The Art of LiveOps

Empowering Collaborative Creators w/ Jay Jodway: The Art of LiveOps S2E10

October 15, 2020 James Gwertzman and Crystin Cox Season 2 Episode 10
The Art of LiveOps
Empowering Collaborative Creators w/ Jay Jodway: The Art of LiveOps S2E10
Show Notes Transcript

When you think about the top game platforms in the world, though Roblox may not immediately spring to mind, it should. With over 15 million games and 150 million players, Roblox is the #1 gaming site for kids and teens, and is growing in popularity across all demographics. Listen as Jay Jodway, Head of Top Developer, outlines how, and why, Roblox empowers their community of creators with LiveOps tools from day one.

Support the show

James G. (00:05):

Hello, I'm James Gwertzman.

Crystin Cox (00:06):

I'm Crystin Cox. Welcome to The Art of LiveOps podcast. Hey, James.

James G. (00:16):

Hey, Crystin.

Crystin Cox (00:18):

So who do we have on today?

James G. (00:19):

Well, so today we have a very interesting, non-traditional guest. We have Jay Jodway, who works for Roblox.

Jay Jodway (00:27):

Hey, so I'm Head of Top Developer at Roblox. I work in the developer relations group. The top developer group focuses on the developers that resemble studios and are established businesses making multiple thousands of dollars a month in revenue.

James G. (00:42):

And I don't think most of our audience necessarily thinks of Roblox as a game platform, but it is. Roblox has millions of games and some of the top games on Roblox have millions of players. And I think you can take some of the top games on Roblox and they're bigger than some of the top games on iOS or Android, so this is a real platform. And the most successful games, guess what? They're doing LiveOps. They're actually running themselves like a free-to-play LiveOps game. And so we thought it'd be interesting to go deep into Roblox and see what lessons we can take away from the Roblox community and share back to our audience.

Crystin Cox (01:21):

Yeah. Roblox is fascinating, and I'm really interested to talk to Jay. I know he works really closely with their developer community, so I think he'll have a lot of interesting stories and insights from a community of developers that are starting to do LiveOps.

James G. (01:35):

Yeah, yeah. So let's see what insights and takeaways Jay has for us.

James G. (01:46):

Because that's something that most people I don't think necessarily recognize is you've got games on Roblox with millions of players, even tens of millions of players, right?

Jay Jodway (01:55):

Oh, yeah. Yeah. I believe we have over 110 million MAU.

James G. (02:01):

Wow. Yeah, so that's a pretty serious scale. And I guess the reason we wanted to chat with you is because a lot of the games on Roblox, despite maybe how they started, they're running full on LiveOps today inside of Roblox. And we find that fascinating and so I'm hoping that that's something we can explore on this interview today.

Jay Jodway (02:22):

Yeah, absolutely.

James G. (02:23):

So let's see. Why don't we... Where should we start? Maybe we can start by talking a little bit about... Well, actually, for people who don't know, and anyone who has kids under the age of 15 probably know... But for those who don't know, what is Roblox? That'd be a good place to get started.

Jay Jodway (02:38):

Yeah. We're a platform, first and foremost. So unlike a lot of other live operations conversations that you've had, we actually enable development from creators in a really large developer and creator community, that uses our studio tool to develop their creations and post them on the Roblox platform, which hosts thousands of games. And across all those games, we have over 120 million MAUs, and we've paid out over 110 million to those creators in 2019 alone.

James G. (03:23):

Wow. And that in revenue you're paying out, those are in-game microtransactions for the most part?

Jay Jodway (03:28):

Yeah, for the most part.

James G. (03:29):

Okay. So anyone who's basically a game in your platform is operating a game. They're doing LiveOps. And so, maybe we can start by talking about what... Let's just maybe pick out a couple games that are representative. What are... I guess, what do you see in terms of the level of growing sophistication among your creators in terms of what they're starting to adopt? Where does a typical creator... I know for [inaudible 00:03:59] games on Roblox, it's pretty simple ones. What's the first baby step they take when it starts to adding the LiveOps stuff. So where do they get started and where does it go over time, might be an interesting way to diving into this.

