When to Sunset and When to Stay With Social Games on Facebook

The Facebook games ecosystem has evolved to the point where most social game developers put effort into planning the end of a game’s life cycle instead of just its beginning. Facebook’s Sean Ryan shares some insight with ISG on what developers are doing when a game reaches the end of the line.

“We’re hitting a phase right now where games are two, three, and four years old,” Ryan tells ISG. “So we’re seeing different philosophies on what you do. A full sequel that is a different game? An expansion pack like we’ve seen with Zynga and Pioneer Trail, or Army Attack with the Desert Area? Or just let it go and watch it decline?”

When to Let Go

Even with games consistently releasing new content and monthly updates, we’ve seen several long-running games reach the point of sunset in the last four months after 2+ years on the Facebook platform. This point comes when the game drops below a financial performance point or when the developer hits an opportunity cost of resources spent maintaining the game that could instead be spent on developing a new game. The financial performance point varies by developer, but the most common “evaluation equation” we’ve heard from developers is LTV > CPI. That is, when a game’s lifetime value of users is greater than the cost per install of a user, the game is healthy. If its performance drops to a point where LTV < CPI, the game is dying.

This was the case for several games developed by ZipZapPlay prior to its PopCap acquisition. Curt Bererton, former CEO of ZipZapPlay and now General Manager under PopCap, walked an audience through the equation during a session at Casual Connect in Seattle last month. In his talk, titled “Finding the Minimum Viable Game and When to Kill Your Baby,” he explained how the developer checked the LTV > CPI equation on the first day of launch, the seventh day post-launch, and then a final time a month after the game’s launch. In general, he says, if LTV > CPI after 30 days, you’re doing it right.

The point at which the equation seems to change dramatically for most developers is after the three- and four-month mark, where we see social games lose a lot of their early traffic. At this point, says Ryan, the game’s LTV is usually very high because only the dedicated paying users remain in the game while non-paying users have left. Depending on how a developer manages the CPI, the game will either thrive on its loyal users, or gradually decline over a period that ranges from six months to a year.

Case Study: ZipZapPlay’s Happy Habitat

The graph above shows the entire traffic life cycle of Happy Habitat as recorded by our traffic tracking service, AppData. At its peak in March 2010, the game had an all-time high of 501,918 monthly active users and 79,921 daily active users. At the point of the PopCap acquisition in April 2011, Happy Habitat got a slight lift in MAU and DAU as the new owner announced plans to sunset the title. The game officially ended June 30, 2011 at the age of 2.5 years.

When to Hang On

A game might have a LTV < CPI situation on its hands when retention drops below 10% and average revenue per daily active user is already low (say, less than 1 cent). That doesn’t always result in sunsetting, however, as we’ve seen developers turn a game around by simply changing the gameplay or adding new content.

When a game introduces a new expansion, two things happen: MAU and DAU spike as new users come into the game, and retention (DAU as a percentage of MAU) falls. This creates a second curve in the game’s overall life cycle similar to its first three months on the market. Like those first six months, the non-paying users will drop out and the dedicated paying users will stay — and hopefully there will be more of those dedicated users than prior to the expansion.

Case Study: RockYou Playdemic’s Gourmet Ranch

At the beginning of 2011, publisher RockYou acquired Gourmet Ranch developer Playdemic. Not too long afterward, the developer released an expansion pack that added a new game mechanic — fishing — the core gameplay experience. As we can see from an AppData chart covering the last eight months, MAU and DAU rose while retention mostly fell during the April and May months when the expansion launched. Throughout June, MAU and DAU fell while retention rose and after three months — in July — the game’s overall retention came out almost 5% higher than the pre-expansion figure. *Note that we see retention falling in the last 30 days on account of RockYou pushing its new ad platform in Gourmet Ranch; this is creating a new influx of MAU and DAU.

What Developers Can Do: Keep it Awesome, Reduce CPI

Social games on Facebook are only just now finding ways to prolong the gameplay experience through expansions and sequels. Early examples of sequels like Playdom’s Mobsters 2 didn’t perform very well, as users seemed reluctant to migrate from an old app to a new one. As RockYou’s Zoo World 2 proves week after week in our rankings, however, developers can get around this complication by running the sequel in the same app ID as the original. Other games, like 6 Waves Lolapps’ Ravenwood Fair, are hoping to find success through franchise extension in new games rather than with direct sequels. Zynga, meanwhile, continues to explore expansions like the upcoming Pioneer Trail add-on for the 14-month-old FrontierVille.

Aside from introducing new content, developers can also change the evaluation equation by reducing overall CPI. This comes from targeting user acquisition toward only the users they want to have in their game. It may seem like an obvious conclusion that you should only spend money on the users most likely to spend money on you; but we still see some games going after a large, general audiences in their first two months on the market. While this “race to critical mass” approach seems to be working for some games like Disney Playdom’s Gardens of Time, most mid-market and small developers struggle to reach this point unless they go through a publisher with a preexisting critical mass.

What Facebook Can Do: Make it Easier to Acquire Users

Facebook scaled back viral channels available to social games last year in an attempt to clamp down on a spam problem. Some of this virality is beginning to return on the platform, now that Facebook feels it has spam under control. Beyond that, though, the platform is working to improve social discovery for games and introduce a better new-user experience. All of this will help developers connect with the kind of users that can support high LTV.

“We focus on greater ability to find the right types of users for a lower cost for your game,” Ryan says. “It’s not just a sunsetting issue — sunsetting in creative fields is life. You move on. [Facebook] focuses on better ways for you to expand, for you to go viral, for you to acquire customers. That should help, whether it’s sunsetting or expansion.”