Product Management Stories: Scott Koenigsberg, of Zynga

Scott Koenigsberg, who leads Zynga’s product management efforts, spoke to me about the risks and rewards of spear-heading game development at Zynga, the leading developer of the world’s most popular social games. Koenigsberg, who is senior vice president of Central Product Management at Zynga, also spent time as the general manager of Mafia Wars and Cityville, two of Zynga’s biggest hits. He shares some learnings below about taking risks in product development, how to keep and maintain user engagement, and other tricks of the trade. 

BING GORDON:  Scott, you helped lead Mafia Wars, one of the first big social game hits on Facebook.

SCOTT KOENIGSBERG: It was an amazing ride. It was unprecedented, and we were inventing on the fly. We had no boundaries; we took risks all the time. 

BING: You created the role of “revenue product manager.” How did that happen?

SCOTT: When I joined, Mafia Wars had millions of daily users, and accounted for a lot of the company’s bookings. Nobody knew how they made money exactly; they were just putting out offers as fast as they could. In my first week, I catalogued and graphed the game’s revenue streams, and presented it in one of our first “numbers review.” Everybody was impressed because of the graph I had created (and the content), so I called myself a “Revenue PM”, or revenue product manager.

BING: Was it always hard?

SCOTT: No! One “aha!” was simply to move the buy button above the fold. I thought most people might be too lazy to scroll down. So I moved the BUY button from the bottom to the top of the page and turned it to green from yellow, which I thought might imply “caution.” It increased revenue by 60 percent. 

BING: Tell me more!

SCOTT: Then we launched Mystery Crates, which we borrowed from TCGs (Trading Card Games)--Magic the Gathering, Pokemon, and things like that. What worked so well was that you didn’t know what the rewards would be. Should we tell people what was included?  Yes; it turned out, people were more likely to buy if they knew the potential outcomes. Would crate art matter? Absolutely; we tried wood, gold, platinum and finally diamond, without changing the drop rates. Diamond increased sales by like 25 percent over platinum. 

"Mystery boxes always work because they’re fun; they should be used all the time."

BING: Another innovation in Mafia Wars was the three levels of Mastery. It was maybe the first mastery feature I’d seen in a social game. 

SCOTT: Mastery is another one of those systems that works everywhere and has a big effect on retention and engagement. And basically all we had to do was multiply the difficulty variable, and drop some higher level rewards.

BING: The Player versus Player feature created so much emergent behavior, too…

SCOTT: We had an amazing player base, who loved the game, rabid for it. We created the core game and mechanics, Declare War, but they created the clan game, the revenge game, and Hit List. The Hit List was one of the greatest things that we didn’t capitalize on enough. 

BING: Did you play a lot of player versus player on Mafia?

SCOTT: I did. I had seven different accounts, 6 for testing, and one for my serious playtime. 

BING: Wiped out while you sleep by investment bankers from New York and oil barons from the Mideast?

SCOTT: And one high-powered woman, too. She was a clan leader who had leveled up so fast, that other players reported her. We were sure she was cheating; she was clicking at such a high velocity that it didn’t seem possible that she wasn’t using Bots. 

So I telephoned her, which I did with a lot of our VIP players. Turned out she was such a nice lady, a former tech exec. I said look, I don’t want to be insulting, but you’ve been reported as potentially cheating. She said, no, I would never do that. I asked well, how do you click so fast? It doesn’t seem humanly possible. She said, let’s switch to videoconference; I’ve got a setup like you wouldn’t believe. She had six different high-end gaming PC’s in a semicircle, all dedicated to playing Mafia Wars. Oh my god!

"When you see that kind of response to the product you work on every day, you are in game creator heaven." 

BING: Social links were so important. What about the “501 friend helper mechanic?”

SCOTT: You could add the weapon power from up to 501 Facebook friends, so anyone who wanted to play the max game needed to find more people, but it was hard to find that many friends. A lot of our players just created multiple fake Facebook accounts. I think we probably caused 10 percent of Facebook’s growth in 2008. We would post “how to find friends” on our own community page and immediately there’d be 10,000 posts following that just said “Add Me” with their Mafia Wars account. 

BING: Did it ever lead to real friendships?                                        

SCOTT: Sure, we even had Mafia Wars weddings. I have tons of anecdotes from Mafia Wars players who became very close friends. Zynga.com was great at building relationships with strangers. 

