In this episode:

Jordan interviews Gordon Rowe, a seasoned product manager with extensive experience in monetization strategies and product management within the gaming industry. Gordon has worked at Zynga, where he managed Mafia Wars and was the sole product manager on Zynga’s 3rd party publishing team. He later joined Daybreak Games and now works at Facebook. In this episode, Gordon shares his expertise on free-to-play (F2P) game economies, player retention, and balancing revenue generation with player satisfaction.

Topics covered:

• The hypothesis, test, iterate framework for improving key performance indicators (KPI)
• Using funnels to understand player behavior
• Anti-cheat considerations in competitive mobile games
• Zynga monetization tricks and managing in-game currency “hangovers”
• Avoiding "pay-to-win" dynamics in free-to-play games and why League of Legends has elements of it
• Balancing retention and revenue in game design
• How to evaluate F2P games using Day 1 (D1) and Day 7 (D7) retention metrics
• Building a balanced game economy and when to introduce monetization into the player journey
• The maturation of the mobile and PC F2P markets

For more game industry tips:

• Learn more about Jordan & Bright Black
• Listen on iTunes | Spotify (Support the show and leave a review)
• Visit www.playmakerspodcast.com for the full blog post and more!

Timestamps:

[03:16] Gordon’s journey from Mafia Wars to Daybreak Games and Facebook

[07:05] Product management and the hypothesis-test-iterate framework

[10:12] Anti-cheat measures and team structures at Zynga

[13:00] Monetization strategies and avoiding pay-to-win

[17:14] The importance of actionable analytics in game development

[21:00] Managing game economies to avoid “currency hangovers”

[26:00] Day 1 and Day 7 retention metrics: What they tell you about your game

[28:56] Monetization strategies in League of Legends and World of Tanks

[34:24] How to test and iterate for better monetization in F2P

[40:52] The role of Steam reviews and player feedback in PC game success

[44:43] Gordon’s thoughts on product management and monetization strategies

Resources & media mentioned in this episode:

Games & companies mentioned:

• Mafia Wars (Zynga)
• Daybreak Games
• Facebook
• Steam Spy (Analytics tool)
• League of Legends (Riot Games)
• Rocket League (Psyonix)
• Clash Royale (Supercell)
• World of Tanks (Wargaming)
• H1Z1: King of the Kill (Daybreak Games)

Hello, sir. Hey, what’s up, Gordon? How you doing?

Pretty good, man. How about yourself?

Do you remember that interview we recorded a little while back?

I do.

I’m going to publish it.

Awesome. It was funny because I had wondered if anything ever came of that interview.

So what happened is my editor, he was working on another podcast that kind of took off, and he sort of did a fade away, and it just kind of tanked the whole thing. But yeah, that happened. But now I’ve set up everything to do it myself, and I’ve been going through the interviews that I recorded that didn’t make it out, taking the best ones, and I’m calling them lost episodes.

Yeah, awesome.

I think the one we did is really, really good. I think it was a ton of value, so I’m excited to get it out there.

Awesome.

And my plan is to use this right here as, like, the intro to the episode.

Ha, okay. Well, I’m excited if you’re excited.

I think it was a really, really useful interview.

Okay. Yeah, I mean, it’s interesting. I’m trying to remember back on the things that we talked about, and I’m wondering what has changed since then because it feels like this ecosystem is just moving at a million miles an hour.

Some context changed, right? Like the status of Zynga has changed and things like this, but a lot of the principles really haven’t. One thing that was funny is we talked about Steam Spy, and I hadn’t heard of it. Then, now everyone knows about it, so there are things like this.

So it came and went. Steam Spy got clobbered by Steam. They reduced their API access and prevented them from—now, instead of doing deterministic data, they’re doing aggregates and estimates, right?

So there are things like that, but in general, the interview totally holds up. I know people are going to get a ton out of it, and I just want to thank you for being a part of it.

Exciting. Are you doing intros of who's who? I’m kind of in a new role now and I have a little bit of a not quite public face yet, but that is sort of intended to have a little bit of a public face in this role at Facebook.

Say a little bit about that.

Yeah, so I am in a new role at Facebook called a game development partner. I’m working internally with Facebook teams and externally with a handful of selected clients. The thing that they’ve recognized over time, that you and I have always known, is that there’s a whole other section of what drives the value in a game—that’s the bottom of the funnel, basically. Facebook is really great in the top of the funnel, but the bottom of the funnel is in the game. I’m helping both educate internally the Facebook client partnership teams to understand better how to operate and optimize these games, as well as plugging in with a handful of clients and doing essentially direct consulting, very much like the work I’ve been doing for the last 10 years.

Well, that sounds super cool. And I’m excited to talk with you again soon.

So, Playmakers Podcast has a lot of really, really amazing reviews on Apple Podcasts, and we got a new one just last night. And this one may have meant more to me than any other review we’ve ever gotten, and I want to tell you why. It’s because it’s been so long since we did a show, and I feel so bad—I feel so guilty and ashamed. When it’s been a long time like that and you put out an episode, you don’t know what’s gonna happen, and it’s scary. So, to put out the interview with Michael Bross that I was very proud of and then to have someone almost immediately write a review and tell me how, how the show has meant something to them and helped them—well, they didn’t exactly say that. Let me tell you what the actual review said.

So, this is from VSPV:

“He’s back. Awesome news. Five stars. I love this podcast. Re-listened to a couple episodes a few times. Hoping for something new. 😢 But it’s back. 😺💃🎉”

I seriously got choked up when I saw this last night. Thank you. Thank you. If you like the show, these reviews, they’re like manna from heaven. I think you're going to like what we’ve got going on. Let me get you right to the intro. Don’t forget to subscribe because we’ve got a lot of good stuff coming, and this episode is no exception. It’s awesome.

This is my weirdly late-in-the-episode intro to Playmakers, the podcast where I, Jordan Blackman, talk to legends and leaders in the game industry to get useful information that can help you grow in your game, your game career, or your game business. This week, our guest worked as a lead product manager at Zynga and a director of product at Daybreak Games, currently serving at Facebook as an internal consultant. We’re talking free-to-play economies and so, so, so much more. This is ridiculous. Ladies and gentlemen, Gordon Rowe.

Gordon, thank you for coming to PlayMakers Studios.

It's my pleasure. It's good to be here.

We made it, and I’m excited because I think you have a lot to offer our audience. Why don’t we dig a little bit into your background so we can talk about that?

Absolutely. How did I get into this crazy world of free-to-play gaming? Really, mostly by luck.

Good luck or bad luck?

it’s arguable. You could really make an argument both ways. But at this point in my career, absolutely it’s good luck. There’s no question. I love what I do. I love working in games. And I do feel very lucky every day that I get to do what I’m doing today.

And I wouldn’t be where I am, working in the industry on products that I actually like, which is kind of an amazing opportunity. I moved to San Francisco after working in IT consulting for a couple of years following business school. An investor friend had invested in Zynga and was very close to the HR department, and he knew they were hiring rapidly in the Bay Area. They were looking for people with consulting backgrounds who were interested in and could sort of speak the language of games.

I’ve been playing games my entire life. My first job was working at the Babbage’s in the mall, selling PlayStation games.

Yeah, I had an EB Games gig very, very, very early on as well.

Nice. Yeah, that’s the thing, you got the store discount. We had the unofficial video game rental policy where you could take games home for three days to play them—to explain them better to the customers, you see. Then you’d bring them back in. And this is the old days, just shrink-wrapping it in the back of the store and putting the label back on it, and nobody asks any questions. Really, PC gaming was my core game experience from the time I got to college, playing Quake 2. My claim to fame—my internet video game claim to fame—is I was the number one rated Quake 2 Capture the Flag player in the world for one week in 1999.

That’s actually pretty amazing.

I love using that for icebreakers at new jobs and for two truths and a lie.

At parties?

Yeah, it’s at parties. You ever play two truths and a lie where you have to say three things about yourself, two of which are true and one of which is a lie? I always include that one because it’s just far-fetched enough that people go, “Wait, what? Really?” Is that real?

The specificity!

That’s right. That’s right. Well, because all of them are that specific, that’s just how my mind works. I’ve always been a gamer. That’s where I was, almost 10 years ago—eight years ago now—and just this opportunity… I had just moved out to San Francisco and was looking for work, passing my resume around. A mutual friend said, “Oh, I’ll give it to the HR director over at this company that I’m also invested in, and we’ll see what happens.” Zynga was really aggressively hiring everyone they could get their hands on right at that moment.

When was this?

This was right around 2010.

Oh yeah, that was about when I was being recruited.

Yeah. A lot of people. I mean, Zynga bought whole companies—30, 40 people at a time—really just to acquire their employees for, for no other reason than to bring the talent in. Very occasionally, it would be related to what products they were working on. But for the most part, the amount of money that Zynga raised was pretty astronomical in the earliest days of the company. There were three rounds that went from, I think, $50 to $250 million, and then the IPO was a billion dollars. So there was a ton of money happening all around the time that Zynga was growing, and that's what they were putting into: acquiring talent. And there were a lot of amazing people.

