I found a very good presentation on Hacker News which outlines some key metrics that are important for "social games". If you're building a consumer app these days, adding some social gaming elements can make or break it, so those metrics are very relevant.
The presentation breaks down metrics into acquisition, retention and monetisation metrics, and distributes their importance around the lifecycle of a game, which goes through the stages of early, mid-game, and mature development.
- CAC (Cost of Customer Acquisition) (i.e. how much it costs you to generate sticky traffic) is the key metric early on. As you move towards mid-game, virality becomes important.
- K-factor (also known as viral coefficient) is the important metric here.
The focus then shifts into retention, and the important metrics become:
- Sessions per user and average session length, which measure user engagement.
- 1-day + 1-Week Retention, and Average lifetime per user, which basically are some foundation metrics of cohort analysis, explained in other places.
Finally, in the mature stage, monetisation and long-term traffic generation become important, leading to the following key metrics:
- ARPU & ARPPU, i.e. the Average revenue per user and per paying user, both of which need to be trended up.
- % of paying users, ideally broken down by traffic source and other actionable splits (e.g. "Reddit brings us lots of users but none of them ever pay").
Action-driven
The presentation also provides some actions to take to improve those metrics, as well as what to measure to calculate them, so it's really quite worthwhile to read through it.
Keep it (or this post) bookmarked. On the topic, if you didn't see it, Dave McClure's Startup metrics for pirates is also a must-read full of ideas for metrics.
Why read about metrics?
Actionable metrics (as opposed to useless vanity metrics) can and should drive your business decisions. They can make the difference between a roaring success and a complete failure. But proper metrics measurements can be very hard to bolt on to a product later on.
Much like test-driven development results in a higher-quality technical product, metrics-driven development results in a higher-quality business. If you know about the metrics and you build them in from day one, and you make sure not to fall into dangerous data traps like this and this, you may end up with a much smoother ride than most businesses.
One big difference between new entrepreneurs and seasoned ones is that the seasoned ones know what needs to be measured and make sure it's measured to the appropriate level of detail and accuracy from day one, whereas new ones need to figure this out as they go along.
Edit: Actually, it looks like this presentation slipped by unnoticed on HN a year ago, so I must have clicked on it in a tweet, rather than found it on HN!
If you read this far, you should follow me on twitter here.
