J-teams, let’s move from silos to pods

 

As a programmer-journalist studying journalism/computer science, I’ve found myself at the “intersection of journalism and technology” or in the emerging field of computational journalism.

I used to think the technical stuff was the hard part. I spent my time trying to grasp scope in JavaScript, scraping webpages or learning the MVC framework of the week. But here’s the thing — programming? That’s actually the easy part. (N.B.: That’s not to say it’s trivial.)

The real issue: How do we use these tools (see: robots) to tell better stories and present information in a correct yet impactful way? How do we leverage machines to do better journalism?

I’ve learned not to get so caught up in the technical flashy stuff that I lose sight of this overarching goal. And that’s why I say programming is the easy part. What good does it do to code a flashy display of charts/figures/data/tables/insert buzzword here if it’s not doing the story justice?

And that’s why I’d like to talk a bit about silos and pods.

Not the grain container or storage solution, but the metaphorical structures they represent — silos as separate, rather isolated groups and pods containing a mix of these groups.

Silos: Why they don’t work

Consider the genesis of a story. Starts with a pitch. Maybe a crime beat reporter wants to do a particularly compelling story on the prevalence of teen violence, prompted by a few individual anecdotes. This has data potential, too. Anecdotes are great on a micro-level, but more broadly, is this part of some nationwide trend? The numbers can tell us the answer to that question. But data and thinking statistically, numerically, needs to happen at the beginning. If it becomes an afterthought, then it’s too late.

Moreover, let’s pretend that reporter didn’t come up with this story. Let’s say we had a script that scraped crime data and pumped it into a spreadsheet, and a data journalist drew up a quick chart to see if there was a story in the numbers. Maybe the stats tell us that the mean age of homicides in the last month has decreased by 50 percent, and we decide to run with that story.

A newsroom typically contains reporters, editors, data journalists, developers, designers … the list goes on. But traditionally they work in separate departments, unable to seamlessly collaborate together. That’s why silos don’t work. Instead, there should be a horizontal of these newsroom archetypes to truly integrate the journalism with the tech.

In my LedeHub work (as part of my AP-Google scholarship), I’ve been playing with the idea of version control for journalism. One question I’ve been struggling with is this: How do you support a process that allows cross-disciplinary teams to combine contributions and work in parallel?

Journalists can learn a lot from programmers. Namely, how to make machines do the grunt work and automate tasks. But there should be a feedback loop here, because programmers must also learn from reporters. Deep beat knowledge, how to find and talk to the right people to suss out a great story, these are critical to being able to do good journalism.

So we need to bridge the gap, because we’re all on the same J-team. It’s about information. It’s about the story.

Pods: Why they work

So instead of silos, I’m proposing newsroom pods.

Think of a pod as a horizontal slice of the varying newsroom archetypes. Ideally, a pod would consist of a reporter, data whiz, developer, designer and editor, resources permitting.

But there’s no need to be overly prescriptive here. What I’m really trying to get at is the integration across disciplines that this model affords. I’m not saying we should throw beats and desks out the window, but introduce some pods to make it easy for the story and the technology to meet.

The benefit of having a beat reporter’s intimate knowledge and expertise in a given topic area is invaluable. A developer helping journalists clean up messy government data sets is also pretty useful. There should be cross-pollination between the reporters and the news developers. I firmly believe that this is a pivotal step to producing great journalism.

So, how do we do it?

Newsroom geography is a good place to start. The news applications teams at the Chicago Tribune and ProPublica sit in the center of the newsroom. Plus-one for visibility. ProPublica’s news applications developers are highly embedded in the reporting process and work closely with editors and reporters throughout the entire story process.

The task of designing digital newsrooms is not a new concept. Various factors — audience, platform, brand, goals — play into building a newsroom and ecosystem that works for a specific publication, and I recognize that there is no “one size fits all” solution. But pods do exist.

The BBC’s redesigned newsroom layout also exemplifies that organization’s recognition of the need for integrated spaces. Integration helps ease information sharing and, in turn, collaboration. Steve Herrmann, editor for BBC News Online, emphasized that in the new floor plan, global and UK journalists all sit in the same space. I think the same must be done for news applications teams.

In today’s online news world, people have talked about unicorns. You could try your luck at breeding them, but in the meantime, we should work with what we have. People are awesome at what they do. Reporters cover beats like nobody’s business. Programming is often cited as an aggregate skill, but within that, developers have different specialties, too (at the most basic level, front end vs. back end). Throwing all these roles into a pod results in breadth and depth, and just might just help us do some better journalism.


This is one in a series of blog posts from the first ONA class of AP-Google Journalism and Technology Scholars describing their experiences, projects and sharing their knowledge with the ONA community.

Katie Zhu is a junior at Northwestern University, working on LedeHub, a project for collaborative, open journalism.