Lesson: Accelerating Early Product with Amy Jo Kim

Step #9 Fit: Creating a good hypothesis

When we set out to do Rock Band, I had been talking to the team for at least a year about other projects they had, and I had known them for five years before I ever worked with them. We were friendly. I love music games. I'm a musician myself. It's just my favorite genre. We were very friendly, we were collegial, and when Rock Band time came they said, "We have this amazing team, but we don't have anybody who deeply understands social systems, and you do." That's what I'm known for. "We want you to design all the social system and help us bring the thing to life." So I said "Great".

So at that point it was all of us in a room, sort of a secret room, with these spit and bailing wire instruments, just really ugly, every week different. We were trying, there were controllers, wires sticking out and that kind of hardware. But what we were really focused on was the core loop. And this is a lot of where I learned it, is working with these amazing game designers. We weren't working on onboarding so much. And we had a few early adopters that were not just music gamers, but casual gamers and social gamers because we knew that this was going to be a party game if it was going to be a breakthrough.

But we had this question in front of us. "Is it even possible for four people to get together in a room, pick up plastic instruments and feel like they're in a band?" Now we say yes, because everyone's played Rock Band. Before Rock Band existed, a lot of people said, "That's ridiculous. You're never gonna achieve that," because it's actually hard. It's nontrivial.

So that was the question in front of us. We didn't answer it with polish, we answered it with iterating the core loop. There was a point where even though the screens were not very pretty, because we always had screens for feedback, even though the screens weren't pretty, even though we had wires sticking out of the instruments, it felt like we were playing in a band. And that's when we started to do other things and add polish and do all the other stuff.

And that's when we had fit. Fit was about, "Can we pull this off?" this thing that if we could, it would be amazing, but then when we were doing the marketing and stuff, part of my job was to do a bunch of narratives and scenarios to say, "Well, how would people use this?" and we would say "Okay, what about a bunch of guysn a frat? How would they use it? Okay, they have to spend $160 on plastic instruments. Then they have to have enough room to set it up." We went through that. "Are there enough people in the world with enough money and enough room that they would even do this," and we weren't that sure.

So we did a lot of research and it was a big gamble until it got released. But do you know why it was a success? Because it delivered on the very first thing I said, which is that four people who aren't musicians can pick up instruments, and if they stay at it for a while, they will feel like they're in a band. And that's the core. That's what I mean about core product experience. But to make it work for the market, some of it was dumb luck, and a lot of it was a lot of hard work figuring out how to communicate to people that a $160 bundle was worth their gamble.

But I had a similar experience working with Will Wright on The Sims. I worked with him for years on many products and it was always the same sort of thing, which is at the beginning it was lots of experiments, lots of small, high-learning experiments to get that core thing right. So, small high-learning experiments with the right people. If you take away one thing, that's what MVP experiments are. They're small experiments with the goal of having the highest level of learning, and the clearer you are about the hypothesis that you go into the experiment with, just like any scientist, the better you'll know what caused what.

If your product is interesting and online, think about MySpace. They had it, then they didn't. For product-market fit, it's not like you find it and you're done. Somebody else comes along and does it better.

One of my very early clients was this awesome company called MPlayer. They built infrastructure for online gaming when it didn't really exist. It was a really fast infrastructure. You'd go there and you'd enter a lobby and you'd launch any of a number of games. It's what many . . . it's sort of a really early version of what Xbox is now, but online. And it was really great. We found product-market fit, same thing, I worked with them really early on, designed all the social systems, all the infrastructure, did really well, we're doing great, and then Microsoft entered the market and they did the same thing for free. Boom. The whole thing crumbled.

If you look at the timescale of a business you have to keep finding it, because your market changes. And if you find product/market fit, your market changes and new products come in. I mean, that's sort of the punch line, right, of the MPlayer story. So you can't just find it and rest.