BING: What happened on Mafia Wars 2?

SCOTT: I did not touch Mafia 2, it already had way too many cooks in the kitchen. We had no hope. We had no clear vision and when we did, it changed often and sometime without reason. 

BING:  Looking back, what goes on your Mafia Wars epitaph?

SCOTT: Other than inventing the Revenue product manager job, it was probably inventing the flash sale, because we discovered it by accident. We meant just to offer it to non-buyers, but we had a coding error, and offered it to everyone. It was our biggest day ever at the time, but kind of scary, because we didn’t know if it was going to cannibalize everything.

BING: After a couple years on Mafia Wars, you moved onto Cityville. How was that different?

SCOTT: I came onto Cityville right after launch. It was a chaotic situation, because they were so darn successful and they had delivered a great game. Before launch, most of us thought the game’s going to top out at 3 million daily active users, and I was more optimistic at 5 million. The highest case projection had the game hitting 10 million, but that seemed impossible.

BING: Yeah. And it peaked at 22? In just a few weeks? What was that like?

SCOTT: That game, that team! They invented so much new stuff, like the “guitar pick” (floating identifiers) for social visits, the franchise concept, area effects. They brought “peeps” (virtual people) onto the board with simulation value. You would see moments on the game board that were based on your actions. You’d open up a business and all of a sudden people would flood to it. It was a beautiful ant farm, but it also had the aspiration of building a gigantic city. It worked for both the decorator and the optimizer, and it just caught fire.

BING: What happened with franchises, where players opened businesses in each other’s cities, for mutual earnings and even skyscrapers? The first five levels were amazing, and created so much fun social behavior. And then it stalled.

SCOTT: There was no vision for what franchises should become. As a game grows, you’re going lose a lot of players along the way. Better matchmaking would have kept franchises more vibrant. If you had a franchise in a dead city, where the other person had stopped playing, it was just a bummer. 

BING: You were the big dog on City, the General Manager, for a couple years. Knowing what you know now, what would you have done differently?

SCOTT: I don’t think we pushed ourselves early enough on “Bold Beats”. That’s what we call any risky feature that potentially increases key metrics by 20 percent or more, that 80 percent of the audience is aware of, and, at minimum, 25 percent interact with on a given day. And, it is also something that creates word of mouth, and that Apple might feature us for.

Bridge was one of those. It came a year after launch, and it really should have been the first feature that we delivered, because it gave decorators a new beach community game board, and it gave the optimizers a new power-up. 

BING: What was another example of a Bold Beat in City?

SCOTT: The biggest Bold Beat we did was Downtown, which failed. You teleported to a whole different game board with its own mechanics. And that’s where we broke it. We should never have changed the core mechanics of that game board and we lost the magic of it. But you have to take risks.

One of the other big Bold Beats that I still love to this day was Cops and Robbers. You had to feed the cops donuts in order for them to catch the robbers who are looting your stores, and when they caught them, it leveled up the store. I haven’t actually seen anything that is comparable. 

BING: The best product managers seem to come out of Stanford CS or Wharton, where you’re from. What happens at Wharton?

SCOTT: It’s not really about the curriculum. It’s just a very entrepreneurial culture. When I arrived at Wharton I was clearly going into investment banking. By the time I got out of Wharton, I knew that I actually wanted to build things and tangibly see the results of my efforts.

"I wanted to build stuff that people, my friends, my family, actually use."

BING: You’ve hired and interviewed tons of entry level product managers. How have you improved your interviewing skills?

SCOTT: We set up criteria. We’re now doing case studies with real Zynga data as part of the interview process. So, we give six different case studies like how would you grow LTV (lifetime value) of this game? How would you analyze this market? There is no right answer obviously. It’s just how do these people think? Is this somebody I want to work with? Are they open? Will you make people around you better? I love undergrads, by the way. Haven’t been spoiled by the world yet.

BING: And who’s the best product manager you’ve ever seen?

SCOTT: I’ve seen hundreds, but Eric Schiermeyer taught me the most; he was the Godfather of Zynga product managers. He had a way about him that was inspiring but also super critical. He pushed you to do your best all the time, and he had a very particular point of view of what things should be. That set the tone here. He was, literally, a black belt PM ninja.

