Comment sections are on their way out
In 2007, when I launched my first website, our comment section was central to the product. We scrutinized every detail of it. We incessantly debated the interactions — should we allow ‘up’ and ‘down’ votes? — and considered “comments per article” to be a major KPI.
Back then, a discussion of websites’ comment sections felt like a bunch of sixteen year olds talking about their cars.
“Mine is the best — check out the way we elegantly nest the response comments in up to two layers. Then, we sort them reverse-chronologically and put an [email protected] symbol to highlight the response. Of course, you can vote them up, but we decided to eliminate the down button. For simplicity purposes, of course…”
Our industry’s efforts to one-up each other’s comment sections felt like the business card pissing-contest in American Psycho - and that comparison now seems fitting in hindsight.
Here’s why: though it was never the intention, time has shown that comment sections often create a negative space and provide a negative experience — to the reader, the writer, and to the publisher. That is one of the defining reasons my website — Bustle — made the conscious decision to launch without comments, as reported last week by Slate columnist Amanda Hess while discussing Jezebel’s recent struggle with comment sections.
Because several others writers have discussed the issue of comment abuse and bullying in great detail, I won't elaborate on this primary reason for avoiding them. It is, by far, the main reason to eliminate comments. Both the ongoing struggles at Gawker, and the stunning attacks on Zelda Williams serve as heart-wrenching pictures of trolling.
Rather, I will explain the business reasons for avoiding comments.
Simply stated — the real estate traditionally devoted to comments is too valuable to be wasted on a component that adds minimal value. This slot is typically found directly below an article’s body text and, on Bustle, it is now used primarily for “related content”.
The shift to hosting “related content” underneath an article has taken place over a two or three year period of time, and it was catalyzed by two major events — both of which were advanced by companies like Outbrian and Taboola.
The first was optimization of related content. Picking the stories that readers will want to click is not easy. In fact, it requires incredibly advanced algorithms (or an expensive manual selection), but those algorithms do work. And when they do, it can double the engagement of your readers. Last week, I spoke to an executive at a popular website who told me that her average page views per visit increased from 1.3 to 2.2 over six weeks with the help of Outbrain. That additional page view per visit could generate massive revenue for her site.
The second incentive was also a monetary one. Some websites will pay for other websites to link to them. In many cases, websites use Outbrain and Taboola to both buy and sell traffic. How does this make sense? Well, it has to do with inventory levels. When a website is running a branded advertising campaign, it may need to buy traffic to fulfill that campaign. During windows between campaigns, they may as well sell outbound links. It’s nothing more than an inter-site efficiency operation.
Optimizations to ‘Related Content’ have driven KPI’s, but that is not the only reason why publishers have turned to them. Over the last few years, publishers have better come to terms with their identities as… well… publishers. Those who create great content are just as happy to let the discussion happen elsewhere — Twitter and Facebook for the most part. This doesn’t just enhance the publisher’s focus on what they do best, it also drives meaningful traffic when Tweeters link back to the article in the context of discussion.
But the reason to avoid comments does not end there.
Advertisers have never loved comments — and, in fact, publishers long struggled with how to address the “freedom” of allowing comments while bowing to advertisers’ taste for clean content without the thorns and pitfalls of comments. Sites like the WSJ ended up splitting the baby by putting comments on a separate tab. For its part, NYT answered with heavily curated comments that required costly moderation. Neither is a particularly optimal path.
So why do sites still allow comments? The answer — which is sadly prevalent in every aspect of publishing — is the least satisfying of all: tradition. Comment sections have been a convention for nearly a decade now, and publishers are hesitant to eliminate them. But some new sites like Bustle are launching without them. Medium has abandoned the comment thread as we know it, in favor of an optionally-visible notation system. If a reader disagrees with an article or wants to add their opinion to it, they may do so on Twitter, on Facebook, on other platforms, or in real life. Heck, they can still write a Letter to the Editor at some publications.
A year ago, people mocked our decision to avoid comment threads. Our refusal to provide a comment section was heavily attacked by tweeters and commenters (on other sites). Upon learning that Bustle would not have comments, one anonymous critic said this:
To not allow comments smacks of either haughtiness or fear (of people pointing out errors in the article--either grammatical or logistical). Generally, sites bring about commenters who are as good in quality as the articles themselves. To not allow them, Bustle seems conceited, anachronistic, and afraid.In fact, the exact opposite is true. The Internet is constantly in a state of evolution, which means that we create (and destroy) conventional features on a regular basis. If comment threads go the way of popup ads, I think we can all live with that. Over the last few years, the web has provided us with outstanding tools and platforms for debating content, eliminating any sort of ethical or journalistic mandate to present an on-article commenting option. In short, the forward-thinking move was to get rid of them altogether.
Better for the user. Better for business.
Editor’s note: This is a guest post by Bryan Goldberg. The post went through PandoDaily’s usual editorial process and Mr Goldberg was not paid for his work.