Familiarity breeds success for fledgling companies

Teams featuring at least one "stranger" - someone unknown by the team before its formation - are more than twice as likely to fail as teams of friends, family members or co-workers, a new study has revealed.

Any advantages presumed to be gained from having diverse skills and backgrounds are more than offset by the potential for relational difficulties, which are business-killers.

"Team issues tend to be the No. 1 reason why venture capital-backed companies fail," said Jason Greenberg, associate professor in the Peter and Stephanie Nolan School of Hotel Administration, in the Cornell SC Johnson College of Business.

"There's so much emphasis on having a team," he said, "but you have to go a layer below that and think about the actual dynamics of the team - the governance of the team itself. It's social relationships, and can these people work together? You should not take that for granted."

Greenberg is a co-author of "The 'Devil' You Don't Know: A Test of the Detriments and Benefits of Co-founding with Strangers," which published Sept. 30 in the Journal of Business Venturing. The other co-author is Ethan Mollick, associate professor of management at the Wharton School of the University of Pennsylvania.

Paul Graham '86, co-founder of start-up accelerator Y Combinator, is quoted at the beginning of Greenberg and Mollick's paper: "The people are the most important ingredient in a startup, so don't compromise there." The researchers used a unique survey dataset to explore this idea.

Greenberg and Mollick merged data from two sources - a list of more than 65,000 Kickstarter project creators, surveyed by Mollick in 2014; and data provided by Kickstarter concerning crowdfunding (e.g., amount raised, number of backers). The data gave the researchers insights into whether initial crowdfunding was successful, and whether the business survived past that initial push for backing.

The researchers narrowed the dataset to 3,562 survey respondents (who generated collective crowdfunding of $151 million) who met the criteria for the hypotheses they were trying to prove: Founding teams with strangers should draw a more diverse array of backers for their pitches, and have more varied skills and experiences; and teams with strangers should ultimately be more successful than teams made up of those known to each other.

One challenge Greenberg said he faced in trying to analyze the differences between teams of strangers and teams of familiar individuals: There just aren't that many of the former.

"A lot of the most successful businesses that we tend to see in operation - think of Johnson and Johnson, or Google, for example - are relatively small teams, and almost always are individuals who knew each other prior to founding together," he said. "So I made it like an exercise, to try to find examples of really high-achieving founding teams that were made up of strangers. Based on academic theory, there are all sorts of reasons why teams of strangers should be really successful. But I had a hard time finding any."

Based on their analysis, Greenberg and Mollick found evidence to support their first hypothesis - that teams of strangers bring varied networks - but even stronger evidence to disprove their second.

Teams of strangers, including ventures that included three or fewer founding members, were found to be less than half as likely to deliver the products or services they pitched to potential backers, and two to three times more likely to permanently stop operations following their initial foray into the market.

"There's this remarkable disconnect between academic theorizing and what we see in practice," Greenberg said. "And this paper was effectively an exercise to challenge both of those assumptions. A great team is a great thing, but a bad team is terrible - in that case, it's much better to work alone."

Future work will explore just that idea - that solo founders might have some advantages over teams.

"One of the findings in the paper is how much better solo founders do than teams with strangers," he said. "There are these operating assumptions of venture capitalists and academics that a team is better than an individual. Well actually, maybe not."