User comments are just another form of social media.

They’re a way to connect readers with writers and with each other. They’re a way to extend a story and drive interest through conversation.

Comments can be critical though; even rude or racist as they approach the apex of Godwin’s law. And they’re on your site, published under your brand.

Some highly active sites like Huffington Post have dozens of staffers dedicated to keeping message boards lively, clean and relevant. Other networks, like Gawker, have developed algorithms to automate the process.

For most publishers though, the solutions are somewhere in between.

The First Line of Defense: Registration
Registration is an obvious safeguard against spammers, but for Steve Saunders, managing director of integrated marketing firm UBM DeusM, it helps maintain the editorial integrity of the brand.

Particularly in the b-to-b space, Saunders says, commenters can be extremely valuable resources. They’re leveraged as moderators and contributors on his sites, so it’s vital that the community remains professional.

“Our philosophy is, the audience actually knows more about the subject matter in b-to-b than the journalists do,” Saunders says. “The message board is where the value is, so you have to make sure that the people on it are qualified to be having that conversation. If you have just anyone commenting, there’s no value.”

Aside from getting an email address—“99 percent” of all disrespectful commenters will stop after receiving a personal email from a moderator, Saunders says—registration forces users to agree to the site’s terms of use.

The legalese is often overlooked, even by site managers, but it’s critical in establishing community guidelines.

Finding the Sweet Spot
Moderating comments has gotten harder with scale for Jack Shepherd, a community manager with BuzzFeed.

The site’s eclectic and expanding list of verticals—from politics and tech to food and animals—means that commenters can’t be governed by a set list of rules anymore. Appropriate behavior is determined by context.

“The idea of managing your website is like if you invited people over your house for a dinner party,” Shepherd says. “It’s fine for someone to disagree with you if you’re having a political discussion, but once they start saying horrible things or accusing you of something, that’s when you say, ‘OK, you’re not coming over again.’”

Inappropriate behavior can be more subtle though. Having a foreign policy discussion under a Beyoncé story isn’t technically abusive, but it kills the thread.

However, context gets muddled by platform. Facebook posts and BuzzFeed’s own comment stream are visible under each story—but their rules are different.

There are tools to handle disruptive members in each area, but most of the decisions are manual. For now, Shepherd relies on common sense.
“I have one rule—don’t be a [jerk].”


Predefined Comments
BuzzFeed integrates a set of predefined tags called “reactions” into its comment stream. Users can assign “LOL,” “Cute,” “Geeky” or a number of other monikers to a story, in addition to comments. As part of a larger algorithm, the responses are used to assign stories to landing pages within the site.

The New York Times has drawn attention for adding similarly predefined elements to its comments, most recently in its report on the election of the new pope. Contributors were asked to identify themselves as “surprised” or “unsurprised,” their reaction as “positive” or “negative” and whether or not they were Catholic. Comments were posted under a heading that encompassed the predefined responses, and were sortable by each of the categories.

These forums offer users a lot of options, but even binary systems like up/down voting can be a useful add-on. Adding defined elements to comments can give vibrancy and structure to what can sometimes turn into an overrun section.