On Innovation

April 20th, 2010

0

Steve Perlman is a veteran entrepreneur whose past credits include QuickTime, WebTV, and MOVA, which helped earn The Curious Case Of Benjamin Button a visual effects Oscar. His latest innovation is the OnLive gaming platform, which takes the software-as-a-service concept to an extreme, and may well completely revolutionize the gaming industry in the process. He recently gave a talk on the platform at Columbia, which is, in itself, an interesting subject. What makes the talk worth posting here, however, is that, along the way, he shares a lot of insights on how to innovate successfully.

To begin with, the gaming platform is great example of what innovation is really about. Doing gaming successfully over the Web involved solving a lot of truly hard problems really well and then synthesizing those solutions into a complete package. The extent to which OnLive was able to do this becomes especially clear during the Q&A. While it remains to be seen how consumers will respond to the platform, it’s clear Steve and his team tried to anticipate what consumers wanted, in terms of total package. And it’s equally clear how they basically just invented solutions to do so when existing technology was inadequate.

The first part, where he is simply pimping the platform, is interesting for gaming fans, but where it really gets good is when he gets into how they did it. I think there are several really important lessons:

- To bring real innovation to the marketplace, you have to be really, really good. Anyone who has worked with me knows I’m always harping on “world class” and “velocity.” You have to be able to do the easy stuff fast (efficiently) because the hard stuff is what is really valuable. But notice the polish of the consumer experience: they didn’t neglect the basics, either.

- Real innovation is first and foremost about understanding what is possible and being fearless about realizing it. It’s clear that Perlman realized that the speed limit on the Web is theoretically high-enough that you can deliver real-time experience. He then built his team and pursued that objective relentlessly.

- Innovators pay close attention to cognitive psychology. We laugh about the little tricks we play to make something appear to be faster than it really is, but this is no different in principal than realizing that 80ms is the threshold of human real-time perception. That threshold, and thus the entire premise of OnLive, depends entirely on a fact about people, not engineering. It’s not a trick, really. It’s acknowledging that consumers are actually people whose capacity for consuming information has its own very specific performance characteristics.

- To some extent, it’s about faith in your vision. This is easy to overlook: keep in mind, these guys did not know the solutions to many of the problems when they started. They just knew what was possible and believed they had the engineering chops to bridge the gap between “possible” and “it works.”

- Sometimes the solution comes in the form of re-framing (ahem) the problem. As engineers, often our expertise is ironically precisely the thing that prevents us from seeing the solution. A great example is in the way they re-defined the video compression problem, removing a key constraint (I won’t give away the surprise) in order to make huge performance improvements. Most experts in video compression took that constraint as a given, so they viewed the problem as unsolvable.

- Innovation sometimes comes from embracing a constraint, not removing it. YouTube did this when they decided to start letting people upload videos. Outside of a few other like-minded pioneers, conventional wisdom at the time said that you would go broke buying storage. For gaming, conventional wisdom says the network latency is too high. To an innovator, that spells opportunity. They don’t think, ah, rats, I can’t do it. They think, but what if I could do it? They embrace the problem and follow it through to its logical conclusion. That allows you to see opportunities where most people see nothing. YouTube placed a bet on storage costs continuing to drop, while OnLive realized the margins in gaming would cover having local data centers worldwide.

- The first step with hard problems is solving them, not optimizing them. This is a variation of the famous “optimize last” maxim. There’s no point worrying about performance until you have a candidate solution. And the best way to kill a great idea is to immediately dismiss it as being impractical. Obviously it’s impractical, or it wouldn’t be a great idea, it would be what everyone already does. A great example of this here is the way OnLive developed their compression algorithms on very high-end hardware. At that point, a lot of engineers would be panicking, figuring that would never scale. Which it wouldn’t. No problem, says, Steve. Again, I won’t spoil the video by giving away the solution, but suffice to say once they had the algorithm working, they proceeded to figure out how to scale it very cost-effectively.

- Great innovators tend to think in terms of product platforms and ecosystems, not specific products. Notice how the innovation here encompasses everything from the relationships with network providers to the profit margins for the gaming companies to having the games available the same day they show up in stores. Brilliant engineering, like anything, must work within some larger ecosystem. That’s why so many brilliant engineers aren’t really innovators. Think in terms of the entire solution and you’ll be alright.

- Don’t just break the rules, change them. In the end, OnLive is actually providing an experience that is, in many respects, superior to what is available to most consumers, including being able to play games that their own hardware would not support: “We can add the latest and greatest ray-tracing chips and it’s worth it because everyone can have access to them.” Think about what this does to the gaming industry: they are creating a tremendous incentive for game developers to create an experience that will run on the latest high-end hardware, because now it’s available to everyone.

- Stay focused on the gateway innovations. Pay close attention to the last question. Notice how Steve’s vision for what is possible dominates his thought process. He knows that he can move data back and forth across the Web fast enough for it to appear in real-time, so he isn’t overly concerned with unnecessary latency in existing controller designs. He knows the manufacturers will optimize their designs for their platform if they see they can make money on it. The gateway innovation is the real-time experience: controller designs are merely an artifact of existing designs.

How can we apply these lessons to our own work? What is our vision about what is possible? Do we have the faith in ourselves to realize that vision? Are we making the easy stuff too hard and missing the opportunity to really address the hard stuff? Do we let our own experience prevent us from seeing the path forward? Are we thinking in terms of the entire consumer experience? The ecosystem of which we’re a part? Are we aiming high enough in terms of the end consumer experience? Are we creating a consumer experience that is more responsive to consumer’s real aims, with a higher efficiencies all the way through the ecosystem?

I’ll finish by paraphrasing Bruce Sterling, from his crazy but fascinating talk at SxSW this year: “great innovators don’t threaten the established order, they utterly destroy it.”

Tagged with: innovation, user experience, software engineering

Related Posts

• Setting the Stage for Quality Customer Support
• Creating Tech Needs
• How OK Go and Ignite LA blew my mind
• Where's an app for that

Author