Maybe this is all public already. Whenever I see a business’s numbers I immediately start backing out other interesting numbers. Here is a quick analysis of some numbers for Puzzle Pirates.
Slide 11 of http://www.slideshare.net/capncleaver/metrics-for-a-brave-new-whirled?type=presentation
- 5% landing to registration conversion combined with 4,200,312 registered players equals 84,006,240 visitors.
- Average registrant value is $4.20 so total registrant value (aka revenue) is $17,641,310.40.
- Registration to paying customer conversion is 3.5% so number of paying customers is 147,011.
- Average paying customer lifetime value is $120 which leads to 17,641,310.4. This step verifies the rest of my numbers.
- Slides 31, 33, 34, and 38 are good for understanding virtual item purchase behaviors.
How much should they pay for advertising?
They should be willing to pay $0.21 ($4.20*5%) for a visitor under the assumption that the population of paid visitors will behave similar to their historic population of visitors. If internet advertising rates drop to the point where they could buy traffic for $0.20 for CPC, then they might be able to grow their business profitably through advertising. The assumption of identical behavior of two user populations is not a good one to make without testing, by the way.
Later in the slides, they explain that their LTV figures don’t take any virality into account. If I register and never spend any money but invite 10 people who each spend $100 they count me as a $0 LTV user even though I contributed to their revenue. This invalidates my $0.21 for CPC number. The number is higher as a non-linear function of their virality.
LTV of current users
The customer lifetime value may be wrong because they aren’t counting the fact that currently-active paying customers are not done giving them money yet. This won’t change averages unless their current paying population is large in comparison to their past population of paying population.
Slide 26 says that lifetime average of a paying customer is 14 months. Assume that the distribution of time since acquisition of the current active players is not skewed and half of the LTV of current paying user base is unrealized. That would be 30,000 customers right now (slide 11). Assuming stable LTV of $120 for them, that’s $3,750,000. All other figures are based on the near- $17M values. So the unrealized revenue of current paying customers could be an entire 20% of the entire revenue since 2004 (5 years or so).
The results of the currently paying customers could alter all other revenue numbers and also my CPC threshold above. If current customers end up paying less due to the recession, it could pull down average LTV.
Conclusion
That was fun! I hope my math was correct and valid.
Did I just admit that was fun?
Post a Comment