They really did.

That's literally what I was just going to say. There were a ton of just hugely, hugely talented people that came through the building there, in my mind, especially in the early days, because they were able to attract people with the magic of the Silicon Valley exit story. this was a pre-IPO company that had raised enough money already that the IPO was ordained. Every single person, from the time that I got hired in 2010 to the time the IPO happened in early 2013, I think every single person that came into the building knew that was part of the story. And it’s hard to find that—the unicorn billion-dollar valuation pre-IPO company that you can get equity in, do good work, and then make good on the backside.

Now, the Zynga story didn’t exactly make good for a lot of people from an equity standpoint, but that’s a different part of the story. Having been there, I learned a lot, both from watching all of our games go up and from watching all of our games go down, because those experiences were very, very different. The way people reacted to them, the way people reacted to failure, was very different because of the success we had had. I think that was a large part of the story of the decline of Zynga. It was amazing to me because it happened so quickly. In a four-year stretch, it was the fastest company in history to a billion dollars in revenue in its first three or so years. And then, within two years, they couldn’t make the transition to mobile, and there was a decline across the board. They’re still solid, they’re making money, but they haven’t had any substantive improvement in the core business in years.

they had also attracted an amazing pool of designers. That was also very appealing, to have the chance to work with some of that talent. But I think a big issue they had was all that hiring. you're building your business on this rocket ship, and there’s just no way to—how do you build the infrastructure fast enough that's really going to support that, right? And they certainly were not putting that in place.

No. And they, for a while prior to the, I would say the six months before the IPO, the company was very dedicated to promoting from within. The growth was so fast, there was tons of opportunity for anyone who got in the door. I mean, I knew people that started in QA on a Facebook game. I mean, this is like the sort of lowest level of entry into the video gaming world that you could ever have—started in QA on Facebook products and within 18 months had moved into, you know, principal analyst roles. We had a lead engineer that had come out of QA. We had several artists that had come through the QA department. We had artists that transitioned over into engineering. You know, there were people in the earliest year and a half that I was there that were really taking advantage of the white space. As this company experienced hypergrowth, there were just opportunities to step in. You know, the guy that transitioned from QA to engineering did it because he kept finding bugs, and they were simple HTML bug fixes. He was messing around with HTML, and he just fixed them. He went to one engineer and was like, "Hey, if I want to upload this, how do I upload it?" The engineer didn’t know who he was but just taught him, right? It's like, you know, "Here's how you do it. At least this is how you do a submit. This is how you run the BitTorrent script that pushes all of the updates to our servers."

And you know what? I’ll never forget—the GM came in one day and was like, "Wait, since when is Joe doing updates to the game?" But they were working, you know? So that was it at that point. Like, it was something that needed to be done. He stepped in and started doing it. There was one question, like, "Oh, why is he doing this?" and nobody batted an eyelash after that because the growth was so hard to keep up with for every single person that was there.

It was an amazing place to learn. And I, I also saw lots of people grow, and I think, you know, we're still seeing that kind of Zynga diaspora.

Yes, absolutely. I love it.

I think you're also one of the people who kind of came in, had opportunity, and grew. So tell us a little bit about your experience.

I mean, I had an interesting trip through Zynga. My earliest days there were a little bit confusing in the way that you can, without clear direction, when you land on the ground in an organization that's growing so quickly. I was not mentally in a place in my career at that point where I could just step in and say, "Oh, this is exactly where we need to go. Everybody follow me." I was looking for more direction from the people around me, to be giving me the kinds of indications of what was the right thing. Like, I knew how to play games. I knew the things that I wanted to build for me to play, but I didn't have any of the other tools to understand what the levers were, what the actual large moving parts are associated with making a game—but also making a game that makes money, because they are different things.

And that, I think, is probably one of the most fundamental appreciations that I walked away from Zynga with, is the idea that building a system that generates revenue is not necessarily the same skillset as building a game that people enjoy. It's only when you get both of those things together that you actually have a product that's a functioning product that can pay the salaries of the people that are working on it and that can continue to grow as its user base demonstrates what about the product it is that they like. And it's a hard set of circumstances to cultivate to get the engineers and the company and the vision for the game to align with what customers want and what their vision of the game and the feel and the experience is going to be. And Zynga struggled—they found that alignment by accident in a few places.

Mafia Wars was the first game that I worked on, and it was absolutely one of those games. There were decisions that got made on Mafia Wars very, very early in the design of that game that turned into some of the biggest advantages for what the product was in its peak, especially. It was heavily, it was very, very social. It was a game that, they had set an arbitrary limit where you needed 500 Facebook friends to be playing this game with you to be as powerful as you could possibly be in Mafia Wars. And the story goes, because I was not there when the 500 Mafia limit got put into the game—it was probably 12 months before I arrived at the company—but the story that I heard was that it was completely arbitrary. One of the game designers said, "Well, let's just set a number that's so high, nobody could possibly hit it. And that way, you'll always be incentivized to add as many of your friends to your game as possible," because that was the goal. It's like, "Well, let's get everybody to... so every friend you add will help you. So maximize the number of friends that you can have in this game and see how it goes."

And what it created was really a secondary social network. People who got really into Mafia Wars started going into the forums and adding themselves to these giant threads. They would literally just post a link to their Facebook profile and they'd say, "Add me." And you would get random friends from random strangers all across the country that wanted to play Mafia Wars and wanted 500 friends so they could be as powerful as they could be. They would come into these threads and they would all add each other. And then suddenly, your whole Facebook wall is full up with a bunch of random strangers, and almost all of what they're posting is related to this game that you're playing. And this was the earliest days of Facebook, before the feed algorithms were limiting what people would see, before the notification channel was gone, so you could get that little red dot popping up just literally hundreds or even thousands of times a day.

So it was a very noisy ecosystem, and it was incredibly, incredibly rapid growth because at that moment in time, there were no ads on Facebook. They hadn't turned the ad ecosystem on. They were purely in a user growth mode, but they'd exposed these two channels in the feed channel and in the request channel that Zynga recognized the value of in terms of driving growth in their games. And so this incredible social mechanic is sitting behind Mafia Wars, powering everybody to invite everybody else in a Facebook room that was completely silent. It was like a crazy person ran out into the middle of an atrium and started screaming. Everybody was paying attention to it. They were the only ones screaming. And who isn't going to watch the crazy person screaming in the middle of the room? And that turned into a lot of success very early on for Zynga. I think the challenge was—ask your question.

No, I just want to hear more about how you got from where you started at Zynga to where you ended up.

Yeah. So that, I mean, that story was... the first six months that I was there were hard because I was struggling to adapt to an organization with as much white space as there was. I spent a lot of time standing over the shoulders of our lead analyst, giant. He is... he was, and is, a giant, incredibly, incredibly smart—like one of the smartest people I've ever worked with—a guy named Fallon, and he's actually still at Zynga, has done incredible work for them for years and years at this point. But he was, he was a friend to me when I first got into the company. And a lot of things were... cars are zooming by me on the freeway, and I was trying to figure out what it was that I should be applying myself to that was the right opportunity to make a mark inside of this new company, because you can feel that. When everybody else around you has that sort of entrepreneurial energy of like, "How do I get involved and make a difference? How do I, how do I be a shining star in this organization and make the thing that I want to make? How do I make this game better?"

And that was really my deep focus: how do I make this game better? How do I make the experience for the people who are playing it more fun in a way that I would like it to be fun? And I only mention that because I found it very interesting at Zynga that I would say more than half of the other product managers I worked with didn't have those bones in them. They had deep analytics, but they didn't have gamer bones. And I think in the end of the, of the whole story of the company, that showed a lot in the ability to really micromanage to the numbers without really seeing what the long-term implications were for those decisions.

So my first stop really was ramping up on analytics inside of Zynga. It was learning how data was stored, learning SQL, getting in and writing queries—simple queries—against learning the data tables that we had, that actually the way that we were storing different aspects of user behavior, just to understand what questions we could answer. That was what that exercise turned into—an understanding of, "Do we have the data to answer this question that I want to ask?" And with coming from the ground up in that way, and I have a background in engineering, and I've worked at least a little bit in computer science and know enough coding to be dangerous, basically. Never spent a lot of time doing it professionally, but I've dabbled. And so coming at it from a more technical angle and looking at walking in and saying, "All right, well, so, like, what data do we have and what can we do with it?" was basically my crash course into Zynga. My first six months to a year was a lot of time doing that. And I really honed in on it after, I'd say, the second to third quarter that I was there. I got really involved in exploit detection.

