Top 6 Stats for a Social App
Posted on Jul 27, 2011 by Michael Gugel in Gaming
If it can't be measured, it's not worth doing.
A few months ago, I was in San Francisco and was walking home from work with a friend. When we were about halfway there, he made a startling claim:
If it can’t be measured, then it’s not worth doing.
Whether or not you agree, there’s no excuse for failing to keep track of these basic social gaming stats:
DAU
Daily Active Users is the number of unique visitors that came to your app on a given day.
If Bob comes to your app in the morning and then comes back to your app at night, it only counts as 1 DAU.
MAU
Monthly Active Users is the number of unique visitors that came to your app on a given month.
If Sue comes to your app in the first week of the month and then returns to your app during the last week of the month, it will count as just 1 MAU (but 2 DAU).
Engagement
Engagement = DAU / MAU. In other words, if you have 200 DAU and 1000 MAU, your engagement would be 20%. Engagement for most apps is around 15% – 25%.
Pro Tip: Engagement isn’t accurate unless DAU is relatively flat. Otherwise, it will be artificially inflated when an app’s DAU is rising and artificially deflated when an app’s DAU is declining.
Lisa Marino claims that you need an engagement rate of 20% in order to be successful, but I think that’s a pretty crappy way of looking at it. You can (theoretically) have a great app that users play only a few times per month 
Retention
Retention tells you how likely a user is to come back. Let’s say you look at all the users that installed your app on 6/1/2011. How many of those users came back to play on 6/2/2011? What about 6/3/2011? etc.
Pro Tip: Small changes in retention can manifest into HUGE long-term DAU gains.
ARPU
Average Revenue Per User tells you how much money the average user is generating. Typically, ARPU = Daily Revenue / DAU, but you can also calculate it with monthly numbers.
Pro Tip: Breaking ARPU down by geography, gender, age groups and other demographic data can be particularly insightful.
K-Factor
The K-Factor measures the virality of your game. Historically, it was used by medical professionals to measure how quickly a virus multiplied. A virus with a K-Factor of 1 was at equilibrium. Anything greater than 1 meant an exponential increase and anything less than 1 meant an exponential decline. A K-Factor of 1.1 means that the population increases by 10% in each period. A K-Factor of 0.9 means that the population decreases by 10% each period.
Pro Tip: You can spot when someone is bullshitting if they talk about K-Factor without talking about mean time to spread. For example, if the K-Factor is 1.1, does it take 1 day for that to happen or 100 days? There’s a BIG difference!
The easiest way to measure this would be to look at how many installs were generated by feeds (e.g. wall posts) and request channels (e.g. users asking their friends for stuff) and compare that to your DAU.
Additional Resources
- Jon Radoff created this spreadsheet to help model growth curves.
- Kontagent (one of the leading app analytics providers) put together this deck on more advanced metrics.
One Response to “Top 6 Stats for a Social App”
Leave a Reply
Subscribe
Michael Gugel
28. Jul, 2011
Something with high variance and low sample size can’t really be measured, but can still be worth doing.
For example, imagine you custom-tailor the layout of a brochure for a potential client and they sign a huge contract.
Was it really the custom-tailoring that made the difference? You can’t really measure it since that situation will never repeat itself, but if your gut and your experience is telling you that changing the brochure improves your chances, you should do it!