Over at Strange Attractor, a blog on social media, business and journalism written by Ken and Suw Charman-Anderson, a nice analysis of some of the problems that ail media sites. It’s not that they present any new revelations; but the framing of the issues and paths forward – via adding more context – is a good one.
In a nutshell:
While media companies, especially newspapers, have been cutting staff to cut costs, they have also been creating more content. Digital production techniques make this possible but, again, we’re starting to reach the limits of that strategy. Basically, we have an oversupply of content driving an oversupply of digital advertising space, and traditional markets have one way of valuing a surplus: returns plummet.
Nicely put. Of course, there are other issues for publishers – not least the more targeted (and hence less profitable) nature of online advertising. But the key point made here is that there’s already lots and lots of content out there; while there’s obviously room for new content in particular niches, further flooding an oversupplied market in general is not a path to riches. Or creating value for readers/the public, either.
In essence, the news industry is acting against its own economic interest by producing more content and exacerbating the problem of information overload. It’s like trying to save a drowning man by giving him a glass of water.
And it offers a path forward – not unlike the notion of structured journalism, but with a more technological bent. It’s all about context, they say.
Context is about adding value to content in ways that benefit audiences and advertisers. It makes it easier for audiences to find and make sense of relevant content. Adding context, rather than simply creating more content, is about realising that content is no longer scarce, but audiences’ time and attention is. It helps advertisers by providing opportunities for more highly targeted advertising.
Context, in this case, means finding and surfacing relationships and other information that is related to stories. So a piece on a debate over a pending bill might show as a sidebar the campaign contributions that each of the politicians involved received from lobbyists on both sides of the debate.
This all makes lots of sense; too many stories are too disconnected from each other, and from the broader context of the world. That’s partly because that was the original function of stories; to be self-contained vessels of information, published each day and thrown away (or, hopefully, recycled) at night. But in an online, search- and reader-driven world, that structure now limits its value.
Even the steps often taken to provide more context – timelines, ala Living Stories, or “related stories” links, or other stories in the same location – don’t get over the fact that these are all still stories that generally haven’t been modified in any way. (Living Stories does go some way in that direction, by greying out information that you’ve read previously, but it doesn’t fundamentally modify the stories.)
They offer some good examples of sites and services that are pushing the envelope. One, the Sunlight Foundation’s Poligraft, uses public information about political contributions and marries it to people and entities in stories – pretty much as in the example above. To extract the names from the stories, the site relies on Thomson-Reuters’ Calais service, which has the ability to recognize names, places, concepts, and so on from text.
It’s a smart idea to use the Calais engine to tie two elements together – people in a story and a database on political contributions. The more we build those kinds of contextual relationships, the richer readers are for it.
Of course, that also relies on those databases being up-to-date. And on relevant databases existing. Calais – or the smartest search/context technology around – can’t find what doesn’t exist. And it can’t update outdated information.
I think that’s where structured journalism comes in – in building out newsroom processes that ensure we have a steady stream of updated inputs to databases and that get more information out of journalists’ heads and notebooks and into a database somewhere. That builds the store of data that we can link to, or build applications/interfaces around, as well as ensures those databases are kept as current as possible.
That combination – of human and software engineering – should offer some real advances in offering context to readers.