Because Mafia Wars is a PvP game. It was the only competitive game that Zynga had, was the only game where somebody committing fraud and getting free stuff with stolen credit card numbers, or somebody exploiting a security flaw in our PHP game—which, those that know web security, PHP is not a hugely secure coding technology to be basing your game tech stack on. We had holes in the game, and periodically people would figure out what those were. And they were able to use them to either level unlimited or gain massive amounts of XP or gain huge amounts of premium currency, which was obviously the most detrimental to us. And so for at least a year on Mafia Wars, I was very, very focused…

I would say a third to a half of what I was working on was proactive analysis of the data sets that we saw—looking for outliers, looking for people that were in violation of, like, velocity limits on how fast you can level up without putting money into your account—and also reactive. I was sort of white-glove customer support for our whales, our biggest customers, many of whom were over a thousand dollars in spend. We were giving customer service priority when they came in with calls about their accounts or needs that they had. And some of them were actually able to get, basically, a direct line into the studio. And I was usually the guy on the other end of that phone call. So those experiences really helped me to internalize the player perspective, right?

When somebody who spent over a thousand dollars in your game calls you up and says, "Hey, I got a problem. Here's what it is," and your job is to help solve that problem, I'm looking for opportunities to not solve this problem once, but to solve this problem for everybody all the time. And that turned into a lot of this sort of back-end cheat detection stuff that we were running, looking at velocities of a bunch of different metrics. Mostly, we were looking for people that were attacking too fast because it indicated that they were scripting to hit the attack button over and over again. And it could basically, because of our tech stack, lock somebody else out of their game if you overwhelmed them. It was essentially like a DDoS in miniature in Mafia Wars, and that turned into a weapon. It was definitely against the rules of the game at that point, but it gave me the first opportunity at Zynga to find that thing that resonated with me that I could chase after proactively—something my boss didn't really know was a problem until I... I'll never forget the day I took him a data set where I was like, "So listen, we got like 40 or 50 accounts that just in the last month have walked out the back door with like $30 million worth of premium currency." And he was like, "Holy shit, what?" It was the first that anyone had looked. It was the first time that we'd ever seen anything like it. And through that investigation, we flagged these accounts and started getting deeper telemetry. We basically...

And it's not just the lost potential sales, it's also the damage that it's doing to all the players who do spend money.

Exactly.

Or who just play without spending money, but are playing and following the rules.

Yes. That was how I found out. I found out because one of our whales called me and was like, "Listen, I just got beat by this account, and there are 15 accounts in this game that can beat me. And I know every one of them because I know how much they've spent on the game in order to be that powerful. It's more than me." And the whales—the highest-level players in the game—all knew each other because it's a small world when you're talking about people that'll put a thousand-plus dollars into a game they're playing. And so I had one of our whales escalate it to me and say, "Check out this account for me. Just do a little digging on it and see if it's legit. If it's legit, I'm cool, don't worry about it." And so I did. We loaded up the account in the account tool, and it smelled fishy from minute one. It took more investigation to figure out what was going on there than typical, and that was just a further indication that there was nefariousness going on. It turned into a substantive investigation.

It really wasn't until I put that number on my boss's desk, and he was like, "Oh, let me give you some engineering resources to actually try and go fix this problem in a more permanent way." Because once—that was the rule at Zynga—once you can put a dollar value against something, you can get resourced. You can get prioritized in building a solution because there's a dollar value next to it. And that process is actually probably the thing—it's a great transition—that I took away from my time at Zynga the most strongly: adherence to the idea that hypothesis, test, iterate is the approach to making a game better. Go in with an open mind.

Use your intuition and your history to generate hypotheses. Say, "All right, well, given what I see in front of me, what do I think will work?" But once you spec those out, go through the exercise of actually generating estimated outcomes. Really sit down with a number and say, "Okay, well, this feature is likely to affect X%age of my user base, and I believe in that because I shipped a previous feature that was very much like it." So you have a comp in your head to something that's already happened, something that you believe in, and say, "All right, well, this new thing is going to be a lot like this old thing looked," right? And that exercise, as long as you have that historical data...

That's exactly what I was going to ask. I think it's very easy to do that when you have those comps, but when you're creating something new or you haven't gone through... there's not another PM that's gone through that process before.

Well, if you don't have another PM that's gone through the process before, that's way harder. I mean, figuring any of this stuff out—like, everybody's standing on the shoulders of people that came before. Product management in games was born out of product management in e-commerce. The toolset is very, very similar. You're looking at funnels every single time in a lot of different ways.

You want to understand: How many players got to step one? How many players went from step one to step two? How many players went from step two to step three? But the funnels, they point at different things. Sometimes you have a user funnel that indicates that they're going to come back another day, right? Responding to a notification or a push on an iOS device or on a mobile device is a huge indicator of somebody's likelihood of retaining—the fact that they're interacting with your notification channel. Because a lot of people don’t; they'll refuse to turn notifs on. And that's a powerful channel—that reminder mechanic that comes up that says, "Hey, this game is waiting for you. You want to come back and give it a whirl again?" Because oftentimes people do have a good time, but they're busy, and we all understand this. You get sort of addicted to your email, and Facebook knows how to keep your attention because that's what those channels are designed to do. And games have largely borrowed from a lot of those same tools.

So the short version of my experience at Zynga, without going into a ton of details about all the things that I did when I was there: I came in as a product manager. I had a background in games, I played a lot of games, and I had worked in consulting, which was a skill set they were interested in. That's kind of what got me in the door. Once I got there, I came in as a product manager, and the role of a product manager was typically to take one area of the game—engagement, retention, growth, or revenue—and be assigned to one of those three areas. Over the time that I was at Zynga, I worked in all three of them across Mafia Wars.

Mafia Wars was the product that I got put on when I first joined the company. I was on Mafia Wars for a year and a half, almost two years. I did run it, and I saw that game go all the way up and all the way down. And that was educational to have seen that experience from the inside.

Well, what was the release cadence you guys were doing?

We pushed code on average four days a week.

Wow.

Yes, it was—it was a high, high pace.

What was the team size? Obviously, it went up and down, but...

It did, it went up and down. At peak, it was over 70. We had 10 or 11 product managers, almost 40 engineers, and the rest were art, production, and sort of the glue functions that keep...

40 engineers on Mafia Wars? Unbelievable.

It was.

Okay, so you were saying how you went through kind of all three of those areas?

I came in, worked on Mafia Wars for a year and a half, moved through all monetization—which is funny because later in my career I ended up doing a lot more of that than anything else. It turns out that's what people want to be good at, is making money, right?

Who doesn’t?

Exactly. But I was sitting next to everybody that was doing that job in Mafia Wars, and I learned from them all the tricks. And there were a lot—and they're effective. They’re tools that I’ve really continued to use. Flash sales are a perfect example. The power of a flash sale is very, very real. Also, the negative impact of running a flash sale—the economic hangover that you put your game into by virtue of making the decision to basically pull money forward is what a flash sale is doing for you.

Right. So, for people who might be listening but might not understand what you're referring to as a "hangover," you're talking about when you put your currency on sale, the revenue comes in, but now your players have all this unspent currency, so your revenue is going to dip afterwards.

That's right. Well, until they clear that currency out. If you have a normal game where players give you cash for gold, and then they take the gold in that game and spend it on swords, dragons, ships, and armies—there are two sides, two transactions that have to take place. When you put a flash sale on that first transaction, saying, "We’re going to let you get gold more cheaply," people build up a huge wallet balance of gold in their accounts. It isn't until they spend through that gold that they'll start paying money again. So for the business guys who are sitting there every single day, watching the revenue reports come in—on the day of the sale, you generate a lot of revenue.

In Mafia Wars, just to give you a round number that’s not real,because it’s not a great idea to ever talk about real numbers, if you're making $100,000 a day and you turn on flash sales, we would typically go to 400–500% of our baseline on the days we ran sales. But then, we’d have two weeks of running at 40–60% of our baseline. So we had to do a lot of analysis to make sure that we were, in fact, coming out ahead on those decisions.

Now, what I teach clients is that they should plan for that sell-through. Have a plan for that sell-through as well. So after the flash sale, what are you going to do to drain those wallets of all your players?

Exactly. That was actually the thing we discovered probably halfway through my time there—our ability to manage the hangover was what made us more able to run sales. As the game got older in its life cycle and was distilled down to mostly paying players (but a smaller number of them), we were left increasingly looking for ways to manage the revenue. We had pressure from corporate to help us hit the numbers for the overall company for a given quarter, especially in the run-up to the IPO. All of that financial pressure from being a public company became part of our monthly planning.

We’d say, "Okay, well, FarmVille is... well, FarmVille’s not a good example because they never fell short, but the FrontierVille team is going to fall short this year. So, can you make up some numbers?"

We didn’t fall short.

The FarmVille team—FarmVille made more money, and I think it still makes more money for Zynga than anything else they have.

Of course.

But Mafia Wars in its day was the second-highest-grossing game in the company. And nobody else was close. Those two franchises really led the earliest growth of Zynga, and Mafia Wars was doing it with a combination of economy management and flash sales. That was new, as far as I had seen at that point, anywhere in the gaming ecosystem. Zynga was not reluctant to experiment with monetization mechanics. I mean, we directly sold power. We directly sold power in a game where your ability to compete with another player is determined literally by a single click.