Jay Jodway (04:11):

Sure. So, as you said, there's a lot of different levels of sophistication of game development, but typically, a creator will publish a game, gather feedback from the community and, really, they leverage their social media channels, whether it's Discord and Twitter. But they're really in tune with what the players. So they take that as a feedback loop and iterate on continual development, just as a mobile game developer would.

Jay Jodway (04:50):

And some of the basics that they're creating are, I would say it starts with, again, understanding what the players want and feeding that content demand. So at a really high level, it could be like fun stuff that you see. Like the developers may put a poll in Discord asking, "What type of content do you want to see next in this update," or "What kind of map," or "What kind of theme should we create?" So a lot of the new content roadmap early on is informed by the community, which is pretty interesting. It's very different than the traditional methodology.

Jay Jodway (05:44):

And what we see is as developers continue to level up in an existing game, they start to adopt successful practices that they see in mobile games, console games. There's a lot of inspiration. And at a basic level, you see things like daily login rewards, developer contests that are, again, hosted usually on social media channels. We're starting to see leaderboards really pick up in popularity. But more so than all of the competitive LiveOps features, we see a lot more cooperative LiveOps, which is pretty cool. And there's certainly less emphasis on aggressive monetization and more emphasis on, "Hey, how do we create more engaging experiences with our next update?" So the thinking is a little different and it's more around just serving the needs of their community and doing it rather directly, which is pretty cool.

James G. (06:58):

I guess part of the fun is because a lot of your developers, at least initially, are not especially sophisticated. There's not a lot of the pretense or the pretension, I guess, that a bigger developer might have. I love that it's just going straight at in a poll and saying, "What do you guys want to see," and just asking. I'm sure they get a lot of feedback that's not very helpful, but, at the same time, that theme of player-focused development and really giving players what they want is a theme we've been really exploring a lot over the course of these two seasons on this podcast.

James G. (07:30):

Because that's something I know, Crystin, you've been from the very beginning talking so much about how... especially coming from the MMO background... that notion of really seeing what's working with players and developing further down that path has been... It is one of the hallmarks of a great LiveOp strategy.

Crystin Cox (07:45):

Yeah. I think that what I'm interested in, too, is you're talking about these developers going directly to players and asking their opinion. And traditionally, in live games, one of the things you fear the most is listening to just loud voices. So do you see them starting to get into that echelon as well, where they start going, "Well, how do we reach players who aren't talking to us all the time? How do we reach players who don't..." Do they look at data? What are you seeing them start to do?

Jay Jodway (08:15):

Yeah, absolutely. That's actually, it's funny that you bring that up, because I got that question from our community. I was talking to one of our developers who makes game called Bad Business. And it's interesting, there's a very vocal community on Discord and our social media channels that isn't necessarily representative of the whole wide player base in a game. So I think they are dealing with that issue and they're seeing it more through social media. And I think the ways that they're dealing with it... and correct me if I'm not answering this question correctly. But the way that they're dealing with that is by just prioritizing what they think is best for the players. So their heads in the right place, but they are sensitive to the feedback that they get. They actually, they really want to make as many people happy as possible, but they learned that that's, in gaming, not possible.

Crystin Cox (09:27):

Yeah. So they're maybe becoming... What I'm hearing is they're becoming more confident in themselves as designers, as well, to take that data, the input they're getting from players, and not just react to... do exactly what they say, but really filter it through their own acumen.

Jay Jodway (09:42):

Yeah, absolutely. So they're learning from... The games are built by the community, and they're learning directly from the community. But they're also starting to learn... and this is kind of something that's happened more recently than not. They're starting to learn from the data. And what they're doing is they're not into deep behavioral or predictive analytics quite yet, but they're understanding how to measure the adoption across their core loops or they're understanding the basics of retention, monetization, engagement, how to measure, trap and iterate. They take that feedback into account and then try to incorporate it into their subsequent releases.

