The good PM provides a forum for talented engineers to do the right work in the right way. Without it, individuals literally cannot do quality work, and you end up with something like healthcare.gov.
A product manager is not a manager in the traditional executive sense. Traditional corporate culture is tremendously vertical. Since the 1980s, business culture has celebrated the CEO, the manager of a company who knows how to delegate and set a grand vision that the people under him then execute. After the Great Recession and George W. Bush’s “CEO presidency,” delegation and vision no longer seem like enough. The big top-down visions extolled in the vertical corporate model are cheap—the devil is in the details, as the last 10 years of Steve Ballmer’s mission statements should prove.
In the case of bad product management, you still see the problems of underbaked vision. Take the head of game studio 22cans, Peter Molyneux, who’s a bit of a Barnum-esque figure in gaming. I was reading up on his latest project, the Kickstarter-funded “god game” Godus, when I saw a conversation in a Godus beta update video that sent shivers down my spine: Molyneux asks one of his programmers for an update, the programmer starts to talk about the work there is to do, but Molyneux cuts him off, and then suggests that the programmers race each other to completion. The programmer seems adrift; Molyneux seems clueless and imperious. (For contrast, the affable PM-engineer interaction in this Prison Architect alpha update video bespeaks a team that’s honest with each other about things both good and bad.)
Pretty much every new Microsoft product of the last 10 years fell prey to lousy execution of a half-baked vision, from the Kin to the Surface to Windows 8. But you can see difficulties of product management affecting other companies as well. Facebook’s user experience is simply a mess, still not offering any significant search capabilities and now driven by a desperate need to make amends for its disappointing IPO and subsequent share price flat-lining. Google’s user experience seems to be splitting in two, with its traditionally austere search properties bearing little resemblance to the more gaudy presentation of Google Plus, with Gmail and YouTube seemingly caught in the middle.
Google’s technological acumen is second to none, but when it comes to coherent product design, Apple is the one to beat. Google’s purchase of Motorola Mobility hints at its recognition that it would never be able to coordinate closely enough with another company to make the ideal hardware for the Android OS. With user interfaces now several orders of magnitude more complicated than any pre-1980s electronic device, the details are now the vision.
Another way that the vertical corporate model misleads is by creating the impression that executives control the people below them. It’s not President Obama’s fault that contractors made such a hash of healthcare.gov, but it is his fault for not ensuring that there was a qualified person supervising the project end to end—the role Jeffrey Zients has now belatedly taken on. The legacy of Henry Ford is to think of people as cogs in a machine, each cog more or less capable of being switched out at will. In this model, low-level workers are interchangeable, only becoming critical as they rise up the corporate ladder. Far from the case in reality, this is especially untrue in high-innovation sectors, where skills vary so wildly that one programmer can sometimes do the work of 10 others, and the departure of a single key programmer can be far more destructive than the loss of half a dozen managers.
Recognizing that people management is not superior to product management, many startups have separated the two disciplines. Because performance reviews, HR headaches, and recruitment slogging are distractions from actual innovation, these startups—including Google—created dedicated people management positions. Meanwhile, they made product managers into individual contributors or supervisors of small teams defined not by organizational structure but by deliverables.
Because a PM has a disproportionate impact on setting strategy and deciding what work actually gets done, she is the key to success or failure. She can singlehandedly ruin a project. I remember, none too fondly, a PM for whom “soft power” was an oxymoron; he led by yelling, bullying, and backstabbing those who got in his way. He was eventually forced out of the group after the toxic synergy of his personal and professional defects forced an exodus of talent. This goes to show that traditional managerial politics do not produce innovation. The Vanity Fair article makes the telling point that Mayer tended to be more generous to her talented juniors than to other executives—precisely the opposite of the executive back-scratching mentality. To be a great PM, you don’t have to be especially nice to your co-workers. But you do have to value them.