We're now at the point in the lifecycle of experimentation when it's time to figure out how to make our most high quality work repeatable. The first time we build a thing, it's usually a one-off as we determine what works and how we measure the thing's success. But as you start to figure out what works best, the only way to continue evolving that experimentation is to ensure repeatability and reusability of your baseline.
That's where we stand now with art-directed longform storytelling, and not just at Vox Media. We see news organizations innovating in this space more than we ever have before. But this work takes time, especially if you're going to make it super-special. This week at OpenNews' first-ever SRCCON conference, we tried to figure out how to make it scale. We don't have all the answers, but these ideas from the audience at one SRCCON session we organized are a great start.
Why SRCCON is awesome
First, though, a note and some background on SRCCON's sessions format. SRCCON is a conference for developers, interactive designers, and other people who love to code in and near newsrooms (hello!). Unlike other media and design conferences where people stand in front of a room and lecture at you about all the things they know, SRCCON organizers realized that there are many smart people in the room with lessons to share, so its presenters served more as moderators to a group discussion. The format took the best part of conferences — mingling with people in the halls and bars to knowledge swap — and made it the core focus of the event. Rather than going home later and synthesizing the learns post-conference, we were able to hash it out in realtime with real feedback and exploration amongst esteemed peers, meaning we could leave with more tangible takeaways.
These were some of the takeaways from one session we lead.
The underlying problems
Our group conversation explored the question of how to scale longform storytelling, and we asked the participants to identify common problems on reaching this goal. Some of the problems and questions the group identified:
- Developers are too expensive to use on one-off stories.
- It's very resource intensive, aka it takes a lot of time, aka it's expensive.
- Resisting the urge to art direct everything
- How do you get people to engage beyond just oogling at the flashy design?
- How do you measure success?
- "Business model lol"
- How do you balance between making things automated but also special?
- Responsive is hard.
Overall, the problems fell into four major themes, for which we broke into groups to try to determine best approaches toward fixing some of these problems.
Don't start off by thinking that the thing you're going to build is going to be reusable, because there's honestly no guarantee that it will be. The first time you build a thing, you still don't know what you're doing, and shouldn't pretend that you're going to solve for reusability in its first state, since it's bound to change based on your initial learns. Instead, build with good, clean, extensible code that is well-commented and documented. It'll be easier to turn it into something reusable from there if you decide that's what you want to do.
Abstract out your templates into smaller, more reusable chunks. Rather than building out full templates that define every piece of a story, then trying to reuse that, break out into more modular templates that can be mixed and matched to create combinations that are ultimately different but share the same building blocks.
Actual tools you can use:
- The Chicago Tribune's Tarbell
- Templating frameworks like Bootstrap that you can use as a base and bake your own look and feel over it.
- Sass mixins that people can understand to abstract out common design themes that are being reused.
Give people something valuable enough that it's worth their money to subscribe, if you're a publication that might have a subscription solution as part of your business model.
Use longform storytelling to engage people with your site and draw them into other content. Because the content is so high-touch and bound to get attention, use that as an opportunity to draw readers into other places on the site and hook them.
Make it reusable. The only way to truly be able to justify this high-touch work from a business standpoint is to be able to do it in a way that scales for 90 percent of your work, and requires custom enhancements for only 10 percent of the rest of it. You'll notice this is a common solution amongst every theme we identified.
Build things in pieces, not as one big thing. This idea gets back to the same solution identified by the reusability group, which was that you can hit that sweet spot of making each thing special, but also basing it off the same templates by breaking out into smaller templates that you can mix and match.
Have well-formatted data models that allows your content to flow into these designs. While we didn't get into specifics of how that data model would be structured (at least in this session; there were other sessions that did), we determined that you need to have more than just a headline, byline and body field to make this work.
- Marquee, a suite of editorial tools that make it easy for editors to weave images, audio, video, and media from across the web into engaging narratives.
- Substance Composer is a tool that supports media content such as images and videos. You can even include interactive web content such as visualizations.
- Chorus is our own platform that we've been building to solve these problems. Recently the founding team of Editorially joined Vox Product to help work on these ideas and we hope to share what we learn along the way.
The human aspect is always the most important part of a process. Before you can make any progress toward getting developers, designers, photographers, videographers, data journalists and writers to collaborate in this space effectively, they need to understand the work that each other does and learn how to best communicate and plan around these kinds of projects.
Have core components that are standardized so that the process for creating these types of stories can flow more quickly and efficiently. Again this was a common theme amongst every different group in the session, and getting his right seems like it'll be key for scaling longform storytelling.
Have a formal way of institutionalizing your experiments. Your process can get into a real rut when it comes to longform if you keep doing one-offs over and over again. Part of the process that you have to fix is the part where you allow this cycle to continue. Instead, you should make a rule that by the third time you've redone a project in a similar way, it's completely reusable. The first time, it's probably highly custom, the second time it should be a reuse of the first time, and the third time it should basically be a series of text boxes and check boxes (probably more complex than that, but you get the spirit).
Share, don't just show.
And, overall, one of the best takeaways from SRCCON that OpenNews Director Dan Sinker made in his closing remarks is the idea that we, as an industry of media and design technologists, need to get beyond simply showing our work. This idea helped us get to where we are now, but it's no longer enough. It's now time for us to also share our work, which means documenting it, explaining it, talking about your learnings and failures, or involving new and different types of people and groups in the building of your products. In that spirit, we'll be sure to share with you as we make progress in this arena. One of the first examples is a post that one of our developers, Guillermo Esteves, published today about how we made The Verge 50 feature with our own Editorial Apps approach. We also shared the source code for the feature itself, but you can also find the source for our Google Drive Middleman extension.