Jay Jodway (10:35):

They have a lot of resources that we provide. We have our developer stats, which is essentially a homepage dashboard for them to observe their high-level metrics. So they'll get a monthly read out. They have a ton of resources, excuse me, available to them from third parties and from partners, such as PlayFab. They're using a couple different tools to serve different needs. Some of them are using these tools to go through a beta to identify bugs and technical issues, looking at server logs and things of that nature, that help them optimize the performance of their game.

Jay Jodway (11:28):

Performance is probably more important than retention or monetization, because if players can't get into the game or if the game is constantly crashing, that's an issue. So they're looking at that early on and then continuing to get more involved in the operating analytics, which would be looking at, "Hey, what is acquisition funnel look like?" "Are players getting through the tutorial? How are they spending their time in game?" And, "What are they buying, and "How do we facilitate a healthy in-game economy?" And again, I think at the end of the day, games come down to the content that you create and the experiences that you enable for players. And that's, again, a major use case based on the data side.

Crystin Cox (12:23):

And how are they feeling about data? I'm really curious to hear how do they, as developers, feel about data? Are they excited about using it? Are they skeptical of it? Are they overwhelmed by it?

Jay Jodway (12:35):

Yes. So we have such a huge community of developers, so it really does depend on not only the team, but the genre and just their level of experience. Is it their first game? Is it their third, fourth, fifth game? So what we're seeing is the new creators are leveraging data really that deeply. Some of the developers who start to see monetization success though, that's when you really start to see the adoption of data really come into play.

Jay Jodway (13:28):

So it starts off with them just being interested in checking out the dashboards. "Hey, how did our last update go? Did we monetize well? Did we engage well actually?" It's typically around engagement. And then from there, we're really starting to see these learning experiences create better, let's say, business practices. So now that they know that... Once they see a lot, like 10,000 concurrents in their game or $10,000 earned on day one of their latest update, that's when they're like, "Oh, okay. Let's use all the data. Let's see what we can find out." And they have different approaches.

Jay Jodway (14:20):

Again, some of them are still really looking at the high level engagement data in dashboard. And then, some of the more advanced developers, like RedManta... who's building a really great game right now called World Zero, and they have prior success on the platform with game called Robloxian High School... they are some of our biggest adopters of the PlayFab analytics. They're really pushing the limits of creating custom events to track behavioral data throughout the player life cycle and measuring the core loops, as I mentioned earlier. So we see a wide range of data adoption, and then that just... We're seeing... It's really cool to see the developers level up.

James G. (15:11):

What's really striking me is, I'm like, "Hey, anyone listening to this podcast, if you're thinking about trying to give yourself an education in hands-on LiveOps go to Roblox." Start at Roblox, because you guys have a platform that provides so much of the basics, you can jump in and immediately start building and operating live game really quickly. And it's like thrown into the deep end in a very hands-on way, which I think is amazing.

James G. (15:38):

You talked about the data and that evolution of going from vanity metrics, where you're just looking at output KPIs, like engagement or retention, to starting to go more upstream and starting to look at the inputs into the processes and the custom events around, you mentioned, tutorial stages or player behavior. That is an evolution I think every successful studio has to go through is to start to learn the difference between the vanity metric or the output end versus the actionable data that actually helps you make better decisions upstream. And that's a hard thing to learn.

Jay Jodway (16:11):

Oh, yeah. Absolutely. There's professionals out there, that's their full-time job. And so these guys, these developers, are... I consider a lot of them to be at near-genius level-

James G. (16:25):


Jay Jodway (16:26):

... because of just all of the different functions that their taking on with just one, two, three-person teams. Yeah, so it's pretty incredible to see. And to your point earlier, you could publish a game in Roblox in, technically, a day, but maybe a week, let's say and see live data immediately. There's no other platform in the world with Roblox's scale, at least, that enables that kind of experience. So you're learning really fast, as early as day one. It's really whenever you're ready to put yourself out there and publish and get that feedback from the community, it can happen faster than any other platform on Roblox.

