Combining Learning Processes - Kathleen Eisenhardt (Stanford)

Doing actually means learning. But what we've seen in learning is first of all, you have a repertoire of processes, you understand strategic timing and you use the right problem solving strategy. I'll unpack that and tell you what that means in like real life. So this is. We did a study of two sided marketplaces. Again, we always look at good companies and bad companies. Bad is probably too harsh, but less successful. Like I think I've seen in every study we've done, every consulting I've done, experimentation is a good thing to do. Experimentation just works. And I can talk more in the Q and A if you want me to tell you what good experiments look like. But very experimentation works. But what people don't always understand is it's actually very synergistic with trial and error. So engaging in trial and error and then experimenting and vice versa actually works really well. And so, for example, in this marketplace, one of the companies in the marketplace was a company that we called by a pseudonym, Market Chef. And what they were doing was it was a business that connected travelers in Asia or around the world, but initially Asia, with local hosts, where you could go and have a homemade meal, so you could go to Bangkok and go to somebody's house and have a homemade meal and a more local experience. So it was essentially a marketplace. What they did was start out with some ideas of what that marketplace was going to be. For example, they thought that the ideal host was going to be a rural poor woman. In fact, that was not. They realized that actually they needed somebody who had Internet, which means you actually were more middle class, not poor, and that people actually wanted to go to cities and not so much rural areas. So it was actually a different demographic. They started to figure that out as they did some trial and error and then did some experiments to actually hone what the ideal host was. Similarly, they didn't really have an idea of exactly which cities they wanted, but they tried a few out. For example, they went to Phuket. I don't know if you all know Phuket in Thailand. Well, if you don't know Phuket, Phuket is a party place, particularly for Australians. So Australians, partying, you can kind of get what that might be. It was not the same as the person who wants a quiet meal in a local experience. So they realized no more partier. So they did some more experimentation. No more party locations. What we really want to go to is a place like Ubud, which you may not, it's in Bali as well, but that's a very local cultural place with not a party crowd. So the idea is they combined experimentation with trial and error.