Old Media Interview: Aron Pilhofer, interactive guru, editor at The New York Times

December 1st, 2008

For our last Old Media Interview, we spoke with Stephanie Romanski of The Grand Island Independent in Nebraska.

We’re now going from the smallest of the small — perhaps Grand Island isn’t the smallest place ever — to the biggest of the big.

You know all of the data that goes into, well, everything you see from The New York Times? Aron Pilhofer is the guy responsible. We asked him a few questions.

Please give us a short bio/official title:

I am the editor of Interactive Newsroom Technologies at The Times, which means I run a small team of developer/journalists here in the newsroom. I’ve been involved with managing the group since we started it a little over a year ago, but it’s been my official day job since January. Prior to that I was a reporter on the computer-assisted reporting team at The Times. Before that I was database editor at the Center for Public Integrity; a statehouse, municipal and projects reporter for Gannett and on the training staff of Investigative Reporters and Editors.

What are your unofficial duties?
We’re a small team, and relatively self-sufficient in terms of the technologies we use (Ruby on Rails, Flash, Amazon EC2 and, soon, a spot of Django). So we all have to do whatever needs to get done, in order to get things done. Sometimes that’s glamorous tasks like data entry, or managing servers. Sometimes that’s waking up at 3 a.m. to fix a misbehaving application (which happened often during the Olympics). And on good days, I get to code! My official duties include a lot of meetings and managerial nonsense I tolerate because the rest of my job is so darn much fun.

The Times is well known for its visualizations. What are some key ingredients for a powerful interactive graphic? What kinds of responses have you received from readers?
Last bit first: The responses from readers run the gamut as you can imagine. One thing I will say about New York Times readers is they really, really pay attention to what we have out there, and when you have something wrong, they let you know about it. Fast. In no uncertain terms. Of course, they’re equally responsive when you do something right.

After Super Tuesday, we got gobs and gobs of email from readers praising our coverage, which was extraordinarily gratifying since it was for us maybe the scariest project we’ve worked on, including election night. That was the first big test of my group, whether we could pull off a major, clutch project, so it was pretty scary.

In terms of what makes a powerful interactive, I think it depends on the story, obviously, but the best ones I believe are the ones that are truly … interactive. That term is widely used to describe a lot of web features whose level of interactivity is limited to the user clicking on a button and watching things move around the screen. That’s absolutely a form of online story telling, but it’s not interactive in any way, shape or form.

So, the ones that work are the ones that call to action (like Word Train or the one we launched Tuesday, If You Were President), or the ones that provide readers a way to probe and investigate on their own. One of the first (and still very best) examples of this is our Casualties of War interactive. Using a very simple slider interface, users can see trends that otherwise would be hidden or obscured in a sea of data. It’s still one of the best examples of the form. Our election night maps, of course, are also tremendous. Both of those projects were collaborations with the graphics desk, which I can say is easily the best in the business.

Actually, come to think of it, collaboration is maybe THE key ingredient to what makes a powerful interactive. One of the great things about our newsroom is that it is unusually collaborative, and most of the projects you’ll see involve multiple desks: us, the web newsroom, computer-assisted reporting, our multimedia group and our graphics team. We all bring something different to the table, and that’s how good ideas become great interactives.

Which three NYT interactive graphics are your favorites?

I’m in an election mood, so:

- The house maps from election night [are] something close to magic.

- I also thought the pop-up dashboard was a huge hit. (Click the image link just to the right of the photo)

- And I liked [the] word train, which was very un-NYTimes-ish.

How much do things like that cost generally, and what’s the ROI?

Gotta punt on this: I have no idea on either score. We are a news desk, so what we do is motivated by journalistic goals. That isn’t to say that our stuff doesn’t draw significant numbers, and doesn’t “sell.” It absolutely does. It’s just that I don’t have a whole lot to do with that beyond letting our product people know what’s coming up. And that’s the way it should be.

Are there any free or open-source products out there that small-town newspapers can use?

Everything we use is free and open-source. Our platform is Ruby on Rails backed by Mysql databases running on Ubuntu servers. The cost here isn’t software, or even hardware, which is relatively cheap these days through hosting companies like Amazon EC2 (on the high end) or Slicehost (on the low end). The price most news organizations (and it’s not just small ones) seem reluctant to pay is for people — developers like the ones in my group who can build the infrastructure to support the rich, deeply engaging web features that so many people love about our site.

