Liveblogging Worlds in Motion: Nabeel Hyatt (Conduit Labs), TJ Murphy (Warbook/SGN), Mark Pincus (Zynga) on Social Gaming

Ever since Facebook opened its API up to developers to create apps, games, and widgets, indie developers have found success bringing their content to the network's massive audience. Nabeel Hyatt (Conduit Labs), TJ Murphy (Warbook/SGN), and Mark Pincus (Zynga), all working in the field, look at what it takes to succeed when you push your game out on Facebook.

Facebook And The New Web of Social Gaming “The reason why these numbers are so much bigger is that what TJ is doing and what I’m doing is that these games in their DNA are mass market,” said Murphy. “They may be so goofy and lightweight that people in this audience say, ‘That’s not a game at all.’ They’re also a chance to reach a lot of users.” When Warbook started, it was a side project for SGN. Murphy was going to build “a lame class scheduler,” but went for the game instead. A week afterwards, the game was going down from users flooding the system. “Within a couple months I had grown to a million pageviews a day,” he said. “For a normal website a million pageviews is a lot, and this is just some dinky Facebook game. It’s now grown to 800,000 people who have played it and 8000 logging in each day.” Pincus had originally worked to build Tribes.net, an early social network, where he spent most of his time at the Sysiphean task of taking down adult content. When someone else handled the platform, all he had to focus on was the game. “I thought games was the obvious thing because it’s what you do when you get together with your friends,” he said. Not every game succeeds, though. For the top apps, focus on friends is key. “That’s really the element of virality,” said Pincus. The game must also continue to grow, though not too much. Users typically engage in quick sessions, but both Murphy and Pincus wanted to reject the idea of using games solely as timewasters. Plenty of other people see it as an opportunity to socialize. Hyatt, who recently raised funding, asked if the time is over for growth in social games. Pincus showed off a series of cases: Oregon Trail got to “huge numbers” virally in December after just a month (over 210k), but it fell afterwards because there wasn’t a reason to return to the game. In Triumph, a Zynga game, growth didn’t occur virally. The company pushed through marketing, but it continues to see steady growth as users return over and over again. Similar growth happened when the company changed one of its earlier Boggle-like games. “There’s just as much opportunity for growth,” he said. “But it’s getting harder.” There are non-viral games that succeed. Importantly, Facebook is making it harder to promote games virally as it adds more and more restrictions on the ways apps can contact users’ friends. “99% of the games on Facebook are non-viral, meaning that if you just left them alone, they would not grow next week,” said Pincus. “You have to think in the DNA of your game if there’s something to make people come back tomorrow and the next week.” Murphy disagrees to some degree: “The people who stick on games really stick on there for a while, creating a different circle of friends than their real life friends. It almost creates an entirely new social graph of just their gaming friends, which might create a new opportunity and create the viral aspect all over again.” CPM versus Ad Install The question, though, is whether there’s money to be made. As games get bigger, so do the server costs. One issue to keep in mind is that the industry as a whole doesn’t really support independent gamers. In the existing model, games distribute across networks and see only an estimated 1-2% purchase rate for maybe $20k in revenue. Pincus’ Zynga hasn’t disclosed numbers, but it has been entirely supported on revenue generated from games. “Our model is that we built our own application install ad network,” he explained. “We treid all the networks and found really quickly that we’d make at best $.05 CPM. The way to make quickly is to drive traffic to something.” The company turned then to deals with Zwinky toolbars offering users poker chips to click through. The users did it, and the company saw revenue, but it wasn’t a good experience. “We wanted to do something better. We thought just show them applications to try out and see if they want to use it, and we charge $.50/install,” he explained. “It’s a good enough business to be at least better than the download business today, and we have reason to believe in the future it will be even better. It’s Yahoo without the Biz Dev people. You don’t have to kiss anyone’s ring.” Murphy isn’t making money from installs. SGN is working exclusively through ad networks. While some are inefficient, premium networks, like Video Egg, sell user demographic information that can make more money. “The reason we didn’t bother with Video Egg is that it wasn’t worth our time,” argued Pincus. “It’s a bait and switch. CPM is bullshit and social ads are bullshit. Facebook has the scale that they can maybe make that happen. What you can do is send traffic to somebody who can monetize it. We tried everything and we got $.05 CPMs, and you can’t control it.” Hyatt argued that ad installs are self-consuming economies. VCs fund one program that drives traffic to another VC-funded application for install that then sends it back. There’s the chance for it all to dry up. Warbook has enough pageviews—700M across the different games—to get revenue in that direction, responded Murphy. “The space is also really new,” he continued. “And we haven’t even begun to explore the revenue streams that off-Facebook games have been using for years.” No one had touched on digital goods until a questioner brought them up, but Murphy said that he hadn’t considered it until a few weeks ago he found people selling Warbook goods on eBay. Some users make thousands of dollars. “It really opened our eyes,” he said. Pincus said the company recently introduced upgrades to a race game that is now seeing $300-$400 per week. With increasingly low hosting costs, marketing and development is where the money is going out. So even revenues of $300-$400/week for one game out of a stable starts to look pretty good.