3 ms·

It's the latter. A small number of big spenders ("whales" in the industry lingo) can easily skew average number. Median is a better metric than average here.

by TY 14y ago

objclxt 14y ago
Of course, the social games metric includes a large number of players who are paying nothing. I suspect it wouldn't be quite so low if you discounted those players, which would make it a more fair comparison (as gambling would in the majority of cases require you to actual put some money up).
- davidtyleryork 14y ago
  That's a totally fair point. We think of customers on a revenue spectrum. In social games, that spectrum begins at zero and goes up. For real money games, that spectrum begins at the minimum bet allowed - usually one cent on penny slots. Instead of trying to explain away ARPU by dividing customers into paying and non-paying, we want to draw attention to the inherent monetization problem faced by social game developers. Real money gaming far outperforms virtual play no matter how you slice it. To directly address your question regarding paying users versus non-paying users, traditional social gaming ARPPU is around $30. The real-money multiplier would still be a massive 10x delta.
hollerith 14y ago
>Median is a better metric than average here. Not in this case! Remember that the goal here is to maximize revenue (see footnote) so it is a bad move to ignore the existence of the whales -- because whales represent a big fraction of our potential revenue. Footnote: since most of the costs for this kind of business are "fixed" costs, e.g., the cost of paying the developers to built the site, the way to maximize profit is to maximize revenue.
- TY 14y ago
  I was replying Daeken's comment on why $300 seems to be a high number. This number is not representative of average spent by a majority of users but that's what most people assume when they hear "average $300 of monthly revenue per user". For you to maximize your revenues, by all means go after the whales as that's where the money is. I worked in online gambling industry for a number of years and have seen this happen quite a few times. I.e. an online casino with about 12000 of active users had 85% of its revenues generated by a dozen (as in less than 12) high-rollers.
  - hollerith 14y ago
    OK. Looks like you no longer hold that "median is a better metric than average here," which is what I was taking exception to. >I worked in online gambling industry for a number of years . . . an online casino with about 12000 of active users had 85% of its revenues generated by a dozen (as in less than 12) high-rollers. Thanks for the interesting info.

It's the latter. A small number of big spenders ("whales" in the industry lingo) can easily skew average number. Median is a better metric than average here.

by TY 14y ago

It's the latter. A small number of big spenders ("whales" in the industry lingo) can easily skew average number. Median is a better metric than average here.

objclxt 14y ago

Of course, the social games metric includes a large number of players who are paying nothing. I suspect it wouldn't be quite so low if you discounted those players, which would make it a more fair comparison (as gambling would in the majority of cases require you to actual put some money up).

davidtyleryork 14y ago
That's a totally fair point. We think of customers on a revenue spectrum. In social games, that spectrum begins at zero and goes up. For real money games, that spectrum begins at the minimum bet allowed - usually one cent on penny slots. Instead of trying to explain away ARPU by dividing customers into paying and non-paying, we want to draw attention to the inherent monetization problem faced by social game developers. Real money gaming far outperforms virtual play no matter how you slice it. To directly address your question regarding paying users versus non-paying users, traditional social gaming ARPPU is around $30. The real-money multiplier would still be a massive 10x delta.

hollerith 14y ago

>Median is a better metric than average here. Not in this case! Remember that the goal here is to maximize revenue (see footnote) so it is a bad move to ignore the existence of the whales -- because whales represent a big fraction of our potential revenue. Footnote: since most of the costs for this kind of business are "fixed" costs, e.g., the cost of paying the developers to built the site, the way to maximize profit is to maximize revenue.

TY 14y ago
I was replying Daeken's comment on why $300 seems to be a high number. This number is not representative of average spent by a majority of users but that's what most people assume when they hear "average $300 of monthly revenue per user". For you to maximize your revenues, by all means go after the whales as that's where the money is. I worked in online gambling industry for a number of years and have seen this happen quite a few times. I.e. an online casino with about 12000 of active users had 85% of its revenues generated by a dozen (as in less than 12) high-rollers.
- hollerith 14y ago
  OK. Looks like you no longer hold that "median is a better metric than average here," which is what I was taking exception to. >I worked in online gambling industry for a number of years . . . an online casino with about 12000 of active users had 85% of its revenues generated by a dozen (as in less than 12) high-rollers. Thanks for the interesting info.