Valve isn’t driven by financial forecasts as none of its choices are made by financial analysis, says Gabe Newell.
In an email exchange, as spotted by Gamasutra and confirmed as real, the Valve boss answered questions about the company’s motives. He said any concerns the firm will become “excessively conservative and driven by financial considerations isn’t super significant”. The firm is set to release the SteamOS and the Source 2 engine for free, and Newell recently told Develop how it aims to provide tools for developers to succeed on PC.
“None of our choices have been made by financial analysis, we don’t have financial budgets, and financial forecasting fails badly in environments dominated by exponentials,” he said.
Newell also revealed more about how its famous flat hierarchy works at the company, and how employees are valued at the company.
“At Valve, the goal is to have as efficient a connection between creators and customers. Each person is thinking about how to create as much value for customers as possible. How do we keep people honest? By creating an efficient market for people’s time.
“I may like you, but if you aren’t helping me get shit done, then I will value you less than someone who does. How do I show I value? By spending my own time working with you, and to a lesser extent through our peer comp feedback.
In 2013, Valve’s flat management structure was criticised by former hardware engineer Jeri Ellsworth, who claimed there were actually “hidden management” cliques operating like high school adolescents.
Newell responded to the criticism of cliques by saying the Steam firm was intended to be structured in a way to solve a variety of issues, of which cliques are an instance.
“Cliques or other forms of logrolling tend to fail in the face of properly designed markets,” he said.
“So I tend to worry more about making sure our internal market for people’s time is working correctly than I do specific problems like cliques. When problems do crop up, which they do all the time, we try to solve both the specific problem (“why in the hell did we do the dumb thing?”) and the more structural (“how can we make it less likely that we do this class of dumb things?”).”