Twitter  A look at the state of local public media via Localore  
Nieman Journalism Lab
Pushing to the future of journalism — A project of the Nieman Foundation at Harvard

Q&A: The Guardian’s Gabriel Dance on new tools for story and cultivating interactive journalism

“I really look forward to the day where the types of people we have in our newsrooms, and the roles they play, aren’t defined by the medium in which they present their work.”

gabrielddanceGabriel Dance thinks a lot about tools. Specifically, what tools does he have at his disposal, and what tools can he and others on the Guardian U.S. interactive team build to help them better tell a story? Dance, the interactive editor for The Guardian U.S., spent a good chunk of the last several months working on NSA Files: Decoded, a multimedia examination of all the information revealed so far about how the U.S. government conducts surveillance on people in America and abroad.

In trying to provide context around the story, Dance and his team used a blend of data visualizations, videos, social media integration, documents, and animated GIFs. Dance doesn’t see it as a collection of bells and whistles, but as a way to take advantage of the tools the web provides to help make stories more engaging. “I’m not above the idea of saying the Internet is a competitive place — there’s a lot of cats and babies on the Internet,” Dance said. “It’s our challenge to engage our readers in a way that captivates them. And the idea I can captivate them while telling them this incredible story, there’s no reason I shouldn’t be able to captivate them with such an insanely interesting story.”

Dance has been building these type of stories for a while, having worked at places like The New York Times and The Daily. Recently we spoke about how the Guardian U.S. built NSA Files: Decoded, as well as the state of online storytelling and how to create a culture of interactive journalism inside a news organization. And yes, we talked about that Times story with the avalanche. Here’s an edited transcript of our conversation.

Justin Ellis: Give me a sense of how you guys went about building the NSA Files: Decoded and how you got involved with the story.
Gabriel Dance: As soon as the story broke on June 5, with the Verizon story, and then the 7th, 8th, and 9th followed up with some pretty major revelations. We also knew they were extremely complicated. Some of these things were technically complicated — computers and hacking and Internet taps. Some of these things were legally complicated, like the FISA courts and the Patriot Act. So immediately we knew that not only was there an opportunity, but there was a need.

I have a good idea of how it works in some other newsrooms, but there’s not any sort of editorial or reporting divide between my team and the rest of the newsroom. I know at The New York Times it works very similarly, same at The Washington Post. The people on my team, I call them “interactive journalists.” But I’m looking forward to the day we can drop some of the prefixes and leave it as “reporter” or “journalist.”

But the simple answer is, I can go to someone on my team, give them a topic, and ask them to report on that topic and the process is very similar to how that might work with a more traditional print reporter. They go out, they explore it, they see what information is available. They schedule some interviews, they gather data, and they come back and we work on what form it might take.

Are we producing articles with strictly text? No. But I think what we are producing is much closer to what you saw in Decoded, which is what I guess I’d call web-first journalism.

When people talk about this as new and different, it only really is if you’re looking at the old things as the status quo. The way I see it, the web has evolved and changed the way we have the ability to tell stories. What we’re doing is simply adapting to all these new ways we have to tell stories.

A comparison I like to draw a lot is photographers. Nobody ever questions whether a photojournalist is a journalist. I don’t see anyone asking if Tyler Hicks is a journalist. This bizarre notion that because we’re doing this — what is, because of technology and the Internet, a new form of storytelling — there’s this natural, human trait to be defensive of what it was before.

So I really look forward to the day where the types of people we have in our newsrooms, and the roles they play, aren’t defined by the medium in which they present their work.

Ellis: For the NSA Files, what’s interesting to me is the combination of text, interactive, and video. You’ve got this whole cast of people talking directly to readers. How did you decide on that?
Dance: That fundamentally goes back to the goal of this project as a whole. I know we’ll get to the point where we’ll discuss this and Snow Fall and things like that, but NSA Files Decoded is explanatory reporting. The goal of NSA Files Decoded is to engage and interact with our readers while making a very complicated series of stories accessible and relatable.