BING: How has Zynga graduated so many entrepreneurs and chief product officers from its Product Management ranks over the years?

SCOTT: It’s the culture of co‑opetition here: I wanna do better than you. You’d sit in the numbers meeting and see how they did it. And you’d walk out of the room and be like screw those guys; there’s no way that they’re beating us this week. But it never turned toxic, because we were raising all the boats together. We also regularly celebrated all the big and small wins.

BING: You’ve been pretty good at forecasting both revenue and daily active users on new games. What do you do to get better at it? 

SCOTT: It’s vitally important because it predicates how much we can invest and what we build and prioritize. We keep trying to improve our market forecasting, but we have actually found that executional risk can have a bigger effect, and is harder to predict. If the team misses dates or quality, an accurate forecast doesn’t matter. Product Managers need to be able to judge whether we have a vision for what the game is going be. Is the design locked?  Do they have the right people?

"And if you have a team who doesn’t have conviction about where they’re going, they’re probably done. We should just scrap the whole team and get a new team in there who does believe in what they’re doing."

BING: What external tools do Zynga product managers tend to use?

SCOTT: I spend a lot of time on Excel. We use Tableau. We try to use Google Docs, we put comments in it, but you can’t really chat on it. What we need is really a robust workflow system to manage the processes, assumptions and discussions that we use to track inputs and outputs, especially for big models and projections. And real time collaboration tools. A mash-up of Excel, charts and Slack would be perfect.

BING: As chief product officer, one of your jobs is to pass along best practices to all the new Product Managers in the company. How do you do it?

SCOTT: We talk a lot about our concept of a Playbook. We benchmark our own games, as well as other games we admire. We try to capture the game flow in screen shots, and then show the math that we know from our own previous games. Then we benchmark technical performance and UI/UX. But ultimately I learned that tuning is king. First get tuning right and then focus on everything else later. It takes time to do that, because simulations don’t tell you enough. The data doesn’t actually tell you the player motivations either, so you need to combine your intuition with the data.

I also chair a weekly Growth meeting, where we go around the room and compare numbers and learnings. It’s really where the rubber meets the road.

BING: You are a big believer in OKR’s, objectives and key results. How did you rate yourself first half of 2016?

SCOTT: Well, we’re not growing fast enough as a company, so, my results can’t be an “A,” right? And we don’t have all the best practices in our games…yet. And we need to test and innovate more quickly. I take accountability for those things. 

BING: So, what apps do you use every day?

SCOTT: I’m using Snapchat a lot. It’s a platform that is not actually targeted at me, but has such amazing growth and engagement, I need to understand it. And I play a lot of games. 

BING: How many hours a week are you playing games? 

SCOTT: Twenty. Maybe more. If I watch TV, I’m probably playing a game at the same time. When I think of my favorite game of all time it was NHL Hockey on Sega. I came home to the fraternity from the bar at three o’clock at night and I would play for three hours with my buddies. That was so social. You sat there and you played with your friends and everybody cheered for each other. And that’s what our games need to be. We need to bring back that feeling of playing with your friends and talking smack in person.

BING: Speed round. Best notification email?

SCOTT: Gotta be Facebook when you’ve been tagged in a photo. You cannot not open them. A good notification has to have a combination of social, relevancy to you, and also surprise on the other side. 

BING: Best peer-to-peer community? 

SCOTT: League of Legends has a great community on Reddit, Twitch and their own site. The CEO at Riot is an amazing guy. Humble, experienced, smart. 

BING: Best first time user experience?

SCOTT: Messaging apps. They get Day 1 retention of like 80%. There is a lot to learn from Snapchat.

BING: What would you tell your 25‑year‑old self to do differently?

SCOTT: Take jobs you’re passionate about. 

Don’t chase the money. It doesn’t usually work out well. If you’re not happy at a job, then get out. Pick jobs where you can surround yourself with people you can trust and you want to spend 15 hours a day with. The most important thing about any job is the people and the culture you are able to curate. 

BING: You’ve always been one of the most positive personalities at Zynga. When you were in high school, were you the one who picked up everybody’s spirits?

SCOTT: Absolutely not. I’ve become more optimistic as I get older. Life’s too short not to be happy.

BING: And finally, what’s kept you here for 7+ years?

SCOTT: It’s always about the people. We have some of the smartest, best, most passionate people I’ll ever get to work with.