Our experience on a local site is that 75% of traffic is data and 25% is narrative. Yet I imagine that most newspapers spend in inverse proportion.

Don’t get me wrong– I Love narrative. Storytelling is why I got into this business. But our best stories are digital fishwrap in a day unless they get “databased.” It is a verb: “To transform a piece of information that is useful for a moment into something that is useful for ever.” That should be the motto of all news AND information organizations.

Here’s the gazillion dollar question: Which will draw more eyeballs, engagement, comment and therefore advertising?– A three graf story on the City Council meeting wedged onto page 3 of Metro or an exhaustively-linked rundown on the night’s votes, who cast them, and who supported their campaigns? (Substitute nonprofit charity event, soccer game, concert or anything else you like for the council meeting.)

Ideally you have all of the above. But in a world of tough choices, I’ll bet on the data every time. If you provide it in a usable format, the “former audience” will provide plenty of commentary. If you don’t, they’re not likely to spend a lot of time volunteering their services as unpaid “citizen journalists” to prop up your business model.

The only way to really do it is to:

1) get people to write detailed posts.
2) Build prefabed searches on critical keyword combos
3) Turn those news searches into RSS (you can do this on Google News now)
4) Create a synthetic RSS stream that combines multiple RSS feeds centered around a topic of interest
5) If you really want to get fancy, create a tool that autogenerates combinations of search based RSS into a custom synthetic RSS feeds for different topics of interest

All of this is simple tech. It is also doable and perfect for a fast-paced news room.

With the benefit of hindsight, we newspaper people are realizing that we’ve done many things that haven’t been in the best interests of the public we’ve tried to serve. I hope we’re learning.

What is needed is a mechanism that can expeditiously, reliably and accurately categorize content *after* it has been published. And, to be really useful, the categories must be theme-based categories (rather than broad ‘national’, ‘international’, ’sports’, etc.) where the themes persist over a decent period of time.

This is the true ‘Holy Grail’.

PS: I wouldn’t characterize Wikipedia as a “framework desperate for information.” I’d say precisely the opposite - term-based indexing is not, IMHO, much of a framework per se.