What Matthew Liddy learned in his first year running an interactive storytelling unit
The Australian Broadcasting Corporation (ABC News) has been producing interactive stories for many years, but it wasn’t until February of 2014 that Australia’s public broadcasting company formalized the process and created a team to focus exclusively on interactive storytelling.
Matthew Liddy heads up the team of seven, which also includes two journalists, two developers, one UI/UX designer and an illustrator/animator.
Over the past year, they’ve developed a series of clever interactives designed to help Austrailians process the news in different ways. For example, when Australia hosted the G20, Matt’s team developed a quiz that asked Australians to figure out how they compared — in terms of life expectancy, literary, and other statistics — to the rest of the world. And when Austraila’s most famous pair of movie reviewers retired, Matt’s team analyzed all of their previous reviews to discover trends — a really, really good way to resurface archival content.
I wanted to check in with Matt a year after his team had been put together to find out what they had learned and could pass along to other newsrooms — particularly newsrooms that might not have an entire team of developers working to create new products.
MK: What's your daily workflow like?
ML: Our workflow varies quite a bit from project to project, depending on the complexity of the story we're trying to tell, how we're trying to tell it, how long we have to work on it and which specialty skills are required.
We generally start with a discussion about the subject we're looking to cover or the story we want to tell, and we ask ourselves questions about what audience needs we can meet, what we can offer that's unique and how it will provide value for users beyond what other parts of the ABC are likely to produce. From there, we get to work: doing journalism, designing, illustrating, coding.
Given we operate in a newsroom, we turn some of our work around very quickly. I find the workflow above can scale quite well, and it gives our team flexibility. Sometimes the initial discussion might be five minutes, and we'll immediately set to work and publish something in a matter of hours. Other times, we may hold a series of planning meetings and workshops with stakeholders and then step our way through a more detailed research, design and development process.
MK: What hardware do you use? What software?
ML: The more I think about this question, the more stunned I am at just how much software I use each and every day. It's quite surprising when you actually stop to think about it. Here's a snapshot of some of the key tools we use:
- Almost all our work is published via the ABC's content management system, CoreMedia. So I spend a lot of time writing, editing and uploading images in there.
- We use Hipchat for team communication, alongside email of course.
- Trello helps us organize our projects, though I don't think we're using it to its full potential at the moment.
- I use Evernote for capturing notes, archiving information I might want later, and general organization.
- We've used Shorthand for a couple of long-form multimedia stories, on the fallout of PTSD in the military, and the life of Australian media magnate Kerry Packer.
- We've used loads of different tools for various types of stories and interactives, including Mapbox for maps, Polldaddy for quizzes and surveys, Timeline.JS for timelines, Storify for social media curation and integration, and DocumentCloud for hosting documents.
- Our coders, of course, use all kinds of languages and libraries. Leaflet for maps and D3 for visualizations are a couple of favorites.
MK: What tools would you recommend for people in resource-strapped newsrooms? i.e. they might not have a developer but would like to create an interactive or quiz.
ML: There are so many tools out there today that it can be a bit overwhelming just trying to keep up with them. My advice would be to choose a few and really spend some time getting to know them and use them. Some worth checking out would be:
- Datawrapper: A really simple tool for building and embedding interactive charts. It's recently changed its pricing levels, but still offers basic usage for free. It has a really low barrier for entry for anyone who wants to dig into a bit of data journalism and basic visualization.
- Chartbuilder: The name says it all really: a simple tool for building attractive, useful charts. Quartz built this tool for their journalists to use, and then open-sourced it. They have a version live on GitHub that anyone can use, then you just download your chart as an image.
- You may have heard of TimelineJS already, but did you know the Knight Lab has also released tools that let you embed audio in-line with your text (SoundCiteJS), create those nifty before and after sliders everyone loves so much (JuxtaposeJS) and use maps to step users through a geographic story arc (StoryMapJS)?
- The Engaging News Project has a simple quiz creator; you can use it to create a quiz and then embed in your site.
- I mentioned Storify earlier for curating social media. SAM is also an emerging tool in the same domain, and is worth checking out. It can displays social media posts as galleries or maps.
- GIFs aren't strictly interactive, but they're unarguably powerful storytelling tools, whether you want to laugh at cats, find out what's banned when the G20 comes to your city, or learn how to use Github and the terminal. Gfycat is a useful tool for hosting GIFs; it'll deliver them as video to help minimize file size too, which is especially important on mobile.
- Twine: Have you ever wanted to make your own text adventure game (think Zork or Hitchhiker's Guide to the Galaxy)? Or write a choose-your-own-adventure narrative to help people understand a complicated story? Twine is an awesome free tool that makes it all possible - you can even add graphics and sounds if you want to.
- CartoDB is another great tool for mapping, alongside MapBox, which I mentioned earlier.
- Marquee, Storehouse, Shorthand, Creativist and Immersive offer tools for long-form visual and multimedia storytelling.
- Detective.io promises to help you conduct collaborative network analysis.
- Tinyletter lets you set up your own personal mailing list, simply and free.
There are hundreds of these tools out there, and more every day. (You can find plenty more at places like the Journalist's Toolbox, Poynter's Digital Tools Catalog and, hey, the Journalism Tools collection on Pinterest.)
You'll never learn to use them all, so don't even think about trying. Instead, start by thinking about a story you're working on, or one you've been dying to tell but struggling to find the right treatment. Then choose a tool that'll help and start there.
I should also add that not all the tools I mentioned are free, but have some free or relatively low-cost entry points.
If you're creative, there’s also no end of creative ways to use tools like Twitter, Tumblr, Pinterest and Facebook for telling stories in inventive ways. This choose-your-own-adventure Twitter story is but one example.
