In gaming, analysis is critical and needs to take place in real-time. You would think the pressure of real-time analytics would cause stress for something as complex as video gaming, but that’s not the case with Zynga. At Hewlett-Packard, Co.’s Vertica Big Data Conference last week, Zynga’s Yuko Yamazaki, General Manager, and Joanne Ho, Engineering Manager, joined John Furrier on theCUBE to discuss how a standardized taxonomy and personalization models are the keys to its gaming analytics success.

For critical data, Zynga has a standardized taxonomy as a first tier. The company lets all of its games lock their own game specific tracking into its HP Vertica database and logs all user activities. It can then analyze that data at a later time. Zynga continues to improve on any success with a particular metric logged by a product team that, as Yamazaki put it, “makes sense to standardize.”

Yamazaki also mentioned that Zynga first has product managers and engineers decide what they want analyzed. This gets product managers and engineers thinking about what data that the analytics team needs to log and analyze for it to be able to move the business value forward.

“We don’t just capture everything that happens and then decide. We let them decide first and then click on what makes sense,” said Yamazaki. This process truly creates a data-driven culture at Zynga as everyone from game designers are thinking about what data needs to be analyzed.

Personalization models enhance user experience

Ho explained that in the past, Zynga developed personalization models to enhance user experience. She used the company’s model to increase the install of new games as one example. This model looks at current playing patterns, such as user engagement. It also examines the playing patterns of users’ friends to figure out how likely a user will be likely to install a game. This social graph helps Zynga introduce users to a game that they would be interested in.

Zynga has successfully launched a great deal of games within the past five years. The systems it has in place to collect and analyze data has helped the company identify patterns from one game to another, continue to improve user experience and have repeat success with new games.

See the entire segment below:

photo credit: slettvet via photopin cc