Monday, January 23, 2006

More on the MMO numbers game

Man, my WoW related post struck a nerve, both with feedback by email, and on an internal thread at work where I'd posted a similar comment. I guess for many, WoW is almost religion, and you don't go pokin' fun at someone's religion!

While the post centered on WoW, the frustration I was (poorly) trying to express was really about the hype around MMO numbers and the apparent lack of questioning people give them.

It's not unlike the website stats quoted in the dot-com bubble era. Remember all the confusion of page loads vs unique users vs registered users vs user minutes, etc, etc?

Anyhow, someone pointed me to an interesting post by Mike Sellers on Terra Nova, in which they also throw another spanner into the works: That different MMOs have different mechanics in which the frequency & length of play may differ (so concurrent users isn't the thing) and which may use different subscription models vary the revenue per user (so large number of registered users may not matter either, if few of them pay anything). [I can attest to the latter being a concern based on some recent conversations with Korean MMO vendors claiming *crazy* numbers for in-game item sale-based MMO's.]

Anyhow, Sellers and those he interviews, conclude that ARPU/Month or Average Revenue Per User/Month is the best metric, as it lets subscription revenue be compared with game-room revenue or in-game item sale revenue.

I think that's *definitely* better, but it's still not the whole story, for two reasons:

  • Revenue != Profit. I might make $12 per player per month, but my operating costs might eat all of that. Now, perhaps they are assuming (or know) that operating costs are a constant from one MMO to the next, thus it's factored out. However, it's worth including because I'd imagine it will vary as types of games and platforms they are played on change over time. (e.g. An MMO played on phones might have MORE infrastructure to handshake with carriers, and then might need to hand over 40% of the subscription revenues to the phone company. Not sure it's the case, just making an example)
  • Second, and more importantly, there *is* value to the raw audience size, for two reasons. (1) There's a 'critical mass' threshold at which point you get enough users so that they all run into other people on a regular basis, and (2) there's a 'phenomena mass' at which point your app becomes part of pop culture and gets talked about outside it's core audience. (e.g. Justin's post on WoW is the New Golf)

Anyhow, the post is an interesting read. Check it out!

No comments: