How Valve and Zynga utilize data differently when building business models

Both Valve and Zynga focus development efforts around data collected from its users, although Valve has utilized this information for positive growth while Zynga has done poorly, according to a post on Wired by Creative Industries director Hasan Bakhshi and Nesta research fellow Juan Mateos-Garcia.

According to Bakhshi and Mateos-Garcia, Valve has "made a science" of its pricing models, and coupling these models with its "experimental approach to online sales" have made the company more profitable on a per employee basis "than even Apple and Google."

The pair writes that although Zynga has sought to maximize its revenue through gameplay optimization driven by user data, the company has been "haemorrhaging talent and users" since its 2010 IPO filing citing company value at $1 billion. Zynga's "obsession" with building monetization models on data may have provided the company with high short-term revenue, Bakhshi and Mateos-Garcia argue, but the system "burns out" users and employees.

"It appears that contrary to conventional wisdom, being data-driven is not always better," the pair writes. "Which begs the question: how can businesses make the most of their online data, to become more efficient and more innovative, without alienating their users and employees?"

According to the writers, companies need to be smart about analyzing the data they collect and implementing strategies from it. Data collected in a survey conducted by Creative Industries's independent research charity Nesta suggests that companies need to invest in areas that compliment their findings, such as hiring employees with specific skills or changing the culture of their business. Bakhshi and Mateos-Garcia use Valve as an example, citing Valve's workplace organization without hierarchies as one way to empower employees and let those who have the knowledge to address specific data trends do so.

"While Zynga's experience shows that you can have too much of a good thing in data, that's not a reason for not trying to use it," they write. "You just need to know how, which is why we think more and better research about data is so important."