James G. (17:18):

And I've got this [inaudible 00:17:22] question of, "Are we going to see Roblox acting as training ground for the next cohort of big game developers in the world?" You mentioned, just for example, I think it was RedManta. It sounds like they're building their next... They had success, and now they're building their next big game also on Roblox. So obviously, this is a community they know and they're sticking with. Do you see... Is that a typical pattern? Do you see developers stick with the platform or do you see them... I don't want to say graduate, because that implies that it's different... but just moving on to building out of the need of mobile games or moving to Unity? Do you see this progression or do you see them really doubling down on Roblox and becoming more sophisticated? Because I think Roblox has such a huge audience. I know people right now doing startups, creating new game studios, just for Roblox. Building games and making you their primary platform.

Jay Jodway (18:15):

Absolutely. Yeah. It's definitely the latter. I think that we're going to see the next... And I'll go on a limb and make a bold prediction here. But I believe that we're going to see the next Will Wright and the next Sweeney come from Roblox. These developers are at such a high level of accomplishment. When they see success, they keep building on that success and they keep learning from the community. So they learn from their own experiences. They learn from the broader Roblox community. But they're also looking at what people are playing on console and PC and they're starting to bring those experiences to Roblox. And what that's doing is it's going to enable, or it's going to unlock a lot of, let's say, a new breed of game development leaders. I think they're going to come from Roblox. I say that a lot.

James G. (19:16):


Jay Jodway (19:18):

Yeah, it's exciting.

Crystin Cox (19:19):

So making a slight shift here, I'm curious to talk about Roblox itself. You guys are a service and maybe a game... You're a collection of games. You're a platform. You must have your own relationship to LiveOps in the way that you manage the Roblox community. Can you talk a little bit about that, too?

Jay Jodway (19:50):

Sure. So I could... Maybe you could repeat that, so I could better understand what you're asking. Are you talking about Roblox live operations at the platform level?

Crystin Cox (20:01):

Yeah, I'm curious. Your creators are your players and you provide them a service, right? And so you guys also delivering LiveOps. You are also thinking about how to best serve them and iterate and change. I'm curious what that's like, doing LiveOps for something that isn't really a game. It's a platform for game creation.

Jay Jodway (20:28):

Right. Yeah, we are a platform and we do think about how to enable better operations within each game. So through our studio tools would be one example. So we're thinking of... We're always thinking of ways to make new updates better, promoting content easier, surfacing new content and new experiences on the platform, encouraging a wider adoption of play across the platform. I want to say our developer... or our players play about maybe more than 30 games a month, so they're hopping from game to game. And in the past, Roblox has created not just tools and resources to enable better game development, but ways to LiveOps the platform. So one way of making it easier to move from game to game is an example would be Egg Hunt-

James G. (21:41):

Sorry, it was what?

Jay Jodway (21:44):

Egg Hunt. It was a Roblox-hosted event that was obviously around the Easter season and each develop... We had about 50 developers participate and they created a unique egg that they distributed to players inside of their games. And so you had this collection mechanic that was a platform [inaudible 00:22:12] collection mechanic. So they would go from, like I said, Robloxian High School into Wild West and then into another game, to collect these unique eggs. And this was a something that Roblox facilitates in partnership with our developers. And then, obviously, they want to collect all the eggs so that they can get a cool hat.

Jay Jodway (22:37):

And then, there's also some really interesting stuff happening where we're seeing developers start to partner with each other. This isn't planned by Roblox, but we have a really popular game called Piggy right now, and they're partnering with Jailbreak, which is another really popular game. Two different genres, so I guess not in direct competition with each other, but they are cross promoting content in each other's game. So that's as an interesting way of actually developer-driven LiveOps or promotion rather.

Jay Jodway (23:20):

And then, on the IP side, we're even seeing developers pull in or partner with major movie studios. Adopt Me recently had huge success, over a million concurrents in their game, when they partnered with Scoob. So they're promoting, LiveOpsing Scoob content in their game. [crosstalk 00:23:49].

James G. (23:48):

Okay. That explains... My daughter plays Adopt me and she had a Scooby-Doo dog and she was trying to tell me all about it, and I was ignoring her at the time, But I now... That makes sense now. I was wondering if the developer is ripping off the IP. No, actually, that was a real partnership.

