How NHPR Uses Related Content To Keep Site Visitors Around

Jan 15, 2014

We sat down with NHPR Digital Director Rebecca Lavoie to talk about their recent efforts to drive engagement to their site.

Credit Todd Mundt/NPR

When someone clicks on a story link and lands on our site, we want them to stick around as long as possible. Most users don’t; but a small percentage do, and what causes them to stay longer? Our compelling content. Most blogs and news websites use Related Content to try to keep the audience around a little longer: “If you liked reading this story about apples, we think you may enjoy these other stories we have about apples.”

Lavoie asks, “But what if the reader who loves apples is also interested in oranges?” She’s been thinking about related stories and how to harness them more effectively to increase engagement on the NHPR site. Read the interview...

How would you make the case for the value of related content?
Populating the related content field is an open invitation for a story visitor to consume more of our content. Not doing so is an invitation to leave our site. Think about it as you might about taking breaks on the radio; you’ll rarely hear a well-trained host say, “we’re taking a break.” Instead, there’s a constant flow of forward promotion of what’s coming next, and how it ties to what the listener just heard.  

With an increasing number of our web visitors bypassing our homepage and going directly to NHPR story posts via referral, we should aim to replicate that “listener-focused” story flow experience with our site visitors whenever possible. Home pages are designed to do this, but story pages are not.

On the ground for reporters, related content is also a great opportunity to give older, evergreen stories fresh life, and even for reporters to promote their own work.

What, in your view, makes for successful related content?
Content that makes the reader click on it without thinking about why it’s there. Remember the last time you visited a news or aggregation site and just clicked around? You probably made choices you thought were your own, but that content was there for you to choose because someone put it there, either a person, or a pretty sophisticated algorithm. I believe that in public radio right now, people are best equipped to make those decisions, because we know our audience so well.

How would you evaluate the stories that typically show up as related content for NHPR stories?
Core Publisher makes suggestions, and sometimes they’re good ones. Often, though, those drop-down suggestions are old news or related to a story’s subject (say, the Governor), and not what the reader might be more inclined to want.

When I worked on our program Word of Mouth, it occurred to me that stories should be categorized a bit differently when it came to making related content choices. For example, if a reader clicked on a story that had a quirky topic or headline, I would try and think of other quirky stories we’d done, or other headlines that popped. I’d tell other producers and interns to think, “If someone likes this, they’d probably love that!”

This idea works for news just as well. If, for example, we report on the legislative debate over raising the minimum wage, we should consider readers attracted to the story might be interested in other current policy debates in our legislature, like the gas tax, Medicaid expansion, or casino gambling. The instinct might be to simply go for all minimum wage-related content, but what is motivating that reader, an interest in that particular policy or policy in general? Knowing the level of political engagement in our state, it’s more likely the latter.

That’s not as easy as letting the CMS generate a few stories based on some keywords.
Thinking about related content this way relies on the memory and instincts of the reporter or producer posting a particular story. We all know that churning out stories on deadline often means we forgot the work we did yesterday, never mind a few months ago. One trick is to use the related content fields themselves as a search tool. If you vaguely remember, for example, that you or a colleague once wrote a story on heating oil or sled dogs or anything you might be looking for, type in one of those words and look at the headlines that pop up. (It pays to double check stories to make sure they’re a good fit!)

NHPR’s reporters and show producers are absolutely receptive to the idea of always, always, always populating related content, but it’s still a task we are trying to make a routine part of our regular workflow. I never mind jumping into a published story and filling in those fields myself, but I don’t always notice that they’re empty. As with other tasks we’ve incorporated over time, I’m sure we’ll get there.

How have you been working with your reporters and editors to help them make good choices?
I’ve found much more success in talking with reporters individually, rather than creating blanket policies around posting content. Everyone’s job and work style is different, so I’ve been finding opportunities to make this point with content makers using their own stories or beats as examples, and thinking about whatever motivates or interests each reporter. For continuing stories, it’s easy to make the case that previous coverage is a good fit – one that comes to mind from our recent New Hampshire news is the story of a missing New Hampshire teenager named Abigail Hernandez. Our reporters shouldn’t assume that readers of updates on that story have read everything else we’ve posted, so we should give them that chance to catch up on the content. And, of course, self-promotion can be a convincing argument. I would rather a reporter post his own work in the related stories fields than nothing at all!

The method for adding related content to a post is most often based on keywords or tags in the CMS. It’s straightforward and easy for just about anyone to drop in some tags and get a list of stories, but that also exposes the weakness of the process: it’s great for related content but not for truly interesting content. How would you make this better while still keeping it easy for reporters posting stories?
 I’ve been thinking a lot about tags lately. As you point out, they don’t do the very human job of answering that question, “What would someone who likes this story also want to read or hear?” One solution could be to create some custom tags such as “quirky,” “outdoorsy” or “evergreen policy,” and have the newsroom and our programs maintain limited lists of these in order to make interesting content easier to identify. This is something I’m planning to experiment with, but in a limited way at first. As those who work on the web know, tag management can be tricky unless there’s good communication around why they’re being used and how. That will definitely be a big part of the process.