Social Games are Not Social. They should be called Viral Games. Part I

I recently read a great blog post at Gamasutra that examined how social games like Farmville and Mafia Wars are not really social at all:

“It’s also important to understand something about ‘social games’: Most of them are not social. They tend to be single or multi-player games that use social networks (mostly Facebook) as an easy way to drive player adoption. What the industry is calling ‘social games’ are more accurately described as ‘viral games’. ”

When you dig into the success of these games, you realize why Facebook has been such an incredible opportunity.  Facebook offers incredible distribution for these games, making the formula for success much more viable.  Let’s take a closer look:

If Life Time Value of a User > Cost per Acquisition then the game  will be financially successful.

This is the basic formula for the financial success of any social game.  The interesting part about it is when you realize why distribution on Facebook has made the economics so much more favorable.  Let’s break down the components.  The first is Life Time Value of a User (LTV).  This is driven by 2 key factors.  The first is retention of the users.  In order to make money, you need to get that user coming back and coming back often.  A successful social game on Facebook is able to get over 15% of their Monthly Active Users coming back each day.  An extremely successful game gets that stat above 30%.  Facebook is so successful at driving high retention because it already gets their users coming back.  With over 27 billion minutes spent and nearly 45 billion page views per month, Facebook has quickly become for most people, their home on the Web.  You add the fact that as a social game application developer you can integrate your game into the Facebook core UI with communication points such as News Feeds, notifications and wall posts, you have this incredible opportunity to increase your retention rates by folds versus what they would be as a standalone web site.  Of course, you need a game that is also compelling to play.  However as important as good game play, your social game needs to be well integrated into Facebook’s communication system.

Lifetime value of a user also depends on your ability to monetize game players.  A good social game will convert about 1 to 2% of its Monthly Active Users to paid users who spend on average about $20 per month.  It doesn’t sounds like a high conversion rate and it isn’t.  However with such large scale from the massive distribution on Facebook, it can add up pretty quickly.  I think one way that this may improve is when Facebook rolls out its own payment platform to streamline purchases as how Apple has done with apps on the Iphone.

I have estimated that the LTV of a fairly successful social game is somewhere around $.40 to $.50.  That type of LTV is actually pretty hard to make work if you were a standalone web site because of the Cost per Acquisition of that user is usually significantly higher than the LTV.  However in the case of Facebook distribution, the LTV is actually much higher than the incredibly low CPAs that can be found on Facebook.  To be discussed in the next post.

5 Comments

Filed under Uncategorized

5 responses to “Social Games are Not Social. They should be called Viral Games. Part I

  1. Nice!, discovered your blog on Ask.Glad I finally tried it out. Not sure if its my Safari browser,but sometimes when I visit your site, the fonts are really tiny? However, love your post and will be back.Bye

  2. Andy Artz

    Hey Ben, curious on how you estimate LTV – if we assume 1% convert per month, and each pays $20, then this implies a monthly ARPU of $0.20.

    How do you get from ARPU to LTV?

    Did you include virality in your $0.40-$0.50 estimate for LTV?

    • Ben Sun

      Andy – there are a bunch of assumptions here. First is that a 1% to 2% of user MAU (Monthly Average Uniques) is paying $20. So that equates to $.20 to $.40 per month from a user that comes back. Then you have to figure out what the average cohort analysis looks like. A good game should have a 4 to 6 months usage lifetime. So you can see the cohort analysis looking something like
      Month 1 = 100%
      Month 2 = 50%
      Month 3 = 25%
      Month 4 = 10%
      Month 5 = too small to matter
      Apply those percentages with the $.20 per month and you get an LTV of about $.40.

  3. Hey, I just hopped over to your site online via StumbleUpon. Not somthing I would normally read, on the other hand I liked your thoughts none the less. Thanks for making something worth reading.

  4. Pingback: Kontagent 2.3: Day 1 Retention Visibility « The Kontagent Blog

Leave a reply to Andy Artz Cancel reply