Saturday, July 14, 2007

Post-Launch Subscription Curves for Online Games

There appear to be two main types of post-launch subscription curves for online games. (By "subscription curve" I mean the plot of current subscribers versus time.)

The start-big curve is populated by the high-profile mass-market games that cost $20-50M and several years to make. These games start off with a subscriber population in the hundreds of thousands, then decline slowly over time (and may decline significantly when some similar but "newer" game launches). Such games therefore need to spend heavily on pre-launch marketing because they must recoup their development costs as quickly as possible.

The start-small curve is seen for games like EVE Online, RuneScape, and WWII Online. They start out relatively small, with subscriber numbers from 10,000 to ~100,000, then grow slowly over time. These games depend on low (or zero) subscription costs, and accumulate players over time by being perceived as a polished small game (possibly in an underserved niche market).

Most games fit seem to fit pretty closely to one of these two curves. In 2007, however, there were two glaring exceptions, both of which began life on the start-big curve but soon diverged dramatically from it.

The first exception is World of Warcraft. Based on its size and pre-launch buzz, WoW began on the start-big curve. Unlike other start-big games, WoW then grew much bigger very rapidly, and continues to grow though at a much slower pace.

It's tempting to conclude (as some have) that World of Warcraft owes its success to its gameplay, or its content, or its low technical requirements, or its polish, or to some combination of these. But I don't believe that any combination of those things (which other online games have had, at least in part) is sufficient to explain WoW's meteoric rise. As I discuss in The Secret of World of Warcraft's Success, I believe WoW has gained its massive numbers not solely by any intrinsic features (although those are necessary conditions) but by what I call the Hula Hoop Effect: once something hits a critical mass of popularity, the popularity itself makes it even more popular regardless of action quality or content. Like the hula hoop, going viral generates an order-of-magnitude increase in adoption that is otherwise inexplicable. That's not to say that WoW is a bad game; it's only saying that the features and quality of WoW's gameplay aren't enough to explain its abnormal subscription curve.

The other exception to the two-curves theory is Vanguard, which, though entering on the start-big curve, completely skipped the "slow decline after launch" phase and (according to MMOGData) plummeted directly to the "server merge" phase. This, I think, is the dark side of the Hula Hoop Effect: a product or service -- or person -- that's widely perceived (even if wrongly) as unpopular will become even more unpopular.

What's important to note here is that in both the cases of WoW and Vanguard, I don't believe gameplay alone can satisfactorily explain the levels of subscriber adoption and retention.

My guess for the mechanism behind the exceptional performance of WoW and Vanguard is that the core MMORPG gamers are so engaged with these products and so connected to the media and to each other that the Hula Hoop Effect is hugely magnified. That's just a guess, though; it's probably worth further consideration.

All that said, two questions arise.

1. What can an online game developer do to delay or minimize the "slow decline" part of the mass-market MMOG curve? Is there some way that elements of the game's lore can be used (gameplay design? gameworld design? marketing?) to stave off this slow-motion bleeding of subscribers that other games can't? Other than expansions, what can a game developer do to help keep their game constantly fresh and appealing to new subscribers?

2. Beyond just flattening the standard curve for a Big Game, what, if anything, can a game development studio do to induce a positive Hula Hoop Effect for their game? (Sub-question: Would they want to?)

No comments:

Post a Comment