It seems that some pretty major news organizations have built on-site social networks: There’s TimesPeople, the Journal Community and the NPR Community. What are your thoughts on news organizations in this space? Are on-site social networks the future for news?

Well, the person to pose this question to is Derek Gottfrid, the absolutely spooky smart developer/renaissance man behind TimesPeople and TimesMachine. He will give you an earful on the subject.

To my way of seeing it, the answer is yes. But as a means, not an end. We’re not building FaceBook on the Times, and that’s not something people really want anyway. What it does show is how the Times‘s view of the web is changing, and how we’re starting to approaching the web on its terms, rather than trying to bend it to ours. So, viewed that way, it shouldn’t be surprising to see us embracing the web more fully than in the past: data, documents, multimedia, community, RSS feeds, blogs, APIs, etc.

TimesPeople itself is not the future of news, but as part of a larger transformation from “newspaper” website to just plain website, it is.

We’ve seen more and more New York Times reporters get on Twitter. What does it take to get them in the space? Have these reporters been able to develop a rapport with readers? Any success stories yet?

Honestly? I have no idea. I follow/am followed by a couple Times folks, but I really don’t know whether they are using it for work or not. I guess I have to come clean and admit that I only very recently discovered the appeal of Twitter, which I used to think was kind of silly. But I don’t know that I could go a day without it now. It’s like writing headlines to your friends in real time, so, what’s not to love for a newspaper guy like me? I’m probably the best success story I have for you there. Sorry.

What three things can newspapers do to be more like the Times — in a new-media sense, of course.

Wow, big questions. I’ll give you three high-falutin’ answers:

1. Invest in your website. This is the time to bring in as many talented coders as you possibly can and now is the time to let them experiment with new approaches and new models. Sadly, many news organizations seem to be doing just the opposite. You can’t outsource innovation or buy it off the shelf, but that’s what a lot of news organizations are trying to do. It’s puzzling really.

2. Change how you think of news. Jeff Jarvis wrote an amazingly on-point blog post recently in which he argued that the atomic unit of journalism is no longer the news story, and he’s absolutely right. I’m not ready to go as far as, say, Adrian Holovaty and call every data point… news by definition, but I do think that journalism will increasingly involve more than writing and reporting. It will be about the aggregation and curation of information, and news sites will be more and more centers of authoritative information. Authority is one of the rarest commodities on the web, and we have it is spades. We as an industry need understand that, and follow that idea to its logical conclusions.

3. Get the journalists thinking more like developers, and developers more like journalists. Ask any print journalist you know how a newspaper gets produced, and chances are they’ll be able to tell you in great detail how copy flows from the reporter’s fingers to your front stoop in the morning. Now, ask them how that same copy gets onto their website, and chances are quite good they won’t have the first clue. Is it any wonder so many news websites look, feel and read like print on the web?

If the web is our future, we either need to master the medium ourselves, or embrace those who have, or both. By no means am I suggesting The Times is anywhere near close to that point, but we have taken several important steps in that direction. Sadly, those baby steps place us far, far ahead of the pack. There is time to change, but, again, few media organizations seem willing or able.

What’s next for you guys? Where do you see the biggest growth potential for The Times?

I can’t speak for the entire organization — there’s just way way too much going on. But for my little group, 2009 is going to be about community and about enabling the journalist to be more of an aggregator and curator as I mentioned above. In part, that was the motivation behind the DocumentCloud, which is a joint grant proposal I wrote with Scott Klein and Eric Umansky of ProPublica for the Knight News Challenge. That project, if we’re lucky enough to get funding, will enable everyone to treat documents like structured data. It’s a project I’m pretty excited about.

The other big focus for us this year is on community. We have a couple of projects (which I can’t really talk about) that will build on TimesPeople in some new and interesting ways.

Thanks, Aron.

Aron has also agreed to answer a couple of YOUR questions. If you’re one of the first few to post your questions as comments on this story, he’ll (most likely) respond.

Entry Filed under: Old Media Interview