It is said that with a large enough sample size, behavior in a closed system can be predicted with near perfect accuracy. One of the many ambitions of the new emphasis on big data is the ability to not only acquire that sample size, but store it and analyze it on an ongoing basis.
This process is of particular interest to game developers, as in many cases video games succeed or fail based on their ability to determine with some level of accuracy whether players intend to continue playing and continue trying to meet now challenges, or whether they will quit and do something else. "What if," say those developers "we can predict when and where they are ready to make a purchase and/or when and where they are becoming frustrated enough to quit?"
As it turns out, the answers to those questions are apparently now within reach.
Games as Sales Funnel
The quest to pinpoint the moment someone makes a buying decision and/or abandons their pursuit of a purchase has long been searched for by professionals at all levels of the industry. For many of those professionals, a video game is the ultimate expression of a sales funnel. Making progress in the game is no different than gathering information on the possibility of another purchase. So it stands to reason the player who quits or abandons their journey must have a reason, and if that obstacle can be removed, preferably in real-time, then sales will become easier.
These kinds of analytics are becoming easier and easier with the advent of technologies like Hadoop OLAP. The ability to visually portray a plateau in buying interest is a powerful one, and it can be used to dramatically increase confidence in whatever conclusion is reached with regard to player and customer behavior.
Competition Drives Improvement
With two billion players and 50 terabytes of new data being produced every day, the intensity of competition among developers and publishers has never been higher. Those who subscribe to the value of big data know the answer is somewhere in their rows and columns of information. Properly analyzed, their data will tell them what players like and what they don't and the data will also tell them how to change their product offerings to improve their revenues.
This is one reason why big data is having such a large impact on the video game business. If the answers are recorded in the players' behavior patterns, then every company has a potential competitive advantage in their own databases. First to reach the right conclusion gets a big head start.
Incrementalism vs. Breakthrough
Gradually accumulating data is far more likely to generate gradually accumulating improvement than it is to produce a 'eureka' moment beyond which everything is different. This is consistent with the approach most data-driven companies take. When a pattern emerges, a savvy game company can turn it into a positive feedback loop without a lot of extra effort. Believe it or not, it is precisely this kind of approach that helped establish both the satellite radio and satellite Internet businesses. It wasn't clear what customers really wanted until executives had a chance to look at the results in various test markets. The businesses both improved and established themselves as a result.
For game developers, things are a little easier. If players choose one quest over another, or pick a certain kind of cosmetic upgrade at the expense of all others, it can become rapidly apparent which features are likely to drive sales. The real advantage is in those companies able to take advantage of their new insights quickly, preferably in real time. Automating it is the likeliest next step.
Any business able to measure its customers' reactions to their product is one that is very likely to improve more rapidly than others. This is good news for game developers, provided they can improve their manufacturing and design work quickly enough to capture each opportunity.