The Siqi Chen Interview

The following is the full text of the interview I conducted with Zynga Director of Product Management Siqi Chen for my column on social gaming:

Q: How did you decide to begin working on social games?

Back in mid 2007, right after the Facebook platform was announced, I made the decision to get into social games. This is before we had a name for what we now call social games.

There was a visceral sense of new possibilities opening up with the availability of the social graph data for any given user. This new data combined with low friction distribution opened up brand new possibilities around social gameplay – you could for the first time create online games that involve each and every one of your real life friends. This was a big deal – it’s something that simply wasn’t possible before. I actually remember telling a few friends that I believed the largest company on the platform would be the “EA of Facebook.”

Q: Which titles have you worked on at Serious Business and at Zynga? What has your role been on these teams?

Our best known title at Serious Business was Friends for Sale, which was a social network game created by my cofounder Alex Le and myself. The idea was to take HotOrNot and add a market economy component to it. I was (in equal parts with Alex) a designer and developer of Friends for Sale, while also running the company as the CEO.

After our acquisition by Zynga, I run the product management team for Zynga’s Treasure Isle, which just launched. It’s a really exciting launch for Zynga because, by our count, this is the fastest growing social game anyone has ever launched.

Q: In your experience, what has been the biggest challenging in making successful social games. What separates the hits from the misses?

Successful social games are a magical blend of art and science. The art involves understanding what people want, being ruthlessly disciplined with simplicity and accessibility, and creating polished, fun experiences you want to enjoy with your friends. The science involves tracking, storing, and analyzing the billions of actions your players take and figuring out how to retain, grow and monetize your players in the best and most sustainable way.

Q: Steve Meretzky, VP of Game Design at Playdom, has made the point that, with social games, business and design are unified as never before. Do you agree with this assertion? What do traditional game designers need to learn to work on social games?

I cautiously agree with this. Certainly the distribution methods and the virtual goods based business models of social games encourage closer cooperation between designers and the business side. That much needed tension between the two still exists though. Underneath all the data and analysis, at the end of the day there still needs to be a delightful experience as a foundation.

Keep in mind that this advice comes from somebody who is not a traditional game designer, but I think that in order to be successful in this space, traditional game designers need to understand: 1. The context of social games – how and why do people play them? 2. A solid grasp of the economics of distribution and monetization.

The biggest distinction from traditional games is that social games have to basically distribute themselves and monetize themselves. This distinction is the root cause of a lot of design decisions made in successful social games.

Q: How do social games “distribute and monetize themselves” in ways with which traditional game developer might not be familiar? What are some of the best examples you’ve seen?

Because social games are built on top of a social graph, you have the opportunity to get your players to interact with their friends, even if their friends aren’t already playing the game.  These are social distribution opportunities that, when properly designed for, can make your game “viral” and allow you to acquire millions of users in a matter of days. No ad budget, no publishers, no retail shelf space required.

Since these games are distributed socially, all large social game experiences are free to play, monetizing through digital goods. You can play with all your friends, even though chances are only a few of them will ever buy anything.

I think the gold standard on both fronts, objectively speaking, would have to be FarmVille, at 30 million daily players.

Q: How does Zynga make a decision about what to charge for and what to give users for free? Is the internal decision making top-down or bottom-up? How do you know if you are being either too restrictive or too generous?

It’s a combination of experience, data, and intuition. We know a lot about the type of things people have purchased in the past, we have a sense of the things our users want through community feedback, and sometimes there are just awesome things that we think users would pay for.

What’s interesting to me about Zynga is that we are huge fans of our own games, so a lot of the people who work here are some of our best customers. Sometimes they would just come up to me and tell me what they would pay for. This type of feedback mostly comes from the bottom up.

Q: Sid Meier defines a good game as a series of interesting decisions. What are the interesting decisions in a social game? Or does the social aspect trump the need for interesting decisions?

The question I’d ask is, interesting to whom and when? When you’re on Facebook during a 5 minute break from work, the kinds of decisions that are interesting are pretty different from the ones that are interesting to you when you’re engaged in a 4 hour Civ4 marathon.

In a certain context, to certain people, deciding whether to plant the grain or the grapes, whether to buy the chicken or the fruit tree, whether to make my farm look Christmasy or French, are all interesting decisions.

The social aspect doesn’t trump the need for interesting decisions, but it does open up new avenues to create interesting decisions that involve some social tradeoff (Do I help my friend finish his Treasure collection or keep it for myself?)

Q: How do you accentuate these social tradeoffs? What are the best ways to get players to face a tough, but entertaining, choice involving their network of friends?

The mechanics that can work well are going to look pretty different from game to game.

Social interactions are either cooperative or competitive. Letting players gift things to their friends is a good way to build up social currency, but it becomes an interesting choice when the gift itself is scarce, or it costs the players something real.

PvP systems are another great way to provide that interesting choice. However, most players play social games to have fun and relax for a few minutes at a time, so we have to be careful not to design PvP systems that are too punishing.

Q: Charles Hudson, who also worked at Serious Business, said about the genre that “who wants to play a game that’s almost always up and to the right so long as you do what you’re supposed to do?” Social games do primarily seem to just reward time. Is there a place for skill or challenge?

You can make a pretty strong argument that most MMOs primarily seem to just reward time, but they’re awesome games. There’s nothing inherently negative about rewarding people for spending time playing your game. Good social games are about creating the right play experience for the right people at the right time.

There’s a place for skill and challenge, but skill and challenge in a social game might not be as obviously punishing or competitive as a traditional game. Making your farm in FarmVille look impressive is a real skill, and making it look more impressive than your neighbor can be a real challenge.

Q: Let’s talk about the time-based mechanics of social games which have emerged, built around an expectation that play is done in small, five-minute bursts. Some games (such as FarmVille and Social City) use an appointment mechanic which locks players out for a certain period of time once a task is started. Other games (such as Mafia Wars or Treasure Isle) use an energy mechanic, which limits player’s action but is constantly refilling, perhaps encouraging players to micromanage more. What are the pluses and minuses of these two systems?

Sorry, this is too specific for me to be comfortable answering.

Q: How does the design process change when development is based so much on metrics and user feedback? Do you feel less like the “author” of the game experience?

I’d argue the opposite. The biggest difference is that as the author of your game experience, you can see and react to the results of your design decisions and quickly change them.  This doesn’t reduce design to a rote process – you simply have more information available to you, and can quickly iterate and improve your designs.

Q: What specific lessons have you learned from metrics that you would have never guess intuitively? What were your biggest surprises?

Without going into anything specific, the biggest surprise to me is that often, the smallest changes can have a huge impact, and vice versa. This is hugely counter intuitive, but true.

Q: Would you be able to give a specific example? It might help deliver this point home.

This was one of the more ridiculous tests we ran, but it was interesting. Back when I was running Serious Business, Facebook allowed applications to access the notification channel, and we wanted to find out whether longer notifications performed better, or shorter ones. I guessed that it was probably a wash – the shorter ones are more concise, but the longer ones were probably more noticeable since they were physically larger.

We ran a 30-way split test where we asked our team to come up with a bunch of different copy. As it turns out, there was a roughly linear correlation between how short the notification was and how often players would click through it. The shorter it was, the better the performance. The difference in performance was up to 300%. That’s a huge impact for basically writing a few lines of copy.

One thought on “The Siqi Chen Interview

  1. Pingback: Tweets that mention DESIGNER NOTES » Blog Archive » The Siqi Chen Interview --

Leave a Reply

Your email address will not be published. Required fields are marked *