Getting Started
What a data unit is, and who staffs it.
“Data unit” means different things to different people. Here, it’s a team inside a newsroom that uses data and modern data tooling to do reporting the old tools can’t. That’s the whole idea.
You might already have one without calling it that. Maybe your data people are scattered across desks, doing the work without the shared title, the standing meeting, or the box on the org chart that would make them a team. This guide is about giving that work a shape: how big the team should be, where it sits, and how it runs.
Data journalism is reporting done with data and data tooling, whether that’s collecting a dataset, modeling it, visualizing it, or building the thing readers click on. A data unit is the team that does it.
DUG is free to read, thanks to the Reynolds Journalism Institute, which, being in the business of keeping newsrooms alive, has an obvious interest in this one working out. More on the project is on the About page.
The basics
How big and how shaped depends on your budget and your ambitions. A lot of what’s in this guide comes from a survey we ran of data teams across the field, everything from one person working part-time to staffs of twenty-plus. Most small and nonprofit newsrooms land in between, and that’s fine. A team of one or two can do real work.
When we first wrote this guide, a small newsroom with in-house data capacity was a novelty. It isn’t anymore. Local news kept collapsing, nonprofits grew into the gap, and the tools got cheaper and better. Data capacity is now one of the clearest ways a small newsroom produces something nobody else in its market can. That is the opportunity this whole guide is about.
A data unit tends to be built from four kinds of people:
- Researcher-reporters,
- Visual journalists,
- Full-stack devs, and
- Data editors.
At a small shop, one or two people cover several of these at once; the smaller the newsroom, the more each hire has to stretch.
The best data teams are wired into the newsroom, not walled off from it. They usually report to an editor, often sit under investigations, and take work from across the building: analysis, cleanup, graphics, the occasional news app. The most common complaint we hear is the flip side of that reach, which is too many projects and too many bosses. Keep an eye on it.
Most teams settle into a preferred stack, though it varies by shop, and much of it is open-source and yours to adapt. As one editor told us: “We’ve got standards for data integrity and bulletproofing, but there’s no one way to do an interactive data project, any more than there’s one way to write a story.”