Like, if I am a more powerful account than your account in Mafia Wars, and I go to your page and I click on you, I beat you. And that's it. You cannot beat me because I have spent more money or played more time on my account. So, a deterministic game—and that, especially six or seven years ago, that user experience was absolutely counter to any kind of game that a real gamer wanted to play.

I think that's true today as well.

It’s still true, though not as much, because microtransactions in the last six years have become much more ubiquitous. You see them in AAA titles, you see them in console games.

Yeah, but we're better at doing them without going absolutely pay-to-win.

Absolutely. Now there's nuance around how we monetize in a microtransaction environment that's entirely aesthetic. For example, one of the games we have at Daybreak is King of the Kill, and there is nothing that you can purchase in that game that alters the physics, the likelihood of getting a shot, or the amount of damage you take. Nothing changes the competitive balance of that game. It's entirely about appearance and, you know, the peacocking game of like, “Oh, look, I have a crazy unicorn mask on and a glowing rifle,” but it doesn’t change the gameplay at all.

But there are other games that people argue aren't pay-to-win. League of Legends is a great example.

I tell people the same thing. It gets such a reputation for fair play.

League is totally—if you look at the ban rates on champions over time, you’ll see that new champions are permabanned. They come into the game overpowered. Now, that's partly because it’s harder to balance without live user data, and you have to release a champion, get a lot of games played, and look at the analytics to say, “Oh, this champion is overpowered.” But it’s also because it makes money. When they release an unbalanced champion, everybody rushes to buy it because it feels good to win. If people can pay a little money to get that winning experience, a lot of people will.

If they pay for something that feels good only for a little while, and then it gets rebalanced and they move on to something else—that still feels okay. But there is a definite pay-to-win echo in the League of Legends ecosystem. It’s not as direct as World of Tanks or Crossfire, where you can pay $5 for a golden bullet that kills people in one shot instead of two. That’s really direct and aggressive pay-to-win monetization, but both of those games are making huge money. Huge amounts of money.

Right? And even that is actually a step back from what you're describing in Mafia Wars, because you can have the golden bullet, but if you don't hit me…

That's right. That's right. That's absolutely true.

So, okay, you have this experience on Mafia Wars and it sounds like it was an incredible learning journey for you. And I know that you then applied that in a new way.

Yeah, so the Mafia Wars learning transitioned—still inside Zynga—into a role that really set me up for the rest of my career. I joined the third-party publishing team as the only product manager. My job became working with external studios that we had signed contracts with to help them improve the performance of their games. We structured our contracts in a pretty similar way to most publishing arrangements. We would give you marketing and help you grow your game in exchange for 30% of your revenue. Zynga at the time had 200 or 250 million monthly active users, so there were a lot of users in the Zynga network.

There wasn’t any sort of metrics? Tiered metrics?

There were, actually. So the setup was that Zynga had 200 million or 250 million MAU—a lot of users in the Zynga network. And we basically told our third-party clients, "If you can maintain certain retention targets, we will drive traffic into your game." We’d set you up with a baseline of traffic, always keeping 500 or 1,000 new players coming in so that we could be looking at your numbers.

Almost like a soft launch.

Very much like a soft launch. We took the soft launch approach Zynga had used for its own web titles and mobile titles and laid it down as the launch methodology we advised our third-party partners to use. Now, it was up to them to listen to us or not, but most were very grateful to be getting advice from inside the company that had invented, frankly, a lot of these approaches—the PM methodology that was rapidly being exported from the big three in San Francisco.

We started signing deals with smaller indie developers, 10-person, 15-person, 30-person shops that had made something interesting at a time when everybody thought, "Oh, well, I can make a little web game and see if it turns into money. Let’s see what this does for us." It was an absolute education—there were so many problems. You’d walk in the door and realize how far behind a lot of these organizations were. The biggest failures, over and over again, were organizations that simply hadn’t invested in analytics. They had no metrics. They didn’t know what their products were doing, even after launch. They were relying on Facebook analytics to give them just the most basic numbers.

There's something I've seen quite a few times where a company—and even some big companies—they think they have analytics, and all you do is you just, like, "All right, cool. Send me the numbers." And those numbers never come.

Yes, yes. Like so many things in game development, it's more complicated than it sounds. In order to actually have actionable data coming back to you, you have to have accurate telemetry in place first. You have to have an engineer who has sat there and said, "Oh, this is an event that I care about," and they've created some telemetry that goes into your database. You've got this whole data pipeline that has to be in place. You've got to get engineering in the client, you've got to get data to transmit properly to your backend, you've got data that has to be organized into the database properly, and then somebody who knows how to write the SQL to pull that data out in a way that makes sense has to pull it out into some kind of tool that will visualize it and turn it into a report.

And those four things are actually not trivial skills to hire for, and they're not trivial things to explain to somebody that doesn't understand them. And that was my first learning. So you walk into these companies and you try to say, "Okay, this is going to be really important for you. You guys have to understand what the retention is of your new players. If you can't calculate your retention, you're not really going to be able to estimate what this product is going to look like if you go out and pay a shitload of money for marketing." Because even in the earliest days, that was where your money went. You'd spend half of your money doing development and then get to the point where you're like, "Okay, I'm ready to launch. How do I get users here?" And as soon as you start investigating that, the answer is you pay somewhere between 50 cents and $3 per person, and we'll get them to you. But that's a large chunk of money for an indie studio that's bootstrapped themselves into having something that's ready for launch. And suddenly they realize, "Oh man, I need to get traffic into my game."

And that was why the Zynga publishing program was attractive to people, because that's what we were offering. But what they didn't recognize is that the retention rates that we were setting were based on our own internal comps. I knew what it took to launch a successful game because Zynga had launched both successful and unsuccessful games, and I'd seen the numbers for all of them. And so looking at a day-one retention number tells me 50-plus% of what I need to know to evaluate the prospects of any given game on any platform.

And would you say that that holds to this day?

To this day?

Let me throw some day-one numbers at you. Are you ready?

All right.

20%.

Oh, bummer. Get out.

30%.

Maybe workable, depends on the state of the game. Like a low-quality game where you're like, "Wow, there are obvious holes in this, but it still got 30% day one." I would put one coat of paint on it. One coat of paint. I would go through one time and say, "All right, let's do the flows. Let's look at the whole new user onboarding experience. Let's look at the early game tuning. And let's decide what the biggest, the low-hanging fruit are in those two areas of the game, and let's make those fixes and see what retention goes to." 30%. I wouldn't launch with it. You wouldn't get out of soft launch for me with a 30% D1.

40%.

That's where you're starting to look like you might end up with a product that'll go. It's at the low end of that range. There's also some nuance now that we get into—whether or not you're a free-to-play game or whether you have a paid upfront. If you're seeing 40% on something with a paid download, I'd be really worried still. If you're seeing 40% D1 on a massive scale, like a web-scale social game, that's where I'm definitely like, "Okay, this is something that I'm going to invest in." I would put more than one iteration cycle into improving that first-time user experience, improving the tuning, and improving your marketing. Your marketing flows also impact—the messaging flows and marketing flows impact the likelihood of players to return, which turns into those early retention values. But going from 40%—

Lining up sort of conceptually the message in the marketing to the product that they're going into, or are you saying looking at the attribution of the marketing to see which channels are giving you the numbers?

It's—I mean, it's sort of all of the above. Actually, the biggest thing in the marketing that I'm looking at is the timing. So if you've onboarded and you've granted us email access, are we sending you a one-day email reminder? Are we sending you a seven-day email reminder? What's the cadence at which we're sending you emails? We're activating the channels we have available to try and bring you back to the game. And the greater the number of channels you have available to reach a player, the easier it is for you to actually get them to come back. Whether they come back on day one, or day four, or day seven—if you get them to come back in that first week, you've radically increased the chances they're going to come back in the second week.

And every game that I've ever worked on tends to have a weekly cadence to it. The day of the week matters. Social and mobile games are played more during office hours, basically. People want three to five-minute sessions that they can get while going to the bathroom in the middle of a workday. If you look at other games with longer session lengths—your League of Legends, your PC games—they tend to spike to their highest DAU on weekends. But the cadence is still there, right? It's shifted one way or the other, but you always see this weekly pattern: the two days on the weekend or the Tuesday/Wednesday midweek are kind of the peak user login days. And so getting people onto that...

Bio rhythms of that.

That's right. I call it "breathing." It's the rhythm of your game—it's that weekly cycle of "Can you get players who played last week to come back this week?" And if you can generally keep people playing once a week, that's the minimum level of engagement that is likely to lead to a player being sticky for their lifetime. And engagement turns into monetization. If you have good monetization flows and they pop up at opportunities that players recognize the value in, all you have to do is get them to keep coming back, and they will eventually see those monetization flows enough times that they'll either engage with them, or they weren't likely to.

Or they're going to do something else valuable for your game, like bring in another user or tell somebody.

Or get beat by somebody that did use a monetization flow. There you go. And you can't discount the fact that the whales need somebody to win against. And if somebody is paying you money to win, it helps you to have non-paying players in that ecosystem who also feel enough like they are winners that they want to stick around as well.

