The empty comment box — dealing with the scourge

It’s 2017 and I still see this guy everywhere I look:

Technically, it’s webkit’s default textarea in focus, but let’s call it by what we know it represents: A comment box. An empty comment box.

In some form or another, it dots the bottom of most content on the internet. It’s become so ubiquitous people no longer question why it’s there or if it’s a good idea. At this point, I’ve seen some incarnation of these little boxes added to new product designs almost as an afterthought, as just ‘something you have’ or even ‘something you deal with’.

So let’s talk about this little pest. Let’s talk about how it got there, and what to do about it.

Engagement metrics

Once upon a time, advertisers would only concern themselves with pageviews. The more eyeballs you had staring at your web page, the more your web page was worth as advertising real-estate.

Naturally, this led to people using tricks to get more of those eyeballs on their page. Sensationalism, SEO manipulation, whatever worked. We’re familiar with the byproducts of this era: misleading pages that didn’t really contain useful content, but did just enough work to get a click from the reader, a.k.a clickbait.

Eventually, readers got wise to these ploys and would instantly close pages that felt even a little clickbait-y. Advertisers realized this was happening, and consequently realized their ads weren’t registered by their audiences. Sure, they were seen, but not long enough to be absorbed. They now wanted proof that readers were awake and attentive enough to take in their ad’s content.

Just like that, engagement metrics were born.

Comments, along with likes and shares, are a part of this newer family. A comment is supposed to be proof that the reader has actively thought about the contents of the page, moreso than if they were casually browsing it (We’ll revisit this idea later).

Just like before, people started to “game” the engagement metrics. They started producing content designed to generate more comments and shares (a.k.a “fake news”) but not really inform people.

Subtle design decisions were also made along the way. Any change that led to more comments would be incorporated in the overall product. If you designed a big green button that said “COMMENT HERE”, you can be sure everyone would copy it if it were proven to increase the comment count.

Unfortunately, the rest of the internet started interpreting these subtle design decisions as universally ‘good’ and began to adopt them everywhere, even if it didn’t make sense in their brand or industry.

So, that’s how we got here. The current incarnation of comments are easy and prolific because they make advertisers think your page is valuable real estate, and everyone else just slaps them down as an afterthought because they don’t know the history behind the design and assume this is just “how it’s done.”

Perceived benefits

Now that “comments everywhere” is becoming more and more accepted, I’ve heard people reverse-justify them. They claim benefits that have never been shown, but kind of “sound good”. I want to quickly address a few of these before going further.

Community creation purity

Some developers and designers take an incredibly purist view of comment areas and topic discussion pages. They see them as a bastion of freedom, and as such, want to leave them to grow organically so their user communities can express themselves in a natural and unbiased way. They might even drive new features based on this “community” and what it might create.

To such people, I’d like to ask: what happens if you have a big, empty wall with a can of spray paint sitting at the bottom? How do most people react?

Most don’t. Most people are socially cautious and look to see what others will do before they act themselves. You don’t want to be the jerk that didn’t know the wall was actually part of an art installation and now you’re being sued by the city because you defaced it. So, you sit back and see what others will do. Who does that leave? Who would go up to the can of spray paint? Precisely the people you’ve just filtered for: the people who don’t care what anyone else thinks; the ones who will draw whatever they want irrespective of social cues (or a lack thereof).

Ok, some of these folks will genuinely create something positive and people will come in from miles around to look at it. This is the gamble you’re hoping to win. But, of all the graffiti in the world, how much of it adds wide appeal to its target, and how much of it is just mindless swearing? Statistically, you’re far more likely to get a rude word or a swastika. Why? Because you’ve designed in a way that specifically appeals to those who would want to cover your page in such content (the ‘lack-of-design’ design).

And with that, they’ll create new social cues. “This is a place for rude words”. That becomes the message.

There’s a commonly accepted patch to this design flaw: adding moderators. Unfortunately, this takes a toll on people who do it professionally. It also adds a new layer of agenda-shaping censorship that you were trying to avoid with your lack-of-design, only now some chunk of your potential users have already been scared off.

