People don’t talk much about categorization anymore. It used to be that if you didn’t have index cards, the only way to figure out what was in a book was to read it from cover to cover. That’s why The New York Times back in the 1800s decided to categorize every single article that was written – something they still do today, though now it’s at least partially automated. We don’t talk about categorization anymore, because with search engines as good as they are, the problem of finding things in text has mostly disappeared.
So why would we, in the 21st century, bother with putting stories in buckets?
Because we want to give our readers the ability to make connections between what they’ve read and what they haven’t read. At Fusion, we’re laying the foundation by allowing editorial to assign any piece of content to one or more narratives. Our narratives include trending topics like the Sony Pictures Hack, events like Oscars 2015, and people like Hillary Clinton. They encompass Section, Topic, Person, Location, Organization, Event, and Story Type. We like humans to create and assign these terms, instead of algorithms, because people are still much better than computers at figuring out what’s important or interesting about a story.
We try to be somewhat methodical about what we turn into a narrative. Proper nouns and broad topics make for good narrative terms. The color “yellow” or “posted on a Tuesday,” probably not. But unlike libraries or The New York Times’ taxonomists, we don’t worry too much about categorizing for posterity. And unlike Tim Berners-Lee, we care less about how robots can navigate what we publish and more about showing our readers and viewers stuff we think they might like. An article about CRISPR could fall into “Technology” (broad), as well as “technology ethics” (narrower) and “CRISPR” (narrowest). These are all very different, but they’re all things that we’ve written about and we want to give readers the option to dig deeper.
One way to think of narratives for a story is to reverse engineer: if you were looking for the post, what words would you search in Google?
Step 0 is adding structure to your content through the use of narratives. The neat thing about structure is that it acts as a foundation, something you can use to build entirely new experiences that you hadn’t initially considered.
Narratives serve a dual purpose of being useful to our audience, but also being useful internally. As one example, we can use them in our analytics tools to better determine what kinds of stories are resonating. It’s often impossible to draw conclusions from a single article, but when you bundle them together into narratives, patterns can emerge. Some kinds of stories work better on Facebook, others on Twitter, others when people are at work, and yet others when they’re relaxing over the weekend.
We also have tests running on the site to play with the concept of “following” articles. In its simplest form, a user could hit a “follow” button to receive email updates whenever there are new stories published on the Topic, Person, Event, etc. However, what you call the button would frame what the experience could be. For instance, the updates don’t necessarily need to be chronological. If I’m new to a narrative, then maybe the most appropriate first email is a collection of best hits from the archive, not the latest piece.
We think there’s lots of potential in structure, but we can’t just add structure for structure’s sake. We like narratives because editorial can get a grip on the concept and (aside from the challenge of deciding on new terms!) they get used quite heavily.
Our litmus test for adding metadata to our stories: how would we use this information on the site, in applications, or for our analytics dashboards? Do we have a user-friendly interface for adding the necessary information, knowing that someone will have to annotate every piece of content we publish? If we come up with new story forms, are those the kinds of things writers would enjoy working on?
You wouldn’t be the first to notice that news online really isn’t that much different from news in a newspaper: a header, some text, some images. Hey, it’s a winning formula. We think it’s subtle things like narratives that can ultimately make news on your computer and on your phone a nicer experience than reading something on paper.