Competition.

That's right. That's right.

Getting the telemetry was one of the major areas where you saw these companies kind of stumble. Anything else that you were repeatedly seeing?

Their approach to monetization—we called it "bolting on monetization." There were a lot of companies that would have a great design approach to a new hybrid game idea, maybe a mix between two mechanics or just a reskin of a current game idea, like an invest-express game with a new theme. But they would inevitably get to the point where they were saying, "Okay, we're three months away from launch. How do we monetize this?" And it's just so backwards at that point. You've baked the entire design of your game, and now you're asking for opportunities that fit that design to turn the game into money. Not everything is that easy to align.

For me, the process of thinking about how you want to manage the economy of a game starts with really understanding the session characteristics of the game that you're offering to players. Is it going to be something that they're regularly logging in to but only playing for five minutes at a time? Is it something that they're going to be logging in for one three-hour mammoth session once a week? Or is it somewhere in between? There's a spectrum. So, how are players experiencing the product—that's one piece of it.

The other piece is—this goes back to what we were talking about before—what is the competitive nature of the game? If you're in a game where you have competitive balance concerns, and the tuning, awarding of loot, and progression status affect the players they're fighting against, then your game balance becomes hugely, hugely critical. You have to make a decision at that point—frankly, you have to make a pay-to-win decision. When you're still in design, you have to sit down with your whole design team and ask, "Is this a core tenet of the product that we're building or not?" And if it is, and you decide that pay-to-win is something that you absolutely want to completely avoid, then you have to look at what your cosmetic options are.

But there is, like you said, this nuanced space in between where you're paying to gain an advantage in time that somebody else could still get to.

Yes, we used to call it "slow boating," having the "slow boat" option for everybody. So everyone can still have that same advantage, but it just might take someone else longer.

Yes, that works in games where your design space is constrained in a way that is constrained in the sense that there's a maximum power. You can achieve max level, and a person that gets to max level faster than me by monetizing isn't necessarily having a game experience that's better than mine because I can still get to max level. So I will be competitive eventually, and the psychology of a player in the chase mode is to chase that max level or that max power. And then it will be about skill again, but you have to have that skill space waiting at the end. Players will say, "Oh, it's okay that I didn't get there as fast as somebody else, as long as by the end, it's still an even playing field."

But it's also like a segmentation issue, right? Because if you're able to segment the users within those different kinds of spaces...

Oh yeah. So like matchmaking, right? It becomes more and more important when you're building and designing competitive games. Any amount of pay for power and pay for advancement turns into pay for power over the short term but flattens out; it's sort of an asymptotic line.

Anyone who has played Clash Royale knows what you're talking about.

Yes, exactly, exactly.

But anybody who got ahead too fast...

But anybody who's played Clash Royale also knows the experience of getting to that—getting to the line between two ranks and realizing that everybody in the rank above you has more stuff than you. Now, at a certain point, like when they launched that game, there's kind of a land race. Your early monetizing players are going to sprint ahead of everybody else. And there is value to literally just being ahead because it changes your loot table to be ahead of people you're playing against in that game. And I actually think that their monetization suffered a little bit for it. They went out like gangbusters, and don't misunderstand me, Clash Royale is a really, really, really, really well-built game.

And it seems like they've fixed a lot of the issues as well.

Yes. But even they—I mean, even Supercell, and this is a great example—who are best in class at iteration and testing, they're extremely aggressive about killing products before they ever see the light of day because they only launch stuff that's good. And even they launched with tuning and their sort of competitive bucketing a little off in the first six months of that game. People really noticed it. Everybody I knew who was a non-payer was stuck at exactly the same place.

Yep.

And that, that can hurt your ecosystem. And going back to the point earlier, when you have your whales paying money to try and beat each other, but they need to have somebody that they always win against in order to feel like their money is getting put to good use. Because that's so much of the psychology of how to make monetization land—when players give you money.

Value. It's about showing the value.

And it's about the show more than it is about that. Players have to feel like immediately after they give you money, they feel better than they did before they gave you the money. And that can look a lot of different ways. They can look better. They can literally just be like, "I get to spend 40 minutes running around in the middle of town, showing off my new costume and having people go, 'Wow, those shades look really cool.'" Or it can be the really satisfying experience of popping around a corner and hitting somebody with the instigate rocket and watching their guts explode.

This actually reminds me of something—another kind of mistake on the other end that I've seen. Although not as often as people failing to design for monetization, but just putting monetization everywhere. Just like, yeah, it's just everything in the game is going to approach you. And then you end up with a game where no matter what you buy, you're just being asked to buy the next thing, and you feel bad about it.

And oftentimes, you buy the wrong thing first. You buy the weakest thing, and you don't get the pop—you don't get the experience of, "Hey, I bought something that felt meaningful to me." You get the experience of, "Hey, I bought something, and I may or may not have gotten a pop-up that said thank you, but I don't know what to do with this thing, and I don't feel like it's helping me." And that is a huge problem. If that's your first monetization opportunity for somebody, you're not going to get them back.

How do you go about, in the early phase of design, getting in some of the monetization?

So I'm looking at this—like I said before, I'm looking at the session criteria to try to understand how players are going to be interacting with it, because that's telling me, or helping to inform, the decision about where to put the first pinch. How far into the game experience are we going to go before we present to the player an opportunity to give us money? And so the "pinch" really refers to invest-express style progression games, where you drop in and you're basically just playing a single-player game in your first session, and you go through a lot of mechanics like teaching you how to put a building down and teaching you how to spend some money.

But eventually, at the end of that experience, you're left with some aspirational goal. It's like, "Oh, the next really powerful unit in the game—you're 80% of the way there." And that was kind of the number at Zynga. We always wanted to leave somebody at the end of their first session between 70 and 90% of the way to their next goal, and to give them clarity about how long it would take them to reach that—to finish that last 20%. Because typically, in a Zynga game, it's a timer game, right? You've just built the first three structures on your farm; you're waiting to collect your next round of resources to sell for coins to then build the next round of structures for your farm.

But, so you're in that state where you have something clear that you need to do next, but you don't have the resources to do it. And that's how we like to leave people at the end of their first session because that's the pinch. That's the opportunity where they go, "Oh, I'm out of energy. I can't do anything else." Unless the choice literally is, "Let me log out and wait for the timers to catch up so I can come back and keep playing the game for free," or, "If I give you a dollar, I can keep going right now." And if you get people into the psychology...

Both of those are pretty good outcomes.

Yeah, both of those are... Oh, I mean, if they come back, they're great outcomes. Given the choice, to be totally clear about it: If I'm offered all my players coming back tomorrow or 10% of my players give me money today, I'm taking all the players coming back every single day of the week.

The 80% rule applies just as well to retention, is what I'm trying to point out because if I have something that I'm aspiring to, it gives me a reason to return as long as I bought into the goal.

Exactly. The carrot is the same for retention as it is for monetization. You're asking somebody to monetize exactly in that moment, or you're asking somebody to retain to get that thing at a later moment. But in both cases, if the offer that you're putting in front of them, if the pinch is compelling—the thing that's waiting on the other side of that glass jar is something that those players want—that's the circumstance you want to put people in coming off of that first-time user experience.

Let's talk a little bit about the economy. Once we have these mechanics in place for a particular game, how do we build out an economy that's going to deliver that 80% pinch and kind of just be balanced, delivering the right content to last long enough for players to retain?

The short answer is you're going to test your way to it. At the highest level, the answer for me is that there isn't a prescriptive answer. There isn't a one-size-fits-all "this is the right game economy for everybody to launch with." So my process really is exactly that—it's a process. It starts with understanding the mechanics and looking at who else is using similar mechanics and how they're monetizing. Let's get some comps; let's look at games that are out there in the ecosystem that are doing well. Let's look at how they're making their money and let's look at the balance concerns that are around the decisions of how they make their money, because those two things are always kind of in opposition to each other. How are you balancing the game? And how are you making money on the game?

So what you're saying is it's all about telemetry. And when you don't have the telemetry, you get as much as you can from other apps in the space, if there are.

Absolutely. That's the first step for me in doing any estimates. So if I'm doing a high-level estimate on a new product launch, I don't have any internal data that tells me what that thing is going to do, but I'll go out and look at products that have similar mechanic sets. So if I know—like Warframe is a great example. Warframe is a game that has really been on kind of a slow burn. They're four years old now. They didn't make nearly the money in their first two years that they're making these days. They have iterated themselves slowly into a monetization mechanic that is basically the release of new frames. It's very similar to the release of a new champion in League of Legends. It's essentially a sidegrade. You're not changing your power at all; you're giving people the opportunity to have a slightly different play style based on the equipment that they're choosing to take into the game.

