Get to Know Your Mobile Game Players…Before It’s Too Late

This post originally appeared on Playbook, Chartboost's blog dedicated to the business of mobile gaming.

When it comes to significant metrics for understanding your mobile game’s marketability, Lifetime Customer Value (LTV) reigns supreme. Studios have kept the lights running or totally abandoned projects based on this crucial figure. But by definition — LTV is the sum of monetization events over a player’s lifespan — the metric has one fatal flaw: You have to wait until players leave to calculate it. And there’s a lot that happens between when players download the app and when they sign off for good.

Luckily for indie devs, there are ways to understand and predict player behavior early on. Savvy mobile game developers identify metrics that indicate high-value players — such as whether users complete a tutorial or make a purchase in their first session. Those insights pay huge dividends, allowing devs to focus on finding more users with similar characteristics before it’s too late.

Large studios like Machine Zone and Kabam continuously optimize based on what they learn about high-value players — and, as a result, keep players engaged for six months or so. While those studios’ level of analysis is probably beyond the capability of most indie devs, there are certainly ways to identify valuable players before they delete your app.

From shrink wrap to live operations

First, some historical context. If we look back five to 10 years, the bulk of gaming was console-based. Huge teams of developers would spend years working on AAA titles like Call of Duty and FIFA. At the end of the day, the game would be pressed on a CD, shrink-wrapped and stocked on a shelf in Best Buy.

As players shifted to online games, development became much more of a live paradigm. With Facebook games from the likes of Zynga, Playfish and Playdom, not only did you build and launch a game, but you continued to operate it after it was running. You’d create new content, quests, virtual goods and actions. This eventually came to be called live operations and the process allowed game developers to move from a model with shrink-wrapped goods that players paid 60 bucks a pop for, to a game that essentially was free, with a premium upsell model.

Mobile, then, involved a similar evolution — one in which devs had to build for a whole new platform both from a game design perspective and a technology perspective.

So, what does this changing landscape mean for mobile game developers? As the process became more iterative, devs gained the ability to learn about their users during game play. In fact, they had to in order to make money.

Events as signals

Now not only do indie devs need to understand their live players, but they need to be able to predict a player’s value as calculated by his or her actions within the game. In the free-to-play model, users aren’t paying up front, so devs have to prove the worth of those players. And there are different types. Only a small percentage (2 to 5 percent) of your player base actually pays. Others have a viral value, meaning they might not pay, but they encourage other players to come on board. And others still can bring value through ad monetization.

The sooner devs learn about their players, the earlier they can work on engaging those who bring the most value.

Sophisticated developers will now test gameplay to learn what the payers and monetization characteristics look like. Developers who focus on the U.S. market might do a soft launch in Canada, Australia and New Zealand — places where player behavior is similar to the U.S. — to see if they can figure out how to identify high-value players early on. Their goal is to figure out what information indicates if a person is a good user, track that religiously and then go find similar users.

Savvy devs are looking for events that correlate to high percentages of players who stick around and pay. They may look at people who make it through the tutorials or who make it to a certain level of a game, for instance. Let’s say a game has a lot of graphics and sounds, so it takes 10 seconds to load. Devs may want to pay close attention to the patient users who wait for it to load.

One of our customers — a very sophisticated developer — looks at a series of events. The average lifetime for their player is more than a year, but on average, the first purchase happens on day 14. Since that’s too long to wait to get to know players, they identified a series of events in the player’s first day with the game that indicate whether someone will stick around.

What are you waiting for?

Developers who make games that last a long time, such as Game of War or Paradise Cove, have this stuff down to a science because they have to optimize players for a long time. And some go overboard by developing all kinds of mathematical models to predict player behavior. While most indie devs probably won’t go that far, it’s still important to define a basic set of metrics that are good indicators of player behavior.

Start with simple ones like day one retention or payer percentage at day one. Then you’ll start to see patterns and can get more sophisticated as time goes on. Once you get a feel for the metrics that will help you identify your MVPs, focus on how to find more of them — and maximize their fun.