Y Combinator Model “Too Messy” for Microsoft, Says YC Founder
On Wednesday, Greg highlighted the provocative suggestion of Kevin Merritt, CEO of Seattle startup Blist, that Microsoft should return to its “risk taking, entrepreneurial roots…” by investing in early-stage startups along the lines of the Y Combinator model. By supporting the startups in return for some stock and a commitment to use Microsoft development software, product platforms, and operating systems, Merritt argued, the software giant could potentially catalyze a lot of innovation that would help keep its products relevant and on the leading edge.
The story drew a lot of attention. But does Merritt’s idea have merit in the real world? I figured who better to ask than Paul Graham, the founder of Y Combinator. The firm—based in Cambridge, MA, during the summer, and Mountain View, CA, in the winter—provides seed funding of under $20,000 to startups and then works with the entrepreneurs to help advance their ideas to the point when they can raise enough money (from others) to get to the next stage. For its support, Y Combinator typically takes an equity stake of between 2 and 10 percent of the startup. You can read more about the model here.
I pinged Graham yesterday by e-mail. Here’s his thoughtful, albeit brief response:
“It would be good for Microsoft if they were able to engage startups more. One way to do it would be by investing in them. I don’t think they could or should do it in the same format as Y Combinator, though. It’s too messy this early. If they were going to invest in startups they should do it in the next stage.”
I wrote back: “When you say it’s too messy this early, what do you mean—all the counseling, false starts, etc?”
“Both,” Graham replied. “At YC’s stage, investing is a lot more involved than just supplying money, and a big company wouldn’t be good at the kind of things you have to do. Plus the failure rate is very high.”
Having written a couple books on innovation in big companies, I think Graham is exactly right. The culture of a Y Combinator would be virtually impossible to replicate by any big company—and even if it were theoretically feasible, the people like Graham who would be needed to run such an incubator would either never sign up to do so—or if they did, they wouldn’t last long.