Friday 5 September 2014

The only thing that matters is engagement! (Oliver Kern)


Image credit: The Simpson's: Tapped Out
The value of an app install is zero.
For a few years I have been trying to make sense of performance marketing in free to play games. Initially in online, for the last few years in mobile.
What we are seeking, is a player that engages with our game, plays it, enjoys it and - because he likes it - spends some money. What we are doing is paying a bounty and hoping for the best. We have built models that predict the possible lifetime value of a group of users from a certain source and we have built systems that allow us to identify traffic sources in retrospective that are crap, so we can stop them (only after we have paid). But what we are still doing is paying a bounty and with all the data that we are gathering, predicting an ROI. Sometimes it works out (because we are smart), sometimes it doesn’t (because the game is not delivering or the model is not accurate enough). And it is not getting easier to do things right.

Engagement is the new currency
Engagement drives retention, retention drives revenues.
Already in the old days of social games on Facebook, Zynga developed systems in their games that reward loyalty and engagement: Congrats, you have received 200 coins for playing today, come back tomorrow and you will get 500 coins, come back 5 days in a row and you will get 50 Gems that you would have to buy for $2. Today this is common practice and many games will reward a player for returning to their game. Some developers would probably argue that Zynga’s games are not great games and need such tactics, but still many of Zynga's techniques have been applied in games where nobody would question if these are great games.
And for a good reason: The biggest impact on lifetime value comes from strong retention.
If you manage to increase your D30 retention by a few percent points, your LTV can easily go up 30% and more.

Player acquisition should be about engagement
If a game publisher would - instead of paying for installs - be paying money for rewarding engagement, the money he spends would be much more in synch with what he is trying to achieve. It would mean much less risk, as unengaged players mean less cost. So how could we attempt to get to such a solution?
There have been a number of attempts to be more in line with such an approach: Get 20 coins for completing registration or - to bring users deeper into the game - get 50 coins for finishing the tutorial or get 100 coins for completing level 1, etc.
These attempts are flawed and in my opinion miss the point: Engagement is more of an ongoing pursuit than a one-off event.
On top of that, when paying a network, you are bidding against other offers and at that moment it’s not about engagement but about which game gets you the biggest reward.

Free to play is good, paid to play is the new evil
If an app developer reduces his risk when acquiring a user by rewarding engagement, I don’t think it’s a bad thing. Paid to play is already among us, if we pay a network 2$ for an install. It could however be, that you have paid and the player hasn’t even started playing.

No comments:

Post a Comment