Storytelling with data is having its moment in the journalism world. After leaving The New York Times last year, Nate Silver has launched his new data journalism venture with ESPN , former Washington Post blogger and columnist Ezra Klein is expanding his chart-filled approach beyond public policy at Vox.com, and both the Times and the Post are starting their own data-driven sites to replace their departed stars.

While none of these ventures rely solely on charts and data to serve readers, The Post’s is explicitly betting that data on its own isn’t enough. Its forthcoming policy-focused website is premised on the notion that data and narrative go best together. The lesson for anyone looking to communicate or persuade using data is that the addition of human drama goes a long way.

I recently spoke with The Post’s Jim Tankersley, the economic correspondent leading the project, about what he sees as the benefits of pairing narrative and data to get your point across. An edited version of our conversation follows.

I wanted to start with this idea of “In One Chart Journalism” to get your sense of the ways in which this has succeeded, what’s missing from it, and if it is fair to see this as a starting point for what you’re building.

I love “In One Chart” because I think sometimes that way, and I know a lot of people who do. Where I think it’s been really successful is in drawing people into something they might not have read a 700 word article about but can understand the thrust of quite quickly. And then very often they go on to read that article.

But I do think it’s just a jumping off point. We will do cool data visualizations on our site. But even in the context of those visualizations, we want to make them do a better job of telling stories. We’re going to be very interested in the sort of data plots that really help you understand change over time — a narrative thread that you can represent visually. If we can give [readers] themes, and narratives, and plot, that helps them understand really complicated concepts. And I think you can do that visually. People already do it, they do it really well, and we want to build on those kind of successes.

There’s a sense that the narrative, the story, is a human thing that can draw someone in. Embedded in how we talk about that is the idea that the data is the medicine. You’re going to draw someone in with a story so that they’re consuming the stuff they didn’t want to consume — the data. One of the things that this “chart journalism” has done is make the chart the part that jumps off the page, what draws people in. How do you think about which is the lure and which is the substance?

Data has gone from a thing that we try to bury to now the shiny thing that makes you want to swallow it. I do still think there are things that you can’t do in charts, that you have to do in exposition, that by themselves are not that sexy if you just made a story around the numbers.

Data and visualizations and storytelling — human drama — when combined help readers understand concepts. I can get a lot from one great visual chart, I can get more context from several paragraphs just really laying out the conclusions of data. And I can get a lot out of a story about someone like me doing something that is interesting to me. But I think that when you bring them all together what you get is this experience where people have a process of discovery for themselves which I think is important for learning. If you are just looking at a chart you can try to figure out what it’s telling you and there’s a little bit of discovery as [you] decipher it. If that chart occurs in the natural flow of narrative or storytelling, then I think it’s almost a clue popping up as you’re trying to solve this mystery. I think that when they all work together, they’re more powerful than any of them alone.

I can imagine a case where you find an interesting story and you find data to buttress it and to better tell that story. Or the reverse: you find a really interesting dataset but you think you need a story or a narrative to really draw people in. There’s potentially a little bit of a tension here. As a reporter there can be an impulse to go out and say “I want the best possible story I can find.” The most extreme, interesting, compelling, sometimes bizarre story. And often that can be in tension with a dataset. How do you find stories that are doing service to data and vice versa, rather than having them be in conflict?

It’s quite often that you find stories that lead you then to, “Hey is there an interesting data point that can help flesh this out?” Any really great story will be a window into something bigger. And if you look hard enough you can figure out where that window leads you. I think it’s easier — and it’s going to be by far the dominant type that we’re going to be doing — that the story works the reverse way. That you start with the question, you start with the data, you start with the what are we trying to figure out here. Once you come to a good conclusion based on your own analysis of the data, then you go out and find the story.

A lot of time reporters get caught up because they want every person that they feature to be perfectly emblematic of a central tendency in a dataset. But the truth is it’s really hard to find someone who represents that entire thing. So what I like to think about is “Who is someone who gives us the best entry point here, who represents something about this.” And you’re honest with your readers about the things that they are not representative of, but you focus the camera most on the parts where they are.

I don’t think there’ve been a lot of times where I’ve read a story and thought “Wow, I really wish this guy more perfectly represented the dataset underlying what the writer’s trying to tell me.” But there have been plenty of times where I’ve read a story and thought, “This person’s not interesting.” And being interesting is a really important way to get people to read and digest what you’re trying to help them learn.

How are you thinking about trust with the reader, especially since they’re trusting you to handle data properly?

We have an extra obligation to try and earn that trust. People are naturally skeptical of what they read in the newspaper or on the internet, and they should be. We want to build this relationship with readers where they feel like we are treating numbers honestly. We owe it to our readers to be constantly keeping their skepticism in mind, and try to be very very careful with how we use numbers so that they don’t feel like we’ve misled them in any way.

How much of this kind of storytelling is made possible by access to data that wasn’t available before?

It’s absolutely easier. It also ups the expectation for what a good story is going to be. I think the easiest place to see this is in sports reporting. There’s always been a basic set of statistics available to everybody, but now if you’re trying to make the case for why a guy’s contract is fair and all you’re talking about is RBIs you get laughed out of the room, even by a casual newspaper reader. I think that’s good.

To what extent does this phenomenon go beyond journalism — the way that politicians communicate with citizens, or the way that businesses communicate to their investors or consumers, etc.? Are we going to see data as storytelling increasing in other areas?

Think about what really successful trial lawyers do, what really successful CEOs do with their boards and shareholders: they tell stories, they use numbers, and they make arguments. I don’t think it’s unique to journalism, but I hope that we can do it in a really beneficial way to help readers.

This blog first appeared on Harvard Business Review on 4/28/2014.

View our complete listing of Human Capital Analytics blogs.