reality when it comes to tech startups and just how important displays of confidence are in that world,” Shestakofsky says. He was also interested in the startup office culture. When he entered graduate school at the University of California, Berkeley, he gravitated back to these questions.

Shestakofsky began working at AllDone in February 2012 for a class on participant-observation research in his second year of graduate school and stayed until August 2013. A friend of his from grad school who went to high school with one of the co-founders of AllDone made the connection, and Shestakofsky says he told the founder he would be happy to help in exchange for the opportunity to collect data. He began going in one day a week as an intern but eventually continued in a full-time role for a year.

“As often happens in ethnographic research, what you think you’re studying is not what actually ends up being exciting about the place you’re studying,” Shestakofsky says. He initially focused on workplace culture but became interested in how algorithmic systems require a lot of human labor to function properly. People might think “there’s some super elegant algorithm” connecting sellers and buyers, he says, but a lot of that process was happening manually by workers in the Philippines.

He wrote a few papers about algorithmic systems but says when writing this book, “I decided I wanted to step back even further to look at how the structures of capital in this world set that all into motion.” More than a decade removed from his experiences at AllDone, Shestakofsky says the issues outlined in his book persist today.

“I'm not anti-technology or anti-innovation, but I do think we need to think harder about what it takes to innovate in ways that actually do create benefits that are more broadly shared because the VC system is designed to funnel the vast majority of the rewards up to the elites at the top,” Shestakofsky says. “That’s not the only way to develop technology, but it is the dominant model that we have.”

He notes how regulatory policies in the 1970s led to the VC boom, and he points to a few alternative models. There are nonprofits, such as Wikipedia and the crowdsourced subtitling organization Amara. There are worker-owned-and-operated cooperatives such as Up & Go, a housecleaning platform with a model resulting in higher wages for workers than the local average. Citing a book by Jessa Lingel of Penn’s Annenberg School for Communication, he notes that Craigslist is a model of a privately owned company that is successful.

“None of these models on their own is going to slay Goliath,” Shestakofsky says, “but I do think if we start to support a broader range of these ownership models, we’re going to open up more space for innovation that benefits more people.”