“You can’t imagine the wealth of knowledge here. So much so fast! It’s glorious!” – Cortana, Halo
There has never been a point in human history where we have had so much data of all kinds. “Big Data” has become a buzz word in the last few years. How does big data become useful and actionable? Microsoft has published a new case study that explains how they are using the data gathered from their mega hit video game Halo 4 to improve customer service, increase efficiency and improve the overall game experience.
Halo 4, the fourth installment of the Xbox 360 title, racked up $220 million dollars in 24 hours and had 4 million playerswithin 5 days. Since Halo 4 is played in an online environment, The Halo Services Team is able to collect data on all aspects of the player experience. The size of the player base, the length and frequency of games and the rabid dedication of the gamers means that the collection of analytics available is staggering. This is big data.
OK, you have a huge set of numbers, how can you make that data actionable and valuable? The Halo 4 Team has done some novel things with the information they have, and how they analyze it. On a basic level, they use the data to predict how and when players are online, and use this information to make sure there are always enough servers online to guarantee the best user experience. Conversely, the team can keep things efficient by reading the number of servers when they predict demand will be low. What is amazing is that they are able to make these changes almost in real time, even though they are operating at a massive scale.
How is it possible to analyze such a large dataset, in an efficient and timely enough manner, to make changes happen quickly? They innovated. They shifted focus from storing data, and analyzing it later, to doing useful data analysis as the data is collected. That’s a truly novel approach. Creating software that takes the raw data, as it is collected, and immediately plugs it into various analytics and insights generating engines is a great model. Usually, businesses collect data, and then go back and run queries and mine the data pile. This can be a time and resource heavy process. Eliminating the storage step, by going straight to providing usable metrics, makes the data available quickly.
The Halo 4 Team is using this speedy availability of data in novel ways. They make weekly updates to the game itself, which are informed by trends they see in their data. Improving the user experience by adding features so quickly makes players happy and keeps them playing. Suggestions gleaned from user discussions on the game forums are live within a week in some cases. The normal model is to make massive, infrequent updates, so players may wait a long time to see improvements in the game. They also use their insights to inform their marketing emails to players. If they spot a new trend, they can quickly capitalize on this in their marketing outreach. This is akin to making a suggestion at the drive through, and having it implemented by the time you pull out.
The resulting player outreach, marketing efforts and improved game experience has had a positive impact on the bottom line. It has caused an increase in both player retention and game sales. This shows that big data isn’t a tech trend with limited usefulness. Real time analytics, and the processes in place to take advantage of this data fast, is a solid new business strategy that shows measurable results.