Jay Jodway (24:03):

Yeah, that was a partnership. We were in the background to make sure that everything was going smoothly, but we were really hands off with that. That was driven by their team. And that's something that we're starting to see become more popular on Roblox. Once something really takes off on Roblox, the whole community recognizes it and you start to see that idea propagate. So a year from now, who knows how many game IPs... or movie IPs we'll see in games being promoted.

James G. (24:34):

Well, [inaudible 00:24:35] you talked about cross promotion just now. I'm curious. Are there things you do at that platform level within Roblox that either encourage or discourage those kinds of things. For example, do you have a shared game inventory system where one game can grant an item that another game can consume, for example? There's certain things you might have to do at your level to enable those kinds of behaviors.

Jay Jodway (25:01):

Yeah. So there's actually... We have our UGC Marketplace, which is different than the inventory system that you're speaking to, but it is a crowd-sourced content creation mechanism that allows developers to share their creations and put those creations inside of each other's games. And then, we also have a program that is in development right now that's designed to promote engagement and surface more content and, let's say, developer events... so at a platform level. I just, I'm hesitant to get too deep into that, because it hasn't been released yet.

Jay Jodway (26:01):

But we do have a lot of different sorts on the homepage as well which are LiveOpsed. So one example is our Play Together sort. And the Play Together sort is promoting more multiplayer social play on the platform for games that facilitate that really well. And so we have VIP servers, which we recently discounted. I think we put a 95% discount, so they only cost like 10 Robux to create a private server instance for players to play with each other. So we're facilitating a lot of that in those ways as well.

James G. (26:51):

You mentioned-

Jay Jodway (26:52):

Sorry. Robux is our virtual currency, just to be clear.

James G. (26:59):

Yeah. You mentioned earlier the notion of cooperative. You were saying that your platform tends to leave more cooperative play versus competitive play. And I'm fascinated by cooperative play and community-based experiences. Maybe can you expand a little bit on that? What are some examples of play patterns you typically see that skew towards, for example, live events that are less competitive and more cooperative in nature?

Jay Jodway (27:20):

Sure. So the platform... One thing that Roblox does really well is we have this concept of a lobby and Lobby is your social hub, like a room in the metaverse, so to speak. And what that does is it lets players congregate and plan their play. So imagine being in a theme park or on a playground and being with a group of your friends and you're trying to figure out, "Hey, what should we play? What should we do?" "What kind of game do we want to get into?" And they do that together. So they'll hop from game to game going through these experiences as a group cooperatively, which is very unique to this platform and something that's a really powerful mechanism for promoting social play, and that... The reason I bring that up... It's not a competitive or cooperative LiveOps mechanic, but it does enable that group play and that sense of, "Hey, let's enjoy this activity together." Leaderboards are competitive and a cooperative visualization that is popular.

Jay Jodway (28:53):

But on the cooperative side, we have a lot of games that are mini games and similar to Mario Party where you're forced to complete group goals and engage in more group experiences in order to further progress in a game. So it really teaches a lot about how to play together really well and how to cooperate to accomplish a goal.

Jay Jodway (29:26):

And then, on the Discord servers and on social media channels, you see a lot of cooperation happening in terms of creating content, But, again, that's not really a LiveOps mechanic, so to speak. But I'm talking about a broad range of cooperation, because the amount of experiences that I've seen on Roblox in my six months here is this concept of playdates, meeting up with your friends on scheduled basis and spending time in one particular game. Where we recently hosted a virtual concert that had, I want to say, more than 600,000 concurrents. So there's a lot of different types of experiences that enable kids to find these social connections through different types of play, especially now during shelter in place.

Crystin Cox (30:41):

Yeah, so you talked a little bit before about how a lot of the developers are really leveraging things like Discord and social media to manage their community. Are you seeing that move into the games as well? Are you seeing a move towards cultivating that community inside of the Roblox platform or do you think it really is about connecting with all of these other existing platforms?