So when you’re doing explanatory reporting, obviously the goal is to explain a rather complicated or in-depth process to our readers in a way they can understand. So what better way, when we were conceiving of this, then to have them be able to hear the words exactly from the experts? More to that point, I had this vision of this stark page where the interviews are being done, but give credit to Bob Sacha, the freelance videographer who we hired to work with us. Bob and I were talking, I was explaining to Bob the concept and he said “why don’t we sit right behind the camera and have them speak directly into the camera?” As soon as he said that, it was perfect.

We want them to talk directly to our readers, because we wanted this to be like an intimate conversation between our readers and the experts. I have themes I’ve had all throughout my work, and one of those themes is transparency. So this idea that you’re not just getting a quote in a story, you’re not getting part of a quote: You’re getting the actual expert answering a question relevant to the thing.

It kind of removes us from the process in a way where we’re still doing our job of facilitating journalism, but we allow this really interesting, intimate connection between the reader and the content. And that’s actually one place in Decoded we really have tried — and why I call my team an interactive team and not a graphics team or visualization team — to engage and involve the reader in the storytelling in a way that makes them feel as if they, themselves, are part of the story.

Ellis: With shooting all the videos and building the project, how much time went into making this interactive?
Dance: There are a lot of subtleties that go into it. One of which is the idea that content gathering and production went on at the same time. But the interviews were somewhat difficult to schedule and took a while.

I would say there were three full-time people, including me, on this for two full months. Then, in the middle of the project we had a little bit of turnover in my team. I hired two people, Kenan Davis and Ken Powell, and they jumped in. It’s difficult to estimate how much time they spent, but maybe we’ll say they spent three weeks to a month each on it. It was the full team, for a month, and three of us for another month. It was a lot.

It’s a fair question. But it’s such a difficult question to answer. Some people write me and say, “How many hours does it take to build this?” And it’s like, I don’t know. If we sat down with the exact idea that we wanted and all the functionality, everything, in line…who knows? Potentially you could build that in a couple weeks. But these things are a constant evolution. It probably took over a month before we even had any kind of working prototype with actual content in it. Until you put actual content in it, it’s really hard for other people to understand what you’re getting at.

It’s the subtleties, because the time also includes a lot of discussion, changing course, seeing what works, seeing what doesn’t work.

Ellis: What are some of the challenges you had, specifically on the layout or bringing everything together? There are a lot of moving parts there.
Dance: One of the most difficult things was the initial planning for it. I sat down with a couple other people and made an outline of the entire project. This was an outline of the story, not an outline of what copy we were going to write or what video we were going to do or what interactives.

Then I went through the outline of the story and I started to say, “Okay, what’s the best way to tell this bit of this story?” And the “best way” is flexible term — that could mean most engaging, that could mean most illuminating. There’s a lot of different things “best” could be.

There was this whole planning process because we wanted everything to work in concert. The challenge is none of those things can be done outside the concept of the entire story. That’s why, fundamentally, starting from the story and then deciding which parts are best for which mediums was both very challenging and, I think, in the end extremely rewarding and part of the reason why I think Decoded did so well.

But some of the other challenges — all the interview clips you see are one-shot takes. We did do two-camera interviews, but we never once used the second camera. And we never used any jump cuts. Every single quote is a full-on quote. That’s why some of the quotes, they stumble a little, or they say “um.” But it would have broken with our paradigm of speaking directly to the reader to have cuts in it.

For every single bit of it, I could go through and say the challenges of all these things working in concert to tell a larger narrative. And I think that’s actually what distinguishes Decoded. What I told somebody the other day was: This is a web-only project, wherein we’re using all the Internet mediums available. GIFs, video, interactives, maps — all of that, hopefully seamlessly throughout.

