April 26, 2013

As the Associated Press’ new interactive editor, Troy Thibodeaux brings to the role the varied experience you’d expect of a former travel writer, English teacher and member of the NOLA.com and Times-Picayune team that won a Pulitzer for Hurricane Katrina coverage.

As AP Global Interactive Editor Paul Cheung explained in a memo to the staff, “Thibodeaux will lead a team of programmer-journalists to create groundbreaking journalism with a focus on newsroom tools, data-driven stories and interactive features. There will be a strong emphasis on working on global investigative stories, in alliance with news leaders and journalists across the company.”

We asked Thibodeaux a few questions about his new role, his plans for the future of AP multimedia and how rank-and-file journalists can get involved in interactive projects. Here’s a lightly edited version of our email exchange:

You’ve been with the Associated Press for a while now. What kind of projects led to your promotion?

Troy Thibodeaux

I came to AP to work on the 2008 elections as part of the multimedia team in D.C. Our team’s beat was potentially anything that came through Washington, so I was able to work on a wide array of data-rich stories and with a number of reporters who had deep domain knowledge in those beats.

After the election, I began working with our national and regional investigative teams, and I helped put together a training program in investigative reporting and data journalism for our reporters around the country. As a result, a large part of my job already involved expanding our ability to do data journalism at AP, and this new role builds directly on that experience.

The AP is expanding its digital content team. Who is coming to the team you’ll be overseeing? What kind of roles are you still looking for?

The data journalism team is largely made up of folks I’ve been working with in the Interactives department. They all have significant technical skills in data visualization, Web development or DevOps, but I’m really excited about the passion they’ve shown for investigation and storytelling — those things that differentiate journalist-coders from other kinds of developers. We’re also in the process of hiring a fifth data journalist for the team, and we’re looking for someone who shares those priorities and who can help expand our technical reach.

We’re a small group working with a large organization, so we have to be flexible – a team of utility players, rather than more narrowly focused specialists. Web developers often talk about learning the “full stack” — by which they mean learning both the back-end code (the data-centric heavy lifting the user doesn’t see) and the front-end code (the user interface we do see). Our team is learning the full stack, but our stack is bigger than most; it includes things like data vetting, story development and exploratory data analysis.

What are your plans for future projects? What do you and the AP want to do that you currently aren’t doing?

There are a few predictable projects on the horizon. Elections are probably even more important for us than for the average news organization. We’re also going to be working on other big events, such as the Olympics. There are some large investigative and enterprise projects underway and others we’re just starting.

So, we already have a lot on our plates. But we’re also going to discover important projects as a team, through our work with reporters and through our own brand of reporting: diving into the data. In general, I want to continue doing the kind of core work that AP does best, but I want to expand our technical abilities and bring more-sophisticated tools to that effort.

Approaches to data analysis that involve machine learning or natural-language processing, for example, can help us ask questions that have been beyond our reach. And by challenging ourselves to create more engaging data visualizations and intuitive user interfaces, we can offer our readers a deeper understanding of the stories they most care about.

How do you see increasing digital media distribution affecting how the AP reaches people? 

AP is fortunate in that our product has always been our journalism, and it can be distributed in many different forms. So, any development that increases the distribution range of our members and customers means we reach more people in new ways as well. We were among the earliest news organizations in the mobile space, and our mobile team is doing exciting things with our app. I’m especially looking forward to exploring new ways our data team can contribute on the mobile platform.

Are you planning projects that member outlets would be able to incorporate into their own websites? 

Absolutely! All of our interactives are embeddable, and our members and customers already incorporate them into their sites. I’m interested in ways we can customize our data-driven interactives, both for the customer and for the end user. Those efforts are part of our design discussion for the work we’re doing now. I would also like to find more ways to collaborate with other news organizations. AP was born from such cooperation, and it is an important part of our identity.

Are the projects the AP works on more often than not driven from the print side, augmenting what’s being written by a reporter or team of reporters? Or do they sometimes originate from the digital side? If it is a case of digitally driving what reporters write, would you like to see a shift to data journalists promoting more original narrative work?

More often than not, I’ve found that good data projects start with solid shoe-leather reporting, so most of our data projects have begun with the work of text reporters and editors. But the data work that we do begins before the story is written: exploratory data analysis can discover new angles in the data, and the right visualization can reveal a trend we had missed.

We’re really moving away from the model in which visualization or interactive storytelling is an afterthought, an illustration of the story, and toward a model in which this work is central to developing the story and enables us to tell the story in ways impossible for straight text reporting.

I’m also looking for our group to originate story ideas, to develop our own data sources just as reporters develop human sources, and to pitch ideas that could lead to text and video stories in addition to interactives and graphics. There are also some ideas that simply work best as standalone news apps, and in those cases we might not have a text story to accompany the app (at least not right away).

How would a journalist with no data or multimedia background best prepare themselves to step into that role. What lessons can you offer from your own career?

I’ve already written a couple of pieces for Poynter on this topic, so I hope what I say isn’t redundant. [See 5 tips for getting started in data journalism and 10 tools that can help data journalists do better work, be more efficient.]

In the typical progression, a reporter without a technical background discovers a spreadsheet containing valuable information for his or her beat. The reporter learns enough Excel to get some answers from the data by sorting and filtering and then writes the story. Sorting and filtering can take you pretty far, but eventually the reporter wants to look at change over time or some other interesting measure that isn’t available in the raw data.

So the reporter chats up someone who knows a little more Excel and discovers formulas. At some point, the reporter is forced to deal with a dataset that contains multiple tables, and in order to make sense of the relationships among these tables, the reporter opens up a relational database manager, such as Access or SQLite.

And then one day the reporter gets tired of clicking through pages of a website to copy and paste rows into a spreadsheet, so he or she decides to learn enough scripting to scrape the data automatically. Web scraping is a nice gateway task to learning a scripting language such as Ruby, Python or JavaScript, and the reporter is now on the road to becoming a complete news nerd. (Hurrah!)

The point here is that the learning is task-based. At each step, the reporter is doing something real, and there is a payoff in the journalism. I didn’t follow this path. I started learning to code as part of an early job doing technical writing. As a result, I didn’t really learn Excel until I began working with reporters who used it.

Still, very little of what I’ve learned has been in the abstract. I learned to use the tools I needed to do the work I wanted to do, and I’ve found that approach generally effective. Scratching your own itch is a pretty useful approach to creating software and to learning technical skills.

Aspiring journo-coders are fortunate: reporting skills provide a real advantage when getting started with coding. They already know how to ask questions and they know how to do research. (They also probably have superior Googling abilities.) But it’s also nice to have a supportive community to help provide direction, and many of us have found that community in the National Institute for Computer-Assisted Reporting (NICAR) and its email list, NICAR-L.

Sign up for the annual conference, even if you have to pay your own way. Register for the email list and follow the conversation. And when you can’t find the answer on StackOverflow.com, one of the brilliant and generous folks on the NICAR list will inevitably provide a clear and thorough explanation that will get you back on your way.

Support high-integrity, independent journalism that serves democracy. Make a gift to Poynter today. The Poynter Institute is a nonpartisan, nonprofit organization, and your gift helps us make good journalism better.
Joshua Gillin is a contributor to Poynter's MediaWire blog and a writer, editor and pop culture blogger for the Tampa Bay Times and its sister…
Joshua Gillin

More News

Back to News