Jay Jodway (31:10):

We are. We have in-game chat... Sorry, we have platform chat. But we don't really have... We're not promoting... We don't promote too many social media channels on Roblox, but we are seeing most of that take place on Discord. But we really want to create a safe communication environment and a lot of that should be happening on the Roblox chat.

Crystin Cox (31:50):

Mm-hmm (affirmative). Is that something that developers are also interested in? Are they... Would they like to move that more into a controlled space or are they more comfortable, you think, on stuff like Discord?

Jay Jodway (32:04):

I think I'm pretty comfortable on Discord. It's a really powerful tool to get feedback and there's a lot of interaction that's enabled. But I think it's probably skews towards a little bit more of an older audience. But right now, on the platform, we don't have... We aren't creating a feedback mechanisms directly in the game that you're alluding to. I'd see most of that happening on the social media channels, which, quite honestly, is a good medium for that.

Crystin Cox (32:51):

Mm-hmm (affirmative).

Jay Jodway (32:54):

Does that answer your question?

Crystin Cox (32:56):

Oh, totally. Yeah.

Jay Jodway (32:57):


Crystin Cox (32:58):

So shifting gears a little bit, as you have more and more of your developers take on LiveOps functions and start to use LiveOps for their games, do you have advice for developers that are just starting down that path? What have you seen, because you see a lot of the different games, really work?

Jay Jodway (33:20):

Yeah. I always emphasize over index on engagement. Over index on creating fun experiences. And the advice that I give is play as many games as possible. Most importantly, play your game as much as possible, so that you could really tap into the desires of your player base. There's really no better way to find success on, honestly, any game platform than being in line with your player expectations and knowing what your players want and delivering against that.

Jay Jodway (34:08):

There are also... I'm promoting a lot of... We'll, we're seeing this organically, but I'm also promoting developers to create variety events to think about content cadence, to think about other things like time-limited events and easing them into what we would see on a mobile free to play or console games. So we're seeing more of that. And that's a direction that our developers are actually starting to get really good at. So we're seeing cool stuff emerge as well, like rotating shops, countdown timers. Less so on aggressive, mystery box monetization and, again, more on, "How do we create a fun, engaging experience that brings players back?" And I'm trying to help our developers frame that in terms of a LiveOps of structure, which typically includes time boxing and having an engaging activity that's measurable and that has chase prizing and just keeping that fresh with every update. So I'm hoping to see more, a higher velocity, of updates. So instead of once every couple of months, we'll see it once every couple weeks.

Crystin Cox (35:50):


James G. (35:52):

The Discord thing is interesting, because I do think a lot about... We talk a lot about communities and usually communities... Yes, pf course, Discord is there for communities of players. The notion of the community developers being both part of the community with our own players and also developer to developer creator. You have this interesting three-level structure where you've got your own community of developers and they've got their own communities of players and they're all intermingling. [inaudible 00:36:18], have you thought about trying to have your own forums or have you really decided that Discord is, for example, the best way to let your communities communicate with each other, because it's already there, it's already established?

Crystin Cox (36:28):

Oh, yeah. Sorry. I should have brought that up. So our Dev Forum, it does that. That's what it's for.

James G. (36:35):

Oh, I see. So you-

Crystin Cox (36:38):

Yeah, yeah. So we have a Dev Forum where, if you're an engaged developer, you're able to interact with the rest of the development community and look for feedback, actually solicit help for development. We see a lot of recruiting happening. Honestly, we don't really rely on Discord, but we do rely on our Dev Form, because our Dev Forum is where all of that developer-to-developer conversation starts.

Crystin Cox (37:18):

We also facilitate a dialogue amongst developers through our incubator and accelerator program rather. So we have the equivalent of like summer interns come in. And we have the group of 20 to 40 accelerators, which is a program run by Developer Relations. And they work on a game for three months, get constant feedback, present it to each other on a weekly basis. And it's like a three-month long game jam. And we're facilitating this. Putting all the developers in one place, doing presentations and updates out and play tests on a weekly basis. I'll get involved and our... We have a designer who gets involved and it's...

