Deciding how a commenting system looks and operates are choices publishers should make with intent. These decisions affect the nature of readers’ and moderators’ experiences and the content of their contributions.

Talia Stroud, associate professor of communication studies at the University of Texas – Austin, studies how the media affects our political behaviors and attitudes and how those attitudes and behaviors affect people’s use of media. Before deciding what type of commenting system to go with, Stroud said to consider:

  • The type of comments you would like to see on your website
  • The culture of your publication and audience
  • The design components

Consider how these three factors could work together to create a unique participatory environment for your readers. If you create a system that incentivizes a certain type of comment by elevating them, like “NYT Picks” or Gawker’s Kinja stars, a community can rally around the system to drive engagement and discussion value. “These incentives could be very different depending on the newsroom,” she said.

Tommy Craggs, who is the editor-in-chief of Gawker’s sports website, Deadspin, said that this “discussion” can take on multiple forms. ” A joke followed by a round of virtual applause is, as far as Kinja is concerned, a lively discussion.So is a well-informed exchange about hearsay exceptions,” wrote Craggs when the platform was introduced in 2012.

The New York Times selects the “most interesting and thoughtful” comments, expressing a variety of views, to be featured prominently in its “NYT Picks.” These may also be chosen to highlight discussions from a certain region or readers with “first-hand knowledge of an issue.”

What are some incentives news organizations could provide to encourage desirable comments?

Incentives could be in the form of voting, ranking, highlighting, or even through money.

Quartz rewards “great annotations” by both responding to these annotations and displaying them “more prominently.” On the flipside, the news organization removes others that are “off topic or abusive.”

Quartz "annotations" appear in the margins next to specific paragraphs.

Quartz “annotations” appear in the margins next to specific paragraphs.

Cory Blair sought to understand and learn about the Kinja’s audience through “becoming an active member of it.” In his quest to become a “Kinja star,” Blair wrote that Gawker thought carefully about what types of conversations it wanted to take place on Kinja, including the motivations and incentives Gawker uses to encourage “quality content” on the platform. Great Kinja commenters can even be contacted to write for pay through the company’s “Recruits” program.

In announcing the Recruits program, Gawker’s editorial director Joel Johnson wrote that some of the site’s “best writers” rose up through “commentariat,” but that not every commenter is fit for what is required in “real reporting.” And, with this, the idea for Recruits was born.

Recruits are paid, short-term contractors. They receive a stipend, and their own “Kinja subsite attached to one of eight core brands, mentoring and direction from our site leads and their staff, as well as an aggressive bonus structure that will reward the capturing of uniques (using the same Quantcast ‘people’ metric we use across all the sites),” according to Johnson.

How do I get rid of abusive and off-topic comments?

Sometimes quality can get lost in traditional commenting systems as they are posted chronologically and can be overrun with incivility.

More advanced systems like Gawker’s Kinja have also played host to incivility. Gawker’s “weakness” became especially clear this summer when one of its publications, Jezebel, became bombarded by obscene pornography and violent trolls.  Consequentially, Jezebel had to turn off the upload ability and shut off comments on select articles, and Gawker disabled photo comments. As of the publication of this report, no long-term fix for the issue has been decided.

Some organizations have opted to get rid of comments. Last year, Popular Science chose this course, citing a University of Wisconsin-Madison study that found that uncivil comments polarized readers and “often changed a participant’s interpretation of the news story itself.

Others approach this issue by adding a layer of structure into the commenting process.  Another option is only allowing comments occasionally, or on select articles. The New York Times allows commenting on 17 articles a day, which are selected based upon four factors: the story’s news value, predicted interest, staff’s ability to moderate the estimated volume of comments in a “timely fashion” and whether an article on the same topic was recently open for comments.

Additional features can also assist in filtering out low-quality comments, such as voting, ranking, starring, or liking, which helps create a space where “the good stuff stays at the top and the bad stuff recedes,” writes Bob Cohn, publisher and chief operating officer of The Atlantic. The Atlantic also utilizes its most faithful readers to help moderate comments.

There are some additional variables to consider, such as your time and capacity to moderate and engage with commenters and whether you would like to “own” those comments.

Are your reporters willing to invest a little bit more time? Do they have the capacity?

Talia Stroud and her fellow researchers at the Engaging News Project, which aims to “provide research based techniques for engaging online audiences in commercially viable and democratically beneficial ways” studied: What happens when journalists take a more active role in comments sections?

