So NICAR, as usual, was a blast – with a record 600-plus attendees, which either speaks volumes about how much people now realize that data journalism matters, or that word has spread about the wild parties. Not that I’d know anything about that.
Unfortunately, I missed my own panel on business models; the airline gods were firmly against me the entire trip. But we launched Connected China (Chrome, Safari or iPad, please) there, and on time; and I did make it there on time to talk about data-driven beat reporting with Chris Amico, who co-founded Homicide Watch and Matt Waite, one of the creators of Politifact. I’ve written about both sites many times as great examples of structured journalism before, but I’ve never really had a chance to talk through, in public, some of the broader issues that setting up such data-driven – and highly-focuses – sites involve.
There’s certainly plenty to talk about – how to design, build, and modify a CMS to power such a site; the cultural resistance within newsrooms to new requirements for reporting; and so on – but it seems to me that one of the core issues that doesn’t get raised often enough is the question of a site’s mission.
Politifact tracks and fact-checks political statements; it doesn’t do polling, and it doesn’t cover election horse races. Homicide Watch just cares about murders in DC; it ignores other crimes. Connected China has a broader mandate – but then again, it’s focused mainly on understanding power in China, not in any other country.
To be sure, all news organizations have missions and audiences they focus on; but these sites go far beyond that – they pick out a specific activity or area, build a data structure and workflow around it, and then just pound away at it. Homicide Watch doesn’t make a judgment about whether a particular murder is newsworthy or not; it covers it. Likewise, it doesn’t suddenly decide to investigate a spate of poisonings, even if they seem really interesting. It’s not even that the site would deviate from its mission; at some level, it can’t even fit a different type of crime/fact into its CMS or design.
Is that a good thing or a bad thing?
I’d argue: Yes.
There’s much to be said for great focus, and on building coverage around regular updates to a database. It can lead to new – and proprietary – insights. It extends the shelf-life, and hence value, of the work you do (and may yet be the basis for a more sustainable business model). It makes it clear to readers what they can expect. And it serves readers better, by giving them as up-to-date information as possible when they want it, and not when we feel like writing about a subject. It also gives you a better chance of success. As Matt Waite noted:
If you have one short declarative sentence (about what your project/site does) – you have a chance.
It helps, too, in how you design your system and your site. At Connected China, Irene Jay Liu and her team built the data structure concurrently with Ben Fry‘s team at Fathom Information Design, allowing both sides to tack towards a clear single goal. They weren’t trying to design a one-size-fits-all CMS.
But. There’s much to be said for editorial judgment, for being able to change directions and chase new and interesting things. And what if you’re wrong in your original mission? Or you missed something? (As Matt noted, when they first designed Politifact, they didn’t expect they’d be covering multiple elections with it, so they had to go back and retrofit the data structure.)
It seems to me the trick is to be able to marry both sides of the house; more than that, marry multiple partners into a more coherent whole, ideally with some kind of common data structure that helps offer up even more insights. For example, why can’t there be an Assault Watch alongside a Rape Watch alongside a Homicide Watch in DC, paired with a more traditional news organization that can jump in on other types of crime coverage? Why shouldn’t all those microsites – for want of a better word – combine to create a broader picture of what’s happening in a given city or area?
People who are deeply interested in a specific murder, for example, are well-served by the Homicide Watch part of the combined site – and helps pay for it, either with traffic or subscription fees; others with focus on crime in general can surf the broader data trends; and the general news desk attached to all of them gets to jump in whenever there’s something that doesn’t fit the pattern.
If each microsite serves a deep interest in a community – Connected China for China watchers, for example – and as a result covers its cost (or gets close), then the combined value of all those sites, and more importantly, the data within them, provides gravy both in terms of better stories and insights as well as in opportunities to monetize them.
It means keeping an eye on two types of missions – of the microsites, as well as on the collective value of the combination of them. But it seems, at least at a theoretical level, to make sense. Why don’t we try it?