So...how would this work?

The Data Desk's main home would be online, at some page like cbc.ca/data. That page would aggregate all the maps, charts and databases we'd worked on.

We say "aggregate" because much of our work would also appear within cbc.ca news stories; we know that's where the audience clicks most.

Every chart and dataset won't break news, but a key part of the Data Desk's job will be finding original stories in the numbers, that can be told and talked about on all platforms. That doesn't necessarily mean we'll be the ones producing the stories on all platforms; that would significantly slow down the original stories we could generate and likely wouldn't make sense.

It's hard to define a precise workflow without trying it in a pilot, but here's one visual describing how the Data Desk could work:

Note, the audience will play an important role at the Data Desk. They'll be able to search and play with the data themselves, adding to the transparency of our journalism and maybe giving us new ideas.

At times, we can also crowdsource and map information from the audience, such as reports of storm damage.

But precisely...what do you want to do?

We think the best way to develop data journalism on CBC.ca would be with a regional pilot. A pilot would let us experiment to build capacity and find best-practices within CBC News.

Doing it in a region first might seem counter-intuitive, since online resources are more concentrated at the network level. But putting a Data Desk within a region puts it closer to data that touches peoples' lives: grow-op locations, daycare infractions, court delays.

Mapping can also become more relevant and revealing at a local level. Rather than a census map that goes city-by-city, the Montreal Gazette and Global News were able to see the story block-by-block.

We'd still make use of data relevant to a national audience; certainly consumer items like our food price example would interest people across the country. And we'd hope to spark original stories that will get national airplay.

Vancouver is the right place to do it because we already have a popular website, a strong web team, and managers that see the potential of a Data Desk.

What does success look like?

The first step would be getting more data and interactives online, and learning which inexpensive or free tools fit best with our site to make that happen. The number each week would depend on the complexity, but we'd want to aim for at least one a week that would get people linking and talking.

If that seems like too low a goal: consider how much collective knowledge CBC has about putting a story on TV or a guest on radio. By contrast, there are maybe a handful of journalists across the entire network who know how to put a database online.

There will be a learning curve, and that's why this should be piloted.

In the longer term, success would mean projects that get our regional site nominated for journalism awards, and helping meet Rachel Nixon's goal of growing traffic by 10%.

Whether or not the end result is an ongoing Data Desk would depend on many factors; perhaps piloting this will show us a dedicated team is the best way to do it. Certainly that's how several other news organizations are finding success now.

It's also possible that data journalism will become a part of every reporter's job. That's a win too—piloting this would build capacity for that to happen across the network.