Crystin Cox (38:16):

At the end of this program, actually, they present to Dave, our CEO, who provides a lot of feedback as well. And it's just really impressive to see how they learn from each other in this accelerator program. And seeing the final project after three months, it's really cool to see that happen. And actually they said directly that the feedback that they get from their colleagues is one of the most important parts of the program.

James G. (38:48):

Yeah. I love this notion of Roblox as [inaudible 00:38:50] for game developers, which is pretty wild. Crystin, do you want to ask our favorite question?

Crystin Cox (38:59):

Yes. So we always like to ask this, because I think it's good for everyone to remember that everyone has disasters when they do LiveOps. So we want to ask every guest we have to share a LiveOps disaster that you have personally experienced.

Jay Jodway (39:18):

Ooh, a LiveOps disaster.

James G. (39:21):

Or in your case, maybe you've seen one of your developers experience, because it's...

Jay Jodway (39:28):

Let me think about this.

Crystin Cox (39:32):

Yeah, as I say, we like to talk about it because doing LiveOps is live and it's iterative, and so everybody deals with that. I think it's good for everyone to remember everyone makes mistakes and even even the most successful people. It's a little humanizing. And it's also just fun.

Jay Jodway (39:51):

Sure. So I don't... On our platform, I actually haven't seen any major destructive issues really pop up, but I can speak to personal experience from my history in mobile free to play, if that's okay.

James G. (40:13):

Yeah, please.

Crystin Cox (40:13):


Jay Jodway (40:14):

So one of many disasters that I've seen would be the issue that just still hurts to this day. We had a game called... Actually, I don't even want to say the name of the game. But it was a Facebook game turned mobile game, 4X strategy game that went pretty heavy on LiveOps monetization sales. And what we did to level the playing field a little bit was we wanted to create hard currency drops in the game as a loot drop. And it was really successful, because it allowed the grinders to get access to free currency in an engaging way. A really simple mistake happened and someone innocently fat fingered an entry in the database, the configuration to the drop rate, changed it to 1,000% instead of 1%. And we had a massive flood of free currency rush into the game, which, for a high spender, that's really going to make you angry when you see all of this free currency devalue what you've spent... not even a high spender, just any spender.

Jay Jodway (41:54):

And we caught it a few hours after it happened. We were all in the office for 48 hours just solving this problem. Rolling back servers, working with the community and PR to rectify the situation. It's just a weird situation where you gave a bunch of stuff away for free. If you had a store and you just gave away a bunch of items for free, how do you recover from that, and how do you keep those people interested in coming back? And so that was a nightmare. And unfortunately, it took a long time to recover from that. But we had good management and our philosophy was to take the long road and to give everyone in the game a, "Hey, we're sorry" package. We rolled back servers. We did these consolation rewards. And then, we just we had to change our development philosophy in that game going forward, because of that mistake that broke the economy. So I am particularly conservative around game economies and ensuring a healthy virtual economy ecosystem.

Crystin Cox (43:21):


Jay Jodway (43:22):

But we did salvage... We did save the game, but we just took a bit of a revenue hit for six months.

James G. (43:30):

Yeah, that's all. That's a small thing. Well, thank you. I just want to repeat, thanks again for coming on this podcast. And I think Roblox is a fascinating phenomenon. I think a lot of people in traditional gaming circles still are not sufficiently aware of it or what impact it's having, and I think that's to their detriment, because I think a lot of really, really pioneering stuff's happening on your platform. And I think to your point about the next Will Wright coming from Roblox, I wouldn't be surprised at all if we see the whole new crop of designers and developers growing up more quickly in your... I won't want to call it a sandbox... your environment where they can try things at much higher iteration speed and learn much faster. Because, ultimately, it's all about how fast you can learn and try new things and I think your platform does that in a really accessible way, which is super exciting.

Jay Jodway (44:16):

Thank you.

Crystin Cox (44:23):

Thanks for listening to The Art of LiveOps podcast.

James G. (44:25):

If you liked what you heard, remember to rate, review and subscribe, so others can find us.

Crystin Cox (44:31):

And visit for more information on solutions for all your LiveOps needs.

James G. (44:35):

Thanks for tuning in.