Mike Sego on Monster Galaxy’s Monsterous Facebook Growth

[Mike Sego is CEO of social game and online community developer Gaia Online. Though the developer got off to a false start on Facebook with Ocean Party, Gaia entered the social games spotlight with its second Facebook game, Monster Galaxy. Launched in November 2010, the Pokemon-style adventure game now boasts 18 million monthly active users and frequently breaks 1 million in daily active users. The story below is an “as told by” monologue, drawn from a recent Inside Social Games interview with Sego.]

All my life I’ve been passionate about games — my parents used to call me the “Game Boy” because I’d be hunched over in a corner playing my Game-Boy handheld console for hours on end. I distinctly remember the moment I decided to study Computer Science. It happened when I first learned that the maximum integer you can hold with 16 bits is 65535. I instantly made the connection that that’s the most experience you can earn in Dragon Warrior. From then on, I knew I wanted to build games. But enough about me — this is a story about Gaia, and our path to making Monster Galaxy.

Gaia Online started back in 2003 as a project born from passion. A group of artists and their engineer friend had the idea to combine an internet forum with gameplay and create a place where simply hanging out – browsing and posting — would earn you virtual currency to help advance your status. The site quickly blossomed into a large community of anime and video game fans, and over time broadened even more.

What transformed Gaia Online from being a project among friends to a successful business was the virtual goods revenue model. The team was running out of money, so they put up a link asking users to donate. In order to thank everyone who donated, the team gave donors a virtual item for their avatar as a gift, a golden halo. The next day, donations tripled because people were so excited about the virtual item. This system got formalized and refined over the years, and the team developed a deep understanding of virtual goods monetization.

In many ways, Gaia Interactive has always been a social gaming company, but well into 2009, we were operating pretty much independently from Facebook, doing everything ourselves on our own destination site. We built our own network, we built the games, we built the community and all the features, etc… Gaia Online was a superset of social networking and social gaming, and in trying to do everything ourselves, we were limited in how much we take on.

I joined the company in October of 2009 to help us pick a direction and figure out our next steps. My background had been in developing (fluff)Friends, one of the first successful Facebook games, so during my period of transition into the company, we took some steps towards exploring a Facebook gaming strategy. We decided to port one of Gaia’s games, an aquarium simulation, to Facebook and launched Ocean Party. The game’s potential was limited by being in such a crowded genre, but it was clear there was still an enormous and growing opportunity for games on Facebook. We strongly believed the team and the talent behind Gaia Online would be incredibly applicable to the Facebook platform, but needed to do something different than what everyone else was doing, something unique that few people could do really well.

So, the idea for Monster Galaxy didn’t come from just one place. It was a combination of three different things that came together around the same time.

First off, having worked on (fluff)Friends, I had infinite ideas about virtual pet games. I’d been musing on the ideas of what Fluff 2 would be for years. The original (fluff)Friends was simply a collage creator, focused around decorating your habitat. I wanted to create something more engaging, where each time you get a new pet it is a glorious moment. I thought a lot about a “pet game” with much more emphasis on the “game”, where players don’t just treat pets like stamps, but you invest time into your pets, going on quests and making them stronger.

Meanwhile, the team here had very similar thoughts. Fantasy creatures had long been among the most popular items on Gaia Online. The team building Ocean Party discussed developing new features that would take you out of your aquarium and into the world. Instead of selling your fish, you would raise them, take them on adventures, and unlock new creatures to collect. The more we discussed this, the more it seemed clear that this doesn’t belong in Ocean Party, and should be a totally different game.

Finally, all of us loved console games. As we were discussing potential social game ideas, we did this intellectual exercise to explore all the top-selling console games of the past 30 years and discuss what each would look like on Facebook. How would these games look in that context? Facebook is missing so many of the top video game genres, and we grew to believe that social gamers needed to be exposed to what core gamers already know, and that people who already know core games needed to be exposed to social games. To attract both those audiences, we needed to create a game inspired by one of the most popular video games of all time — and for all the reasons mentioned before, we were the most excited about Pokemon.

From there, the team that built Monster Galaxy was primarily four artists and four engineers working for about four months. In designing the game, it was important for us to think critically about how Monster Galaxy would be uniquely suited for Facebook, incorporating our best ideas, and deviating from Pokemon in some big ways. For example, we currently don’t have avatars. That was one of the early decisions we made, to let players quickly jump into action, and make the primary focus of the game the monsters themselves.

