Fail More. Experiment More. Plan Less.

Also, in my experience the initial concept that you start with, that usually changes anyway. If you look at any successful company, the initial concept is quite different from the thing that ends up succeeding. And again, that’s because it’s impossible to predict what will succeed. You have to do a lot of experimentation, and the really critical insights come as a result of trying things out in the real world. You know, you can’t run an experiment sort of in your head and try to figure out what’s going to work.

The other thing that I’ve noticed a lot is in the valley, entrepreneurs understand better that the best order is to develop the product first to get some usage, and then approach investors. Because then you have a lot to show and you can come across with a very impressive product and potentially get better terms, and so on. Whereas here, we do get a lot of inquiries where people are basically saying, ‘I have this great idea, and I’m looking for funding, and once I obtain funding, I’m going to hire a development team, and I know exactly who I’m going to hire, and they’re ready to jump as soon as the funding is there.’

But there again, I think there is a misunderstanding, because that part—the implementation part—is actually the most complicated. That’s where you run into the most problems, in product development and technical implementation. So sometimes, I think people would be served well if they could reprioritize where they put their effort.

Gorman: Just to expand on your personal experience, if you were to look at PayPal and/or YouTube, how many iterations did you go through prior to landing on the models that became the successful companies they eventually turned into?

Karim: There were a number of iterations and very critical insights which were only possible due to real-world usage. Again, these insights are not ones you can gain inside the laboratory.

For example, I remember at PayPal, initially it was a mobile payment service where you could beam money between cell phones. I think that attracted—the total number of users was 5,000 ever, and it was shut down in early 2000 and we decided to go with the Web site instead. And, of course, the way PayPal ended up succeeding was with that amazing synergy between PayPal and eBay.

That’s where PayPal really took off. But I remember the first instances when PayPal users were posting PayPal links on eBay, the reaction from PayPal business development was, ‘What the hell are these guys doing?’—like, ‘Close their accounts, we can’t have this!’ Because we had not anticipated that people would use PayPal on eBay—it’s something we had never thought of.

But then as it really kind of exploded, we saw, ‘Oh wait, this is actually the best usage scenario, so we’re going to go support it’—you know, change our product in support of that pattern.

Same with YouTube. Initially, we did know that it was going to be a video-sharing product, but it wasn’t clear in what space it was going to be. So initially, it had more of a dating focus, which didn’t really end up being very popular, and then we had to expand the concept to make it more broad.

So again, it’s those iterations—and partial failures, even— that help you identify what it is that’s actually going to succeed.