That’s where I think the difference between it and any other projects lie. This piece was designed to be read, or consumed, as a whole. You can’t take the writing out of it and have it work the same. You can’t take the videos out of it and have it work. You can’t take out the graphics and have it work. It’s meant to be consumed as an entire project, with all these different parts being seen, seamlessly, to one another. It’s actually a giant fucking metaphor for what we were talking about before with journalism. It’s all journalism. It’s just being told in different methods.

Ellis: How do you think about keeping people’s attention in a story like this? What have you been able to learn about the ways people engage with interactives like Decoded?
Dance: I think with Decoded we’re definitely trying to evolve the method of storytelling on the web. It’s a total package. Some of the other things we do are different ways to tell individual stories.

I’ll just speak to Decoded, because I do see it as a culmination of a lot of what we’ve been working to. But, I’m just gonna address some of this Snow Fall stuff, OK?

Ellis: Please do.
Dance: It just has to get out the way. I thought [former Times Design Director Khoi Vinh's] post was poorly argued. Let me rip off a few ways this is different from Snow Fall.

Snow Fall is over 15,000 words; ours is barely over 4,000. Let me be really clear: I think Snow Fall was an amazing advancement of multimedia storytelling. On the other hand, it didn’t have any interactive graphics in it. Their interactive team is fucking outstanding — so it’s only because they haven’t decided to do this yet. But it still remains a huge difference. Essentially what they had were words with videos, and these videos had amazing renderings, obviously, of avalanches. They’re infographic videos, I guess you could say.

Like I said, we’re an interactive team, so we’re trying to engage our audience. So that absolutely goes into our thinking of how we did this. We knew the words had to be concise, brief, and clear. I’d say the fundamental difference between Decoded and Snow Fall, is that we’re doing explanatory reporting and they were doing feature reporting.

That doesn’t mean one is better or worse, at all. Both are critically important, but they’re extremely different types of reporting. You wouldn’t compare somebody’s article about the health care rollout and the website failures with a feature story about the pitfalls of maybe one family’s plight in a state without Medicaid.


But I do recognize captivating people means something different with the new technologies of the web. That means allowing people to engage with it; that means allowing them to connect it to their social network. It means making it seamless, and having videos automatically play. Some people hated that. But imagine how annoying our piece would have been if you had to click play on each one of those things?

So this idea of having videos automatically play when they’re centered in your window, and all those things, those are all to engage the reader, to keep their attention. To bring them further into the story. It seems like a lot of people feel like you’re selling out when you say, “I want to entertain and engage and inform and interact with our readers.” That’s our goal.

I haven’t checked the stats, but just over the first two days there was over 10,000 people spending more than 30 minutes on the site. That kind of blows my mind. Those aren’t people I know, even. My mom loves me and I don’t think she’s spending 30 minutes going through this.

Honestly, it’s because of these interactives and these different mediums. If this was all just copy, they’d probably do what I do, which is either print it off or Read It Later or something. Or if the whole thing was a video, I have no idea what it would have been.

None of these things are easy. People want a template for them. There is no template yet.

Ellis: You guys are still a relatively new entrant here in the U.S., and you don’t have the print legacy that the Times or the Post or the Journal does. How can these type or projects create more awareness of what the Guardian does?
Dance: One of the main differences between what we’re able to do, and, say, Snow Fall, or Sharks and Minnows, is they do have to design a project with the idea that this is running in a newspaper. By definition, they’re creating something for two mediums. We don’t have to do that at all. Which is really a fantastic freedom and independence to have.

Now, obviously, the entire Guardian doesn’t have that. There’s a very healthy newspaper in London. But in the United States, we’re a web-only publication. I don’t have to worry about speaking with a copy editor about what the print product’s gonna look like, because there won’t be a print product.

We have this enormous hunger for information about the United States from international people. So we have this really cool opportunity to explain stories from the United States in a way that is accessible to internationals and to American citizens themselves.

In our interactives, we try to express that as creatively as we can. Whether that’s with a gay rights visualization that looks like a huge multi-colored rainbow, or a choose your election game where people are holding balloons. Or it’s a spin the debate, where you can reorder people’s words. We’re taking these really different, creative approaches because we’re quite aware of the media landscape and where other people fit into it.

