How to Predict Churners in Need for Speed
Today, with the advantages of analyzing billions of data in real time, companies can better understand their customers’ emotions. Especially the tech companies should develop bigdata based models to increase user experience and reduce churn.
At this point, I would like to share my experience about Need for Speed iOS app. In the last 3 days, there was a competition in the game that awards people with a Jaguar sports car. However, the car becomes available after a full dedication to the game with a high probability of spending real money.
My wife and I were playing this game during last 3 days in our spare times on the road, at home etc. After achieving 75% progress in the game I realized that I wouldn’t succeed in a limited time then I deleted the game. On the other hand, my wife completed 95% of the game but the end was the same as me: she deleted the game. The point that I would like to emphasize in that story that using big data technology and a churn prediction model, the game company could keep us playing the game at least my wife.
Here, an algorithm may calculate a churn score of each player. By collecting location data from each device, it is possible to identify the couples (small communities). When one member of the community deletes the app, the churn possibility of other members highly increase. Uninstalling the app can be identified by daily usage routine of the user. Finally, the app can propose extra advantages to rest of the members of a single community.
This is my advice to the Big Data team of EA Games. I am aware that they already follow the application usage routines and offer extra money in each trial. However, there is a substantial need for the development of custom algorithms.