one element that’s changing. We talk about pricing pivots, product pivots, a lot of types of pivots. But if you are scrapping the entire thing, that’s not what I’m talking about.
MM: If you’re scrapping it, we would advocate giving the money back to the investors and raising money again if you have a new idea. A real pivot, by definition, takes advantage of business knowledge. Neil Young has a good topic on this on Quora, about the pivot that Ngmoco did. Neil makes a very cogent argument about how they have changed from selling apps to the free-to-play model. If you can’t articulate what you have learned, and it’s not clear that you have a business, you may be better off saying “We ran an experiment and the hypothesis was invalid.”
AMK: Even Chegg, where they moved from a Craigslist model for colleges to textbook rentals, they were leveraging the insight they saw from the Craigslist model that people were buying and selling textbooks. The insight was that students are very price-sensitive and they would rather buy used textbooks, but there was not a great market for that.
MM: Going from Odeo to Twitter was another interesting example. Evan started out doing blogging software and then went on to podcasting. He discovered the art of audio recording created a barrier—that plus the fact that Apple was giving it away [Apple had created a huge directory of free podcasts as part of iTunes]. Then he went in the opposite extreme, making microblogs happen. Some would argue that you couldn’t have had a Twitter without an Odeo.
X: What do you think about Y Combinator, and about the venture incubator model in general? For entrepreneurs, is joining an incubator or an accelerator a good way to start a company?
MM: I’ve become more bullish on Y Combinator through the years. Early on, it wasn’t all that famous. We invested in a company called Weebly, which was one of the first Y Combinator startups, and it’s a great company and has done very well. But when I step back and look at incubators, the biggest problem they have always had is that whoever runs the incubator constrains the potential success. Historically, incubators have always been the extension of one visionary who tries to spin out a whole bunch of different companies. Very quickly, the limits to that become apparent and it stops working.
The thing Paul Graham did that was really smart was that he combined his visionary capabilities with the benefits of being able to select people from the outside world. Bill Joy once said you should always assume that the smarter people are outside your company. And that is what Paul got right. He understood that if you married the incubator idea with the idea of accepting applications from entrepreneurs throughout the world, you no longer had to be the person who generated the good ideas—you could become a picker of good ideas and good people.
Also, I don’t think Y Combinator could have existed 10 years ago. You couldn’t have demonstrated anything meaningful in 10 weeks. Now you can. Although it’s a little bit of a free-for-all from an investor’s point of view. Everybody shows up at these demo days and people crowd around these companies. Overall he has created a very good model, and he tends to select entrepreneurs with a value system that we like. They tend to be product-centric, they tend to be very technical, they tend to be really focused on building something. Where some of the companies could use improvement is in some of the go-to-market aspects, and understanding not just what their product is, but what business they are in. Having said that, you have to assume now that in any one Y Combinator batch there are one or two meaningful, significant companies.
AMK: I think he’s managed to attract some really great talent. Even amongst my students, many who technically could go out to raise money on their own are opting to apply to Y Combinator, because they think it will give them