The team worked heads down for about four months, focusing on the artwork, characters, and gameplay, before we got the game into a “launchable” state. In many ways, post-launch is when the real development process begins, because we start having the most important factor we need to make decisions: Data. It’s humbling at times to see how much of a difference a change to a live game can make, from overhauling major things like the game balance, to even small visual tweaks.

For example, one of our initial design philosophies in Monster Galaxy, which became a catch-phrase among the team, was to “put a cat on it.” Whether it was modals, ads, icons, or logos, our initial instinct was to fill the game with cute cat-like creatures. This thinking stemmed from Gaia Online, where this aesthetic has worked incredibly well. So even though the game was about battling monsters, we rationalized that we should stay with cats because it would help us attract both the female audience and the male audience. You know, like Kingdom Hearts.

It turns out, being gung-ho on cats wasn’t always the right decision. We seriously had cats on everything from welcome screen to the tutorial. When we started advertising our game, our marketing analyst found that one ad with a dragon preformed much better  than our ads with cats. So he suggested we change the permissions icon to a dragon. We gave it a try, and it resulted in nearly a five percent increase  in new users clicking Allow. That’s huge. The change was so profound that you can see its impact on our growth in our AppData graphs:

So that was a lesson — you need to constantly question even the minor details. We A/B test quite a lot now. A question that used to be the source of a major team debate, like — “Should you be able to capture a monster that’s a higher level than your monster?” That can be resolved with A/B tests. We tried a number of different options over several weeks, looking at how revenue, retention, and virality differ in each group. The right answer turned out to be “No, unless you’re using a Master Seed” — and that answer resulted in a 20% increase in our overall revenues.

Right about the time we were figuring this out (which we called the Master Seed project), we also were testing advertising at different CPIs for different demographics. We have different countries that represent tiers and then within the tiers, we divide it up by gender, age, and by ad provider. For example, Nanigans might bring in a different quality of user for a certain tier than Adparlor and we work with both of them and spend appropriately.

Anyway, after the Master Seed project, we did a reevaluation on ad spend. In general we optimize our ad spend around two separate metrics: Revenue and Reach. Previously we could only justify acquiring users at a lower cost, which just meant fewer users. Now that we had improved revenues, we tested the limits of our CPIs throughout April, went a little bit over, and then had to ease off a bit. All of our adventures in advertising can be viewed in our graph:

That new incline you see is something recent we tried that’s both ad-spend and gameplay-related. We continually refresh the ad side by working with different partners and updating the CPIs once we determine the quality of the traffic. But at the same time, we recently had a breakthrough on how we were using Facebook’s viral channels, We’ve hit a sweet spot where we drove a lot of new players. Now the challenge is just retaining and monetizing them.

Switching topics, monetizing via in-game advertising is going to be an interesting challenge for Monster Galaxy. With Gaia Online, we kind of pioneered the deep integration model with campaigns like Scion-branded cars players could race, or Nike shoes that make your Gaia avatar run faster. Sponsorships like these allow players to connect with brands in a more meaningful way than display ads that show up along the side. We built Monster Galaxy with this deeper kind of integration in mind; we actually have a sales package available for just about anything that could become part of the game – a character, a place, or even just any item. But being able to sell these packages has been challenging because despite the game’s broad reach, I think there’s less understanding around social game ad campaigns.

Back in 2003, Gaia Online was a pioneer in online gaming, which meant there was less competition, but also meant there was less understanding and less of a support network for what we were building, so we’d have to do everything ourselves – the destination site, the network, the games, all user acquisition, the payment system… you name it. Now with social gaming, the world has caught up to Gaia in many ways, which means we have more competitors in the space. But it’s also a great opportunity because we don’t have to figure everything out ourselves all of the time. We can focus on making a great game and deliver it on Facebook, and we’re able to reach a much broader audience than we could get on our own destination site.

So going forward, we intend to continue with a multi-platform, multi-product social gaming strategy. That means new titles on Facebook in the coming year, and exploring mobile as well. Beyond that, we’ve grown and learned enough to where we’re excited about working with other game developers as well — figuring out if we’re able to help publish games and collaborate. So if you’re reading this and you’ve got a game you want to be huge, come talk to us.

As a final note, I see a huge opportunity for the mainstream gaming industry and the owners of the most popular video game brands to expand into social and mobile platforms. I would have loved to develop an actual Pokemon game for Facebook, and it’s a shame that doesn’t exist. I hope to see the two worlds, social gaming and console gaming, become more connected in the future, and Gaia is in a great position to help make that happen.

[Gaia Online intends to release two more games to Facebook this year, and a mobile title for iOS in September.]