And so that's one set of mechanics that I can look at and say, "All right, well, they're doing extremely well with that particular set of mechanics. Is this product that we're describing—is it coming from a particular IP? Are we trying to match an IP to a set of game mechanics? We have a set of game mechanics we know we want." So I'm always evaluating these design criteria because typically those exist first. Game companies almost always start with a group of people that say, "Hey man, wouldn't it be great to make a space opera?" And you're like, "Okay, space opera. Well, let me just stop and think about this. All right, how are we going to make money on a space opera? Are we going to sell music? Are we going to sell costumes? Are we going to make money on ads from people who come in to watch the music? And we're going to make the music-making free for talented people who want to do it?"

The space opera opera. Space opera opera.

It's an EVE opera. We can sell the music live to EVE ships to run through the battlegrounds. But you see, they're actually launching that as a separate thing. I feel like anything goes in EVE. I'm only sad that five years ago, when I had a friend of mine who first started playing, he was like, "Dude, you should make an account, log in, and set your ship to train, then let it lapse." Because it's passive accumulation of experience. You just have to start a character. It's like Bitcoin five years ago, right? I should have started a character.

I did start a character. I should log in.

Go back. Yeah, you've got gold waiting for you.

Actually, I'm pretty sure they make it so it fills, and then you have to switch over. But I could be wrong. I'm not an EVE player. I gotta say, it was a lot of work.

Yeah, yeah, I heard that.

But it was cool. I mean, I loved how free it was and how much people could just mess with each other. I definitely miss that kind of gameplay.

Yeah. Well, it's very interesting. I mean, so one of the two products we have at Daybreak that is very open-world like that—we launched a game called H1Z1 that was split into two different titles. One is an open-world zombie survival game called Just Survive, and the other is a King of the Hill variant of that game called King of the Kill. That is basically a single-elimination tournament: 150 players on a giant map, everybody tries to kill each other until only one person is left. Very realistic body sim—two people get into a gunfight, and 20 seconds later, somebody's dead. Like, that's pretty much how it goes every time you run into another human. It doesn't last very long. And the survival sim is very similar, but one of the things that makes these games so compelling is you can meet people out in the world and mess with them. Voice comms are on; you're kind of out running around in a very undefined open-world experience. Yeah, if you want to be super stealthy and sneak up on somebody and say "boo" right before you shoot them in the head, you can, and there's something compelling about the ability to have that kind of self-expression. These games...

The emotions you can have as a user are really cool.

Well, and people who are really good at games often like to troll people who are not as good at games. Because it adds a little bit of variety to their game experience when otherwise they just run around killing people all the time. King of the Kill is actually very interesting in that respect because so much of the outcome of a given game is about the randomization of where loot spawns in the level and where you spawn in the level. And that actually adds a lot of variance to the way that games tend to end. It makes it, I think, a little bit more of a compelling experience because it isn't really guaranteed. Like in the NFL, the best team doesn't always win, and that actually makes it more compelling. You have this "any given Sunday" sense that really, in King of the Kill, someone could be a total noob hiding in a closet with a shotgun while everybody else runs around killing each other. If they spawn in the right place, they can make it all the way to the end of that game. If they use their ears, and the second-to-last person standing walks into their room, they hear them out there, and they’re being quiet as a mouse in the closet. If you put a couple of shotgun blasts through the closet door, you can win.

There’s a luck factor.

Absolutely. Absolutely.

So, okay. Let’s imagine that we’re creating a game. We’re going to make a free-to-play zombie killer game. Okay. So we would create a spreadsheet and start looking at some of the competing titles and writing down all the information we get in terms of: what's the first session length, how much do things cost, what's the buy of currency, what are the timers?

Two biggest things for me at a high level: I’m looking at how many installs did they generate, how much revenue did they generate. I’m building a business case, so I want to know who the competitors are out there.

The successful competitors.

Exactly. And sometimes you see a competitor that generated a huge number of installs but didn’t generate a lot of revenue, and that can give you an indication that there’s an appetite—people want to play that style of game.

Now on PC, where do you go for that? And where does that data come from?

It’s a challenge. Steam Spy is one of the best places to go. Steam Spy is actually a pretty awesome resource. So Steam has a public API, and six or nine months ago—maybe a year now—Steam Spy started sampling a 3% sample rate of the public Steam API. That basically allows you to calculate the total number of owners of any given product on the Steam platform. So you don’t get external information. Okay, so Warframe, for example—they do their distribution on Steam, you can get a Steam client, but they also do distribution via their own portal.

Right, you’re not going to know what’s happening...

You’re not going to know their portal numbers by looking at Steam Spy.

But, so Steam Spy is asking some small%age of the users and extrapolating?

They’re not actually asking users. Every single day they’re querying Valve’s public API, which provides that information. Valve is providing a public API that will give you values for like the last two weeks, and Steam Spy just updates that data every day and then has kind of an aggregated data set. It’s not awesome. One of the things I would love to be able to get from Steam Spy is actually the history. And what I can really just get is a snapshot of the last two weeks.

Is that data just free? Do they give that away, or do they sell it?

That’s just free. They have some paid data acquisition, but we’ve never really taken advantage of it. I’ve been lucky in the sense that everywhere I’ve worked full-time has had access to internal data for our own products. That’s the place you start with, with comps. If you have your own data, that’s where you start.

I'm just, I'm trying to give a framework that our listeners could use. They're probably working on a smaller product, although maybe not. And they might not have been through this exercise before. They may not have their own data yet. And I'm thinking, “Hey, they might put together a spreadsheet like this.”

So that's the second step, right? The first one really is like, what did it generate for installs? Because that, more than anything else, tells you the space—the head, like the maximum outcome—of the space that you're in. You aren't necessarily going to replicate the number of downloads that Kim Kardashian's fashion game got. And in fact, that story is a great story, right? That’s a reskin of a game that Glu had already done, but they put Kim’s name on it, and she did some Twitter marketing, and it radically changed the number of installs they were able to get. Then you get a sensitivity range there, right? You get a variance that says, all right, at the low end with no brand, no IP behind this, just a normal amount of marketing, we have the Glu fashion game. And at the high end, with a huge amount of brand and promotion that outperforms, we have the Kim Kardashian fashion game.

But now you can take those numbers and say, all right, I take these projected install values and I plot them against retention rates that I’m estimating, and I can take the retention rates from a different product and say, all right, I know the internal retention rates from when we launched Planetside. And I know that we’re making a zombie-based shooter in an open world. And I have another open-world shooter to base those numbers on, so I'll pull those retention values and say, okay, these are the retention values I’m going to estimate.

So you're sort of creating like a, like a Frankenstein model.

That's right. I’m creating a Frankenstein growth model. I want to know what the potential is for this product. And then I’m taking really high, medium, and low scenarios. I'm saying, like, if we really knock it out of the park, what's the high end of this range on retention? What's the high end of this range on installs? If we really fail and miss our targets, we don’t get any marketing support. How do we take the experience? What’s it going to look like at that low end? And then that's your range, right? You can go to the money guys. At the top of every company—whether it's your investors in your small game company, or whether it's the VC private equity company in your medium-sized game company, or whether it's the president or CEO of your public gaming company—there’s always a story where you have to walk into somebody’s office with a set of numbers that says, I need to justify this decision that we’re about to make.

Give me a year to put this thing together.

Right. Give me a year. This is what it could be. It's an intelligent bet. It's going to cost us a million dollars to make this game. The maximum payback on it is 20 million; we think the realistic scenario is five. You've still got a great ROI. Let's put the money against it.

Okay, I love this. Let's keep going. So let's say you put all that together, you make your case, they say yes. You have a basic game design, but how do you go through that same process for the actual model—the timers, the currencies?

So the next exercise for me is actually plotting out the beats of the new player experience. What is the timeline of a new player that's being onboarded? Particularly, I'm paying attention to what we call "power spikes." Power spikes can apply to things that are not necessarily about player power, but they're basically the moments of joy—the win moments. They're the experiences that players will be driven to seek. They're sort of the chase elements that you're putting in. And these vary based on what kind of game mechanic you have, but it's about looking at the timeline of what the average player is going to experience when they come into that game.

And then this goes back to the thing about session length and what is the expected UX of these other games that you're looking at.

So in a mobile experience, it's got to be quick. People are going to give you 3 to 10 minutes in that first session. If you can get them to play for 30, you've probably hooked them. So you need that power spike; you need that first experience of getting something that feels meaningful to a player to come fairly quickly. You may or may not want to put monetization into that experience early on. That's oftentimes something that we test to see, like, does early monetization turn people off? It's a little counterintuitive sometimes, but there's a strong argument to be made on most of the products I've ever worked on that early monetization can only help you. Even if you're very aggressive and you turn some players off, you'll capture some people on the first day that you probably would never have seen again. And just by virtue of them giving you money on the first day, you increase the likelihood that they're coming back because they gave you money and they're like, "Oh, I put money into that game. I should go check on it again and make sure that I don't waste this value." It's a little bit of a sunk cost experience that's happening for players in that circumstance.

So I've seen a lot of companies really avoid being aggressive on early monetization, and I make the argument over and over again that even in high-performing games, 50% of your players are never going to be seen after that first day.

