Parcells: “Al, I am just not sure how we can win without so many of our best players. What should I do?” Davis: “Bill, nobody cares, just coach your … - Ben Horowitz

Al, I am just not sure how we can win without so many of our best players. What should I do?” Davis: “Bill, nobody cares, just coach your team.

The trainers at Uberversity, where new employees underwent a three-day initiation, began schooling everyone on this scenario: a rival company is launching a carpooling service in four weeks. It’s impossible for Uber to beat them to market with a reliable carpool service of its own. What should the company do? The correct answer at Uberversity — and what Uber actually did when it learned about Lyft Line — was “Rig up a makeshift solution that we pretend is totally ready to go so we can beat the competitor to market.” (Andreessen Horowitz, the venture capital firm where I work, invested in Lyft and I am on its board, so I was keenly aware of the dynamic between the companies — and I am decidedly biased.) Those, including the company’s legal team, who proposed taking the time to come up with a workable product, one far better than Uber Pool 1.0, were told “That’s not the Uber way.” The underlying message was clear: if the choice is integrity or winning, at Uber we do whatever we have to do to win. This competitiveness issue also came up when Uber began to challenge Didi Chuxing, the Chinese market leader in ride-sharing. To counter Uber, Didi employed very aggressive techniques including hacking Uber’s app to send it fake riders. The Chinese law on the tactic wasn’t entirely clear. The Chinese branch of Uber countered by hacking Didi right back. Uber then brought those techniques home to the United States by hacking Lyft with a program known as Hell, which inserted fake riders into Lyft’s system while simultaneously funneling Uber the information it needed to recruit Lyft drivers. Did Kalanick instruct his subordinates to employ these measures, which were at best anticompetitive and at worst arguably illegal? It’s difficult to say, but the point is that he didn’t have to — he had already programmed the culture that engendered those measures.

It is easy to see that there are many ways for leaders to be misinterpreted. To get things right, you must recognize that anything you measure automatically creates a set of employee behaviors. Once you determine the result you want, you need to test the description of the result against the employee behaviors that the description will likely create. Otherwise, the side-effect behaviors may be worse than the situation you were trying to fix.

Over the past ten years, technological advances have dramatically lowered the financial bar for starting a new company, but the courage bar for building a great company remains as high as it has ever been.