Curating has such a nice, hushed tone to it, conjuring up images of learned scholars debating the finer points of a 12th century vase. Or librarians poring over priceless first editions of Shakespeare’s plays. Which is why quote of the day at the NICAR get-together in Raleigh last week has to be this:
“We’re not building a library; we’re keeping a zoo.”
That’s courtesy of Martin Wattenberg of Google, and one of the creator’s of IBM’s ManyEyes project. The quote is true on so many levels, not least because running a newsroom is akin to herding cats, only more difficult. And that lion-taming skills and access to a ready supply of raw meat can be real assets.
But that wasn’t what Martin was talking about: He was referring to the fact that data projects, or news apps, or whatever we want to call them need care and feeding, or they die. Far too many projects are launched with great fanfare, only to be left to languish within weeks – and once the data in them is out of date, the sites become historical curiosities rather than living resources.
It’s an issue that we ought to address more as an industry, but we often don’t. One of the strengths of Politifact, for example, is that it’s integrated enough into the newsroom that it’s kept alive as a matter of course. The Washington Post’s POTUS tracker seems to be another one; but such examples are few and far between.
During a lively discussion at the conference about the scarcity of news app development teams in US papers, Jonathan Stray made a good comment about how we need to get away from the notion of the story as the (only) basic unit of news. There’s all sorts of reasons to agree with that – and one key one is that a fixation on story as what we produce ties us mentally to a daily (or hourly) schedule of production. At a newspaper (or TV station, or wire service), once we’ve done the story, it’s done. We can write a follow-up, we can chase new angles, but the original story is done.
That’s not the case with a data or news app, which continues to live on after publication – and needs a completely different kind of mentality.
On the 15-hour flight back from New York to Hong Kong, I watched an episode of a show called “The Genius of Design” (as well as eight straight episodes of Glee, but that’s another story) about the role industrial design played in the Second World War. It featured, of course, the Sten Gun, the dirt cheap – but effective – submachine gun Britain produced in the millions when it was facing the threat of invasion. It was designed from the ground up to be easily manufactured, with simple stamped metal components that could be made in small factories, reducing the risk that bombing raids would wipe out key plants. And, over the war, it was further simplified to make manufacturing even easier. The contrast was with the high precision standards of the German Tiger tank, a formidable piece of war machinery that was so over-engineered and expensive to produce that fewer than 2,000 were made.
The point being that, beyond the craftsman mentality of story-writing, we also need to adopt and an industrial design mindset when it comes to data apps. Good industrial designers don’t just create cool new things we want; they think through the entire process of how it can be manufactured, making compromises here and there for efficiency or cost, or taking advantage of some quirk in the factory to build in some new feature.
In the same way, it’s not enough just to design a data app that has a great front end for the public; we need also to figure out an efficient and sustainable way for it to be maintained. That may involve compromises because the newsroom can’t support a more ambitious project; or it may take advantage of some specific skills or structures in the newsroom. But it needs to take a holistic engineering approach.
I’m not sure how to marry a zoo metaphor with an industrial design one; but we do have to move beyond the library metaphor.
There’s more from Raleigh, and I’ll write up some posts in the next few days.