ML: I'm all-in on quizzes. I think they're an under-used tool for journalism. That might sound like an odd thing to say when everyone has read umpteen articles about what huge traffic drivers quizzes have been for BuzzFeed, but I don't think news organizations have really seen that the quiz can be so much more than just a bit of fun. Sure, the NYT had the dialect quiz, but still these types of stories are not widespread.
I try to steer clear of the BuzzFeed-style quiz, like "which city are you?" They obviously have their place but it seems like it's very rare that the data behind those quizzes really stacks up.
What I'm trying to do is learn from the success of that format but apply it to data or information that is actually sound, and in situations where people will actually learn something. (The Engaging News Project has indeed found some evidence that people learn more via online quizzes.)
In the case of the Tony Abbott quiz, we actually opted for that format because we knew we had a fairly difficult to understand dataset, and the quiz proved an effective way to guide people into the story.
The G20 quiz was very popular, heavily shared, and we had a huge amount of feedback suggesting people both learned something and enjoyed the experience. A popular TV comedy program here even picked up the format and content of our quiz, so we must have done something right.
MK: I love that you analyzed 1900+ movie reviews to create a feature about what the reviewers liked over time. I was wondering if you could talk a bit about the process of creating this, and how other newsrooms could do something similar. (It's such a good way to talk about archives.)
ML: Our audience loved that one, too. David and Margaret were basically our home-grown Australian version of Ebert and Siskel - the most respected movie reviewers in the land. So when they announced they were retiring, we decided to dive into the history of their reviews and see what the data could tell us.
All those reviews were sitting there on the ABC's website, and each one had been lovingly metatagged with not just the scores David and Margaret gave them, but all kinds of useful information such as the director, the stars, length, and classification.
One of my fantastic developers (@collypops) set to work writing a web scraper - a program that found every review on the site and pulled down all the key information into a database. He then produced some charts using the D3 library, we added some analysis of the data and a few fun snippets from relevant reviews, and voila, we were done.
It was such a good idea that our friend at Guardian Australia, Nick Evershed, came up with the same thing, independently of us. What I like about that is it shows that it's not only your own archives that you can mine - you can often use other organizations' archives, too.
MK: What kinds of projects are you hoping to work on this year?
ML: I want to work on projects that tell people something they don't already know, that evoke an emotional response and that are fun to work on. That won't describe every project we do, but I feel like those are pretty good markers for us at the moment.
MK: Do you open source any of your projects or publish your learnings online?
ML: We've taken some baby steps in this direction, and I'm keen to make more progress in 2015.
Another developer on my team (@drzax) recently led the way in starting to share some ABC News code on Github. We also released the data we got from Vote Compass, a political engagement tool we ran as a co-production with a group of academics and which became the biggest survey of political opinion in Australia.
These are humble beginnings, but it's good to have made a start.
MK: How do you resurface pieces from your archives that might be relevant for readers?
ML: We work very closely with the News desk, and regularly provide ready-made teasers so they can link out to our content from within daily news stories. We also share our stories again on social media when the issues we've covered come up again in the news. Both of these approaches have proved really fruitful in driving audiences to relatively timeless content such as our interactive explainer on Ebola.
MK: When do you decide to create an interactive? Is there a pitching process for the staff? When is it better to just write a piece vs. building out something complex?
ML: You've basically just described a big part of my job. I take pitches from journalists and content makers across the ABC, and in consultation with my team decide what we can tackle and how to approach it. At this stage, we've kept this as an informal process (phone, email, in-person conversations etc.) rather than asking people to fill out commissioning or pitching documents.
In terms of when to build something complex, that's a great question. There are certainly very many situations when it's better to just write a piece, as you say. This is particularly true if you're talking about breaking news. In that scenario, we would always take the approach of covering the story first and then looking at how we can quickly provide background and context through a map or an interactive.
I don't think there's a magic formula for deciding when to do something more complex and interactive. If somebody has one though, I'm all ears. In addition to the usual questions about how important a story is and how strongly it will resonate with the audience, I tend to ask questions like these:
- What's the best way to tell this story? What tools and techniques do we have at our disposal?
- How much time do we have to create something?
- How timeless will the story be? Will it still be relevant?
- Can we add background and context that will help people to understand what they're reading about?
- Is there a way we can tell the story online that you can't in any other format?
- Are there elements to the story, or source material we have available to us, that demands an interactive type of presentation?
- Will making an interactive or adding other elements to the story help people understand it better, or help them engage with it in a deeper way?
- Are there challenges to telling the story in a traditional way that we can help overcome?
- Will an innovative approach to telling the story help attract an audience to an important story, or help guide them through it?
- Is there a way we can put the user into the center of this story?
MK: If you learn something, how do you share it with your coworkers?
ML: With my immediate colleagues, I'd turn to Hipchat. With wider groups, email. But I don't think this is an area where we excel, and it's one I'd like to improve. There's always a tension between spending time on producing audience-facing work or spending time sharing things you've learned with colleagues. We tend to concentrate on the former at the expense of the latter, but I think we could aim to find a better balance and also develop some different channels for that communication.
MK: Your team is relatively new. How do you manage the team's resources and time?
ML: The demands of working in a news environment mean we have to take a fairly flexible approach, and sometimes pick up and drop projects at short notice.
We don't follow any strict protocols or software development practices, such as waterfall process, agile or scrum methodology - though I remain open to experimenting with them.
We have a daily stand-up where we focus on discussing what we've been working on, what we're planning to work on and any questions or blockers we may have. I have a great deal of confidence and trust in my team, and tend to give them the freedom to manage their own time.
It's probably not a perfect approach but we've been pretty darn productive in our first year, and managed to hit deadlines while we're at it.