More pictures, less policy: Ezra Klein’s Wonkblog launches viral site Know More

Ezra Klein’s popular politics and policy blog Wonkblog is already a digital success story. Klein tends to be mentioned in the same breath as Andrew Sullivan of The Dish and Nate Silver of FiveThirtyEight: He’s a highly influential blogger whose digital brand is more valuable than that of his corporate parent (Wonkblog is owned by the Washington Post).

But, Klein says, he and the team at Wonkblog wanted to go farther in creating a viral site. On Monday they launched Know More, “a site for people who like learning stuff.” In an introductory post, Wonkblog reporter Dylan Matthews explains:

“Each post is a picture, chart, video, or quote that, we hope, will fascinate you, or fascinate a friend who shares it with you. But at the bottom of every post is the option to really ‘Know More.’ Click on that button and we’ll take you deeper.”

The site is highly visual, consisting of a Pinterest-like grid of images and text. Headlines are designed to be catchy — “This congressman just said the scariest thing anyone has said about the debt limit,” “What your state sucks at, in one map.”

wonkblog know more

I asked Klein a few questions about Know More and the strategy behind it. Here’s a slightly edited version of our email interview.

So this is Wonkblog’s attempt at creating a viral site. How did you decide to launch it and what were the models? What did you think sites like Upworthy and Buzzfeed were doing wrong that you can do better?

Ezra Klein [EK]: It’s not about what those sites are doing wrong. Buzzfeed and Upworthy are amazing. It’s about something we thought we could do differently. A frustration I’ve had running Wonkblog — and simply loving journalism — in the age of social media is that a lot of the best work we do simply isn’t built for social. The part of a post or article that might really engage someone often isn’t in the lede and doesn’t make sense as the headline.

So in a way, Know More is an effort to break the slow web into parts that work for the social web. We can go and grab that one quote or one fact or one graph or one video or one whatever that we think is most likely to make somebody want to learn about a topic or issue and then, using the Know More function, we tell them how to go learn more about that topic or issue.

How much of the content is drawn from posts that are already on Wonkblog, and how much is separate? How are you deciding what to put on Know More?

EK: A small minority comes from Wonkblog. Know More isn’t a vehicle to promote Wonkblog. It’s a vehicle to find wonderful things people might want to learn more about and then figure out the way to convince them that they want to learn more about them. And those things live all across the internet.

wonkblog know more

Do you think that your audience will primarily be the people who already read Wonkblog, or are you going for a completely different audience?

EK: Completely different audience. I’d be ecstatic to see Know More develop a large audience that simply isn’t that interested in Wonkblog. We really hope Know More scratches a different itch: It’s more about the joy of learning new things than it is about minute-to-minute coverage of policymaking. If the audiences are exactly the same, that means we’ve done a crummy job fulfilling the purpose of Know More.

Are you testing different types of viral content? The headlines on Wonkblog are often really good for the Twitter age already — what have you learned about headlines that you’re applying to Know More?

EK: We do internal focus grouping where the writer comes up with a series of headlines and the whole team votes on them. Over time, we want to be doing more A/B testing too.

[Re: headlines], you need to come up with a bunch and you can’t trust yourself to know which one is the best one.

Was the Washington Post supportive? 

EK: The Post has been super supportive. The most important resources we had were Yuri Victor, who’s just an amazing developer, and Melissa Bell, who’s an amazing digital thinker. Beyond that, this isn’t so much a resource-intensive project as it is a trial-and-error intensive project.