Link: Should Design Be Held Back by a Tyranny of Data?
The Times’ “Bits” blog today chased the two-month-old story of Doug Bowman, who left Google for Twitter in March. Bowman’s story wasn’t particularly remarkable aside from the fact that he has a great reputation and was expected to have quite an impact when he joined the big G in 2006, and the fact that he was fairly vocal after he left Google that, to his surprise, he wasn’t able to have as much of an impact as he would have liked.
I don’t have any direct knowledge of Bowman’s situation, but it shouldn’t be difficult for anyone who’s worked for a technology company like Google (and I’ve worked for two, and turned down offers from two more) to see the big-tech company dynamics at work here. It’s quite possible that the problem started the day the Google recruiter called him, promising him he’d be able to bring a new discipline to Google, spin up his own team, and “have a huge impact” on millions of users. I hate to be cynical about this, but it’s vanishingly unlikely that anyone brought in to work at a company with thousands of employees and years of established history is going to have any kind of material impact on that company unless they are brought in as a C-level executive (and even then, just maybe). And it may not be the case that “having an impact” is the thing that someone should be shooting for in the first place. To me, it seems to reduce one’s career to a series of video game achievements and sets aside the notion of teamwork and team success. I generally question the credibility of any recruiter who uses it to pitch their company to me as a prospective place to work. If you’re a technologist, you should, too.
I didn’t want to let this go without a few comments on the way the writer, Miguel Helf, approached this story. When you’re writing about something that’s two months old, the expectation in the newsroom is that you’ll add something new to the story. If there isn’t any actual new information, an inferential leap of some kind will usually do. So in this case, Helf makes “data” the villain, pitted against the hardworking, sensitive artist who really just wants to make the world a better place. This is, of course, a narrative framework used by all kinds of writers to make their stories more interesting, but for this story it really clouds the truth and does a disservice to both Google and Bowman (both of whom, I’m sure, felt like they were doing the right thing). Is there a place for a talented visual designer within a large company? Certainly. Is Google the kind of place for a talented visual designer to do his best work? Maybe, maybe not, although I’m guessing a lot hinges on your professional and artistic aspirations (particularly if those aspirations are in line with Google’s minimalist aesthetic). Should the managers of a multi-billion dollar business rely upon data analysis to determine whether the changes their designers propose are sensible? Well, duh. Bowman himself, in his farewell-to-Google blog post, even says “I can’t fault Google for this reliance on data”. So there’s an unfortunate, imagined tension between two schools of thought going on here that is mostly made up by the Times writer, Helf.
Helf’s inferential leap that he uses to keep the story viable is the notion “crowdsourcing,” a buzzword which I’m sure he has devoted a brain cell to, but isn’t really applicable here. Google doesn’t crowdsource much, if anything (since it doesn’t have to — it can generally buy or crunch whatever data it might need). The story as Helf tells it doesn’t actually have anything to do with crowdsourcing. What Bowman objects to is relentless bucket-testing (the process through which a big web property introduces a change on a trial basis, measures the impact, and then implements the change more broadly or rolls back the change based on the responses of consumers). But I think that even focusing on bucket-testing misses the point. In a big technology organization there will always be someone down the hall who objects to what you’re doing, and you’ll always spend a significant amount of time driving consensus. In some cases this is a good thing, because so much of what we do hasn’t been done before by anyone, and we want to make sure we aren’t making expensive mistakes. But there’s obviously a point where it can paralyze an organization, and it’s certainly not the case that that environment brings out the genius in everyone. Those people generally go to work for startups, as Bowman did.