It's just such an easy task.

Exactly. So test it. And if you get any amount of money from those, the half of your players that were never coming back anyway, that is all upside, and you're not hurting anybody else who was going to stay anyway, by getting some early.

When you test it, follow the cohort, see what happens.

Right, right. See what it does to your retention rates to make sure you're not hurting yourself, but otherwise, it's probably going to be positive on monetization.

Let's take this scenario a little bit further. So we've gone to the managers. We've gotten approval to make the thing. We've kind of put together some monetization mechanics. We've built the beginnings of a model, or at least kind of plotted out the experience, and later we can build a model around some of that stuff. Now we've got the thing in software. Gordon's looking at this, and it's 35% day one. What is Gordon looking at now?

So now I'm looking for two things. With 35% day one, I'm looking at the ratio between day one and day seven. I want to see how steep the drop-off is. So I can actually handle a game that's 35% D1 if it's 20% D7, but that's pretty unlikely, right? Because the shape of that curve tends to be a fairly similar shape. And if your first-day retention rate is that low, your seventh-day retention rate is probably not going to be super high. But some really kind of core game experiences, some deep PVP games in the mobile space, actually do have curves that look like that. Their day one is not high because they're sort of an acquired taste. It's like hardcore fantasy settings and things.

I gotta think the Game of War-style product certainly feels like it could have, with the way they're throwing people into that game.

Well now, and the other thing is life cycle matters here too. If your golden cohort, your very first large-scale cohort after launch, comes in at 35%, that's gonna hurt you. But a lot of games will launch with 40 to 50% D1 and D7 in the high twenties. And if your D30 is above 15%, you can grow on that.

I mean, that sounds amazing on mobile right now.

That's, that's—oh, I mean, yeah. That's a great, that's a great game. I'd get on that any day of the week. Yeah, what company do you work for? So, looking at those numbers, I'm looking at the ratio between D1, D7, D30. I want to see how steep those things are falling off so that I know where in the timeline to focus.

That also goes back to mapping out the timing of the player experience, right? When do we expect the power spikes to be within the timeline of this player experience? Do we expect to hit them with the notifications ask—like, do you want to ask players to give you push notification rights on their mobile device on the first play session? Do you wait three to four or five sessions in? What does the test for that look like?

Do you do it contextually? Do you do it right at the front of the app?

Right. My personal preference is contextually. I like to explain why I'm asking for something. So get them to the point in the game where they're going to want to be notified by something. Have them start a timer, and when they proactively start a timer, put something in front of them that says, "Hey, do you want us to let you know when this timer is done?"

And if you don't have that timer, you've got a bigger problem than where to put your notification.

That's right. That's right. And so contextual notifications matter a lot because the ecosystem has gotten so noisy, and people are savvy to it. They don't like to be drowning in "Oh, look at me, look at me" kinds of notifications on their phone, especially when those notifications are not tied to something really actionable. And so it has to be a pretty high-value flow that the timer is on to make a player really want to come back. And you want to be judicious with how much you're abusing that channel because, take it from somebody that worked at Zynga, you can absolutely destroy a marketing channel if you overuse it.

Well, you're also kind of asking them to delete the app at some point.

Exactly, exactly. You're inviting churn at that point. So, we're in our hypothetical model now, and we're trying to figure out where we're looking at our early retention rates. So I've decided now I'm looking at day one, day seven, day thirty, and I've decided which one is the worst—basically, which one do I want to try to address? And I approach these things differently, right? So, trying to address D1...

Let me ask you about that. So you address the one that's worst, or do you address the one that's first?

Really depends. It depends on how I feel the product is aligned with others—again, it goes back to what comps am I seeing? Am I competing with products in this particular space? If your D1 is low but you expect it to be low, then you're probably going to move on.

So if I have a game with exhaustible content—a game like a tower defense game—and there's 15 maps, and you beat map one and you go to map two, and you beat map two and you go to map three, and by the end of map 15, you're done. You've finished the game; there's no additional content. I know at that point I'm trying to optimize for the early part of the funnel. I'm not trying to build for an endgame. I know that I need to get as much money as possible before people run out of things to do, or I need to get as many people as possible to finish the game so that I can ask them for money to buy more game.

But there are two different approaches based on how we've decided to monetize. If we've got microtransactions all through and it's a single-player game, I'm going to be focused on maximizing the number of people that are in the earliest part of that funnel. But if I'm in a deep MMO, where I'm not really getting into monetization flows until the third or fourth or fifth day, and I'm mostly relying on whale behavior from core players that are going to be around for a long time, I'm a lot more concerned about the late-funnel retention with that kind of game mechanic than I am with the early-funnel retention. But it really does depend on the product, right?

Right. And have you noticed some of these numbers trending in any direction, given all the changes that have been happening on mobile?

It's hard for me because I haven't been in the mobile space for really the last year and a half or two years. I've been much more deeply on the Steam and PC gaming side, so I don't have this kind of everyday access to reading the numbers as they go by, like "The Matrix," to really be understanding the current state of the mobile ecosystem.

I will say that in the time I was at Zynga, towards the end of my career there, when I was watching those numbers, it was obvious that the mobile ecosystem was becoming much more mature and a lot more competitive in the same way that the Facebook ecosystem did. A maturing industry is more expensive to break into. The tech requirements and the user experience requirements are higher, which makes your costs higher. The marketing costs are definitely higher.

Your ability to get promotion from the platforms, to be editor's choice or new and noteworthy on the iOS store, is now worth millions in terms of the marketing value and the user acquisition you get from those things. But you're competing with tons of other people, and now you're playing against huge IPs. You're playing into the wind on timing. If you try to launch the week after Christmas, you'll never get the new and noteworthy slot in the mobile ecosystem unless you're a AAA publisher because they understand the value of all the new devices coming online on Christmas day. That week following Christmas is a feeding frenzy for the mobile ecosystem, and the CPI on ads for user acquisition goes through the roof for that week. So you're competing with the deepest pockets in the space if you try to launch at a particular time and your mechanics match a Clash Royale style game. You've got to be really careful about your decisions with how you're going to market.

So let me ask you this: in the transition that you've made from mobile to kind of PC/Steam, has anything really surprised you?

Yeah. Yes. The difference in the acceptability of monetization mechanics—the sensitivity to pay-to-win is a lot higher in the PC gaming crowd than it is in the mobile crowd. They also proactively communicate with each other in a way that is just...

The audience and the developers?

No, the audience and the audience. So in a mobile game, it's very fractured. Even in a mobile game that has "social flows," they’re not nearly like the kinds of social flows that exist in PC games. People aren’t just voice chatting half the time.

Right, exactly. They’re not voice chatting all the time. They’re not streaming and broadcasting the game to massive numbers of people.

They're not living on the forums the way they do in PC games. But competitive games are the kinds of things that foster that level of activity from their players.

And I think competitive games and pay-to-win...

Exactly, that's why there’s contention. Those ideologies don’t play well together. The idea of wanting to be the best in a skill-based competitive game, and then realizing, "Oh, that guy only beat me because he paid to win," creates a lot of tension.

And the most stark version of this learning came when I did a small consulting project with a company that made Dungeon Defenders. It was a little indie game, really successful for an indie title on Steam. They sold a bunch of units, and it was a $5 DLC-type game. While they were developing Dungeon Defenders 2, they wanted to create a new SKU—basically a premium-priced bundle where, instead of paying $5 for all seven DLCs, you could get the core game and all the DLC in one package for $20.

But the branding and marketing weren’t super tight. The homepage for the new product wasn’t clear, so a lot of the players who bought it already owned the base game and all the DLC. They thought they were buying a sequel, but they got the exact same game they already owned. They carpet-bombed the Steam store page with negative reviews, and when your Steam reviews drop below 50%, your install rates drop off a cliff. People stop buying your game. You have much better tools to get out of that situation in mobile than on Steam. In mobile, you can literally push an update, wipe your review score history, and start over.

Kind of, sorta.

Sorta, but yeah. In Steam, you can't run away from that historical review score. You can, over a long period of time, get players to be happier with the experience that they're having, but it's really, really hard to recover from something that drops below the fold—that's sort of 50% positive. Once you're into mixed or overall negative reviews, it really, really, really hammers—especially for free-to-play games—the user acquisition, the number of new players that you're seeing in the game.

And going back to the example before, how you're modeling what the possible performance for your product is—installs is the first number I'm asking for. If you dry up the source of new players coming into a product that you're still trying to optimize and grow, you're in a world of trouble.

Right? No trickle. No data.

Right, right.

No calibration.

No. And the testing and optimization have to happen fast enough that players can feel the product improving underneath their feet. So even if you have the best plan possible—to say, "All right, we're going to do a soft launch. We have 15 tests that we're going to run. We're going to pick the best winning variant on all of these"—if you dry up all the users that were walking in the front door, and you don't have the ability to run those tests, or you need way more time to run them because you have such a small user base, you've almost already lost by the time you get to that state.