So we’re trying to create this niche, that I think fits with the Guardian brand. It’s news, but you know you’re going to get a breath-of-fresh-air kind of thing. It’s also creative. Oftentimes, there’s entertainment and it’s fun. But it’s also, fundamentally, extremely well reported, extremely thorough, and accurate storytelling.


Ellis: One question that comes up with interactives is how to make them a part of everyday storytelling. Is it having people who see things both ways, interactive journalists like those on your team?
Dance: Working with more traditional reporters is fundamental to creating interactive projects. That’s true here at The Guardian, that’s true when I was at the Times — I think it’s true, period.

But the answer to your question is tools. I created a lot of tools when I was at The New York Times, some of which they still use, some of which they don’t. Tools can be extremely helpful; you see them at the Times, the Guardian.

I think Quartz does a fabulous job of doing same-day news-relevant charts. Quartz has obviously made the decision that they’re not going to be creating, for the most part, interactive charts and elements and things like that. We can turn around quick charts and quick data, and quick contextual elements. But we can’t make them into these really whiz-bang interesting interactive social hook-ins. That just takes time.

The second part of that is something Amanda Cox speaks a lot about, which is the idea that you don’t have templates for stories. John Branch writes 16,000 words on Snow Fall, or Ewen MacAskill writes 4,000 words on the NSA — that’s not a template. I don’t see people saying “where’s the template for longform investigative?” You know what I mean?

This idea there’s going to be a template that does interactive longform web journalism? I don’t know if that’s ever going to be the case. Now there’s certainly best practices that are being developed, and things you see people like The Verge, and ESPN, and The New York Times, and The Guardian, which are templates that really highlight visual journalism, that have huge images and huge videos.

Again, I would just say where’s the templates for any longform writing? It doesn’t exist because each story needs to be told on its own terms.

What to read next
Jake Batsell    April 15, 2014
The daylong summit on new models for supporting journalism examines how the Texas Tribune diversified its funding, the injection of venture capital and private wealth into media, and the future of philanthropy for news.
  • Andrew Duffy

    I think this issue of making interactives part of the everyday is one of the most important questions surrounding multimedia journalism at the moment. +1 for developing/sourcing the right tools, that definitely goes a long way (glad you mentioned Quartz there as well)… but I have a slightly different opinion re: using ‘templates’.

    Gabriel — since there’s no template for interactive longform can we expect the next piece you put together like this will look vastly different to the NSA Files? My guess is no. My guess is that you’ll rework/remix this structure if it turns out to be successful… and if you don’t I reckon you’ll be wasting resources. Reworking/remixing something that works is working off a template IMO, but maybe I’m splitting hairs/quibbling over definitions on that point.

    Also… Would The Guardian ever think about opening up all of the source code behind the NSA Files? Since it’s on the web I know we can pretty much pull it down ourselves and re-engineer it… but it would be great to have the whole thing put up on github the same way as it sits on your server. I see a lot of talented teams around the world constantly re-inventing the wheel because of the closed-source nature of a lot of these projects. I think that’s been somewhat true of a few things in the NSA Files as well. We’d all benefit and we’d all move the format forward much faster if there was more knowledge sharing going on.

    Thanks to Gabriel for taking the time to answer these questions and thanks to Nieman Lab for putting the story together. More please!

  • NewsResources

    Hi @Andrew,

    I agree with you : “this issue of making interactives part of the everyday is one of the most important questions surrounding multimedia journalism at the moment.”

    For those who searched for tools into this post
    At the moment it seems there isn’t a perfect tool, but there are some pretty nice efforts like

    sstory from @mrejfox –

    or scrollkit –

  • Andrew Duffy

    Wow, sstory looks awesome… thanks so much for sharing. I have had my eye on scrollkit for a while, as with some other tools, like creatavist (I’m starting to see a lot of stories get built with that). Sstory looks more up my alley though.