They teamed up with a local news network in a top-50 market to study what impact a reporter’s interaction,  a station’s interaction and no interaction would have on the TV station’s Facebook community. The study found that when a reporter engaged readers in the comments section — by  replying to questions, asking direct and open-ended questions, providing additional information and recognizing and highlighting good discussions — incivility decreased. Specifically, the study found:

  • The chances of uncivil comments declined by 15 percent when a reporter interacted with a post as compared to when no one interacted with a post.
  • When a reporter posed a question that only had a few options, such as a yes or no, chances of uncivil comments to a post declined 9 percent in comparison to posts that included a question or those with an invitation for readers to comment.

Though the task of engaging in comments sections may seem “onerous” or another to-do item, the reporter in this study did not expend “extraordinary efforts,”commenting on average four times.

Are you willing to move your comments over to Facebook or another free third-party service?

In a comparative analysis of incivility in online political discussion, Ian Rowe, Ph.D candidate and assistant lecturer at the University of Kent, compared the incivility of comments on The Washington Post’s website to the Post’s Facebook page.  (Note: the Post does not use the Facebook Comments plugin so this study refers specifically to the Post’s Facebook page.) Rowe found uncivil commenting behavior to be significantly more common on the website version of the Post than on the Post’s Facebook page. He hypothesizes this is because on the website, users can remain anonymous, whereas on Facebook, “commenters , are identified with, and accountable for, the content they produce.” This introduces one of the major problems with comments, which is tied into issues of anonymity. There are pros and cons to requiring commenters identify themselves.

Rowe also concluded that the poor behavior in the discussion was more directly targeted toward others when the exchanges occurred  on The Post’s website. The conversations on The Post’s Facebook page were “less likely to be interpersonal, and more likely to be aimed at individuals not involved in the discussion, or used as a way to articulate an argument, rather than offend others.”

Depending on the size of your organization and your goals, Facebook Comments, or another third party platform may be what you’re looking for.

In APME’s recent survey, nearly 57 percent answered that their news organization uses a commenting hosting service. Of these, nearly 62 percent use Facebook Comments as their commenting service.

However, before handing your ownership over, consider the following: Do you want to increase traffic on your page?

Using Facebook Comments can be a way to drive traffic to your website. If a user comments on an article using Facebook Comments, he has the option to share this comment on his Facebook feed. If he chooses to share his comment on his own feed, any of his friends may comment on this. This new comment, made and posted on, can also be seen on the original article’s website.

However, Zucker-Scharff writes that third party sites can exclude you from SEO opportunities too. Using the Facebook “Comments” plugin means that your comments, and all the “data usage” that comes with them, are shared and stored on Facebook’s servers. This alarmed some in the industry, like Dan Gillmor, director of the Knight centre for digital media entrepreneurship at Arizona State.  “The billion-user social network has emerged as the news business’ most potent competitor, capturing attention and advertising on the path to monopoly status,” Gillmore wrote in The Guardian. “Ceding such a crucial part of one’s online operation to a competitor has always seemed insane to me.”

Zucker-Scharff wrote what troubles him the most about using Facebook Comments is giving up control and ownership of these comments on your site. “Even worse, it seems you will never be able to take that control back.” And, what if the system were to “suddenly disappear?” he asks. (It should be noted that DISQUS does let you own the comments, and data, posted through its portal.)

This is one of the reasons that Facebook did not work as a commenting platform for The Seattle Times. “Having Facebook as your commenting system means you have no control over the data and the functionality,” Payne said. “They own those comments, you don’t.”

Payne said the Times installed Facebook commenting as a test on one of the paper’s busiest local-news blogs a couple years ago. Payne said Facebook didn’t work for them because the “the volume of comments went down substantially,” which he said was because people either didn’t trust Facebook, or they didn’t want their real names to appear alongside their comments.

“When you have less volume, you have less constructive dialogue and more drive-by comments,” Payne said and added that the paper still had to ban commenters who broke the rules on Facebook, attacked other commenters and “trolled using their real names.”

More volume in the social sphere doesn’t necessarily imply that you’re building a community. “Social media, and particularly viral successes, seems like a much better way to build an audience — on paper,” said Dooling.

Dooling points out that in her experience, commenters spend more time on-site and visit their favorite news organization’s website more times a day. “They are also much more vocal, which is great for feedback on your website and very different than a social media reader, who may stop by because they were coerced by a snappy headline, and don’t have an invested knowledge in your news,” she said.

Share with your network

You also might be interested in:

  • Successfully and efficiently marketing your work can be hard, especially for local news teams with limited resources, but marketing yourself to your audience is an essential skill for news organizations to drive revenue and promote sustainability.

  • As news teams begin thinking about their election coverage plans, it may feel like adding more tasks to an already full plate, with a fraction of the staff and resources they once had. But that doesn’t have to mean figuring out how to do more with less — maybe it’s doing less with less.

  • We reached out to Danielle Coffey, the CEO of American Press Institute’s parent corporation, the News/Media Alliance, to learn more about the legal fight for news organizations’ rights with AI.