I mean, it's worth going through the exercise because you definitely want to see if you can turn the corner. And there are games—Rocket League is a great example, Warframe is a great example. There are games—even League of Legends—that all started out slow. They were not the huge commercial successes that they are today on the day that they launched.

Rocket League wasn't even Rocket League; it was a different game—it was like Super Rocket-Powered Mega Cars. They only turned that game into Rocket League after they finished the whole build of the first game and launched it. It didn't sell enough units to keep paying everybody's salary in the company; the company went back to work-for-hire stuff. But they were always kind of looking at the numbers on the original Rocket League, and there was just a core group of dedicated players—they had retention.

So they kept going back to, like, "Well, how could we optimize? How could we iterate? How could we improve this?" And that game turned into basically the internal test for Rocket League. And they knew—it was just polish, just polish that took them from the original idea to what is now—I mean, it's definitely a half-a-billion-dollar franchise. I'm not sure if they've crossed a billion dollars yet, but if they haven't yet, they probably will.

Sure. Yeah, that's a great segue for me to ask you a question that I just love. It's a topic that I like talking about, and I think you're very well qualified to speak to, which is free-to-play on console. Obviously, it's here, but it hasn't grown that big yet. It doesn't seem like an opportunity a lot of people are taking advantage of. What do you think of that opportunity?

I think it's constrained to some degree by the platforms. The discoverability—mobile is still way ahead in terms of the quality of the platforms. I have beef with Steam in the PC space as a platform, and I have beef with Microsoft and Sony in the console space in terms of the ability to get good data back from them about your users and what your users are doing. And that's about discoverability and about their management of their marketplaces. Those two things are challenging when your platform is not giving you good access to data and you don't have a good ability to control or influence how you're surfaced and how you show up in their store.

And why do you think that's the case on these platforms? Are they sandbagging on purpose to protect the business that exists? Is it more that it's just new and they're slower to evolve?

To me, it comes down to they want to own the customer. They want to be the last mile in the relationship to the customer. And they don't want—partly, the lip service here is that they're defending the customer experience. They don't want PlayStation gamers to have a bad experience on the PlayStation Store. They don't want them to get ripped off and play a game that asks them for money that they didn't want to pay. And so they make decisions that limit the ability to expose your monetization flows to their in-store UI and require you to use Steam Wallet to deduct funds, which adds basically a third monetization flow to all of the existing flows in your game. Because you have to go top up your Steam Wallet, and then you have to convert your Steam Wallet into premium currency in the game.

If you have to get people to convert three times, it's harder than getting them to convert twice, which is also harder than getting them to convert once. And that's the difference between the mobile space, the web space, and the console space right now. So the number of times you have to get people to jump through hoops radically affects the likelihood of them getting to the end of that funnel.

So, is it me, or is there like a huge opportunity for an Apple or Google to make a high-powered TV thing that sits next to your TV and lets you play amazing high-definition games? And then those games are basically on an open platform where you set it up so that you can actually run an ad on Facebook, and the thing will appear on your console, ready to go.

Yeah, yeah. I mean, so better integration to the shopping experience is one of the challenges.

Yeah, you have to solve the acquisition.

Absolutely. Absolutely. You have to solve servicing and acquisition, right? If you want a free-to-play game to work, you need to actually be able to get hundreds of thousands or millions of eyeballs to see the game. And the digital stores, the PlayStation Store, it's not a default part of even the user experience because users who have bought PlayStations are accustomed to living in a world where they’re not being marketed to—they’ve paid for that hardware. And they pay for the games that they want. So, there’s animosity there, right?

They are getting marketed. I mean, you get marketed to really hard.

True. Now, all those—it's true now. Yeah, I’m dating myself. I’m still on a PS3, and I use it mostly as an entertainment console rather than a gaming rig. I do really most of my personal gaming on PC at this point. But consoles—look, free-to-play on consoles is already here, and it’s going to be really robust in less than five years. It's not all the way there yet, but they're getting way better, and everybody in the gaming world understands the shift to really a software-as-a-service model.

Games as a service is the new model. Microtransactions are 100% here to stay, even in premium-priced products like Overwatch—it’s a $60 box and they've got microtransactions, Destiny, Madden, everything. They’re here, they’re not going anywhere, and companies are now struggling to match them. They’re looking for what is the monetization mechanic that aligns with these products that we’re trying to build and design. Blizzard is doing a great job. EA and their sports games—it's actually very interesting to me. FIFA Ultimate Team was probably the first thing that I ever had whale-like spend on. I spent $60 to buy the box and easily put hundreds of dollars into buying my ultimate team.

Which year, which platform?

This was PlayStation 3, probably the second year that Ultimate Team was out. I want to say 2014, maybe 2015. That was actually my first introduction into the random card pack mechanic. They're trying to also apply Ultimate Team to football. I found it far less compelling in football than I did in FIFA, and I never really put my finger on why. But I think it had a lot to do with the nature of the more determined gameplay in football, where you're running play after play. And the skill of an individual player at any given moment doesn’t necessarily feel like it has an outsized impact on the results.

Most players don’t touch the ball in football.

Exactly. And that’s another thing I don’t think I’ve seen anybody do particularly well—multiple endpoints for the same IP. What is the companion app experience to Madden that actually has some impact on your console experience? Is there a world where—and I do believe that somebody is going to come up with a product that's going to do that well, and it's going to be really compelling to see. Pokemon was such an interesting flash in the pan in terms of—what a flash, right? I mean, it was more of an explosion than a flash. But that was such a craze, and it introduced AR as a concept in a way that I think is going to be beneficial to the whole industry. There's going to be a land rush of people to try and duplicate that success with different IPs and slightly different mechanics, but taking the same kind of core use case and turning it into a new game.

Last question before I let you go. And thank you for being so, so generous with your time.

My pleasure.

What do you think about free-to-play on VR? And, and that might be a platform where they do open it up more.

I mean, I think any platform that has the opportunity for creative expression has the opportunity to be gamified and monetized with a free-to-play mechanic. It’s such an open field. I do think that, again, going back to the PM mind, if you look at the VR ecosystem and you're trying to plot out what a free-to-play success might look like, you have to start with what’s the size of the ecosystem—how many people have the hardware. And so there are ecosystem growth concerns that I have before I’d be really willing to dive in there, because I don’t think that a free-to-play game is the loss leader, right? It’s not the use case that’s going to drive hardware sales to get people into that ecosystem. You need a more compelling use case to grow the hardware base.

And once the hardware base reaches critical mass, then you start to see these products that don’t have a premium price on them start to become viable in that ecosystem. But I think once we get to that point, if we get to that point—because it’s not clear to me yet—GDC last year was just littered with VR startups, everybody trying to do something with it. And there’s going to be a big shakeout in terms of what that turns into from a game experience standpoint. There are lots of questions in my mind about the interaction mode. How are you actually controlling the avatars that you have in these games? Is it a three-dimensional space you’re walking around in? Are you holding a controller in your hand? The physical risks associated with wandering around in your living room and knocking your plants over—a lot of these things.

Have you used the Oculus Touch controllers?

I have. I have very, very briefly. I mean, I've put all the headsets on my brain for a few minutes at different demos. But I haven't spent a ton of time with any of them. I still have, I, I react to low latency. I get dizzy if I leave the headsets on.

A lot of people do.

Yeah. And that's, look, that's a challenge to the hardware growth of the ecosystem, right? Like it goes back to what is the, what's the possible install base that you have available to you? Consoles relative to the PC ecosystem suffer from that a little bit as well. If I was going free-to-play on only one console, it’s, it’s PlayStation 4, because they’ve, they’ve sold the most. I’m not sure it’s doubled, but it’s, it’s a lot more than the Xbox at this point.

Well, it’s, it’s very interesting ‘cause you’re right. Like, so far free-to-play on consoles has been really on whatever hardware has the largest install base. And it’s been kind of like taking World of Tanks and putting it on an Xbox 360 or something like that. But I could see a Sony or a Microsoft getting behind a game and using it to help sell the product because you can give an amazing game away for free. And you could show that on the box, “Hey, this comes with this awesome shooter.”

Yeah.

We’ll see if we get there. Gordon, it was great having you on the show and great having you on-site. This is, it was really cool. I think it actually made for a great interview to be able to look at you and talk to you like this.

I wouldn’t have it any other way. Skype is so impersonal, and we’ve had so many Skype phone calls in the past that this just seemed like the way we had to do it.

Yeah. So, so Gordon and I have done a little work together in the past, and it was all over Skype. So this was our first time meeting in person. And it's been a pleasure. Let's go get a beer.

Let's do it, man.

Thank you for listening to Playmakers Podcast. If you want to show us some love, here’s how you do it. You can leave a review on Apple Podcasts, Spotify, or Google—whatever you use. You can subscribe. You can shoot me an email. Tell me what guests you want to have, what topics you want covered, what you’d like to learn more about. What are you struggling with in your game development process, in your game career, in your game-based adventure or business? I'm here to help you get it done. That’s how we do. Catch you on the next one.