Rather than dealing with a problem you know will occur, why not start off right and explicitly design around the interactions you want your potential community to have? (More on this later) I get that you’re trying to play amateur scientist or sociologist, but, at this point we know weeds grow in an untended garden. You don’t have to be the millionth person to prove it.

Comments as feedback

Comments are data, and data is good, right? And the only thing better than data is more data. Even if people write off-topic comments, we still learn something, right? Any comments are better than no comments, right? So let’s try to ask for comments and be encouraging of maximizing comments for no reason other than to maximize the amount of feedback, and actionable data.

Well, let’s look at what we’re doing here. We’re literally asking people to tell us what’s on the top of their heads before they’ve even gotten a chance to really think it through. We create live timestamps, auto-scrolling new content and counters to illicit a sense of urgency, combined with messaging that talks about how appreciative we’ll be if they share whatever is on the top of their heads right now. We know all of these techniques create more immediate comments.

But, let’s step back and look at how meaningful feedback is generated in humans. People take impressions in from the world, process them into thoughts, refine them, process them further, and eventually produce something reasoned and meaningful at the end (sometimes).

“[I]f ever any talk should happen among the unlearned concerning philosophic theorems, be you, for the most part, silent. For there is great danger in immediately throwing out what you have not digested.” — Epictetus

We’re effectively arresting that process by asking for the undigested contents of their subconscious, because we see more of that as valuable, not realizing that we’ve stopped the end product from ever being created. So we end up with raw, unprocessed, seemingly random thoughts, with the irony being the more aggressively we elicit this feedback, the more raw and useless the thoughts we get back tend to be.

So, why not work with that human process if we want something well-reasoned?

Solution: To Advertisers

“Fake news” exists because it’s economically viable. We need to look at the central conceit that makes it all work: the idea that comments on a page indicate that readers are more involved, using more of their brains to process the contents of the page than they would be if they just read the article and stayed silent.

Advertisers, take a look at the comments on most pages. Tell me, have the commenters really read the article? Do you have any proof that they’ve actually looked at the page at all (including the ads)? Or, have we just created a weird system that lets a small subset of antisocial readers produce non-topical rants?

Even worse, these non-topical rants can hurt your brand. We know users judge the page as a whole. This includes content, ads, and comments together. Do you really want those half-digested antisocial comments to be associated with your brand? Even worse still, do you want to be directly paying for this devaluation?

To designers and developers

We need to set the agenda for our pages and applications, lest it be set for us. The most important thing we can do is ask for as specific feedback as possible, avoiding open-ended generic text whenever we can.

Do you want to know if the readers found your page entertaining? Put a button down there that they can click to indicate whether or not it was entertaining. You’ll get higher participation rates and more relevant data than if you go fishing for random subconscious streams-of-thought.

What if you want text, such as corrections? Be explicit. Create a prompt saying “For corrections, use this address/form” and be sure the address it sends to is something explicit like corrections@example.com, and that the reader knows and sees this.

The more specific feedback you can elicit, the better. If you’re building a page or an app and you really do want open-ended generic feedback, it means you haven’t defined any solid success criteria, and you’re hoping your users will somehow do that for you. They won’t. That’s your job.

So, think through what your product or content is supposed to accomplish, and then you can check with your readers (directly or indirectly) if it has accomplished that specific goal.

Meaningful comments

When you want something deeper, like thoughts, opinions and reflections, allow your users the luxury of actually reflecting. Let them come up with multiple drafts, request feedback, and ultimately share their thoughts with the content creator(s) to be included as a supplement to the content. Set high standards, but make a deal with them: your cognitive time for ours. If your comments truly add to our content, we will wholeheartedly engage you as a partner and incorporate your comment or reflection into our page.

Sure, there may only be 3 or 4 of these kinds of reflections as opposed to hundreds of comments, but they’ll be of a much higher quality, and ultimately bolster the content you created. You can even put a little “agree/disagree” button at the bottom of them if you want to know what your readers think of the additions.

TL;DR

As an industry, we’ve been experimenting with comments for a while now. We know that if you do them the way we’ve been doing them, you’re almost certain to get high-quantity, low-quality comments. The original advertising reasons for these comment factories were tenuous at best and the social damage being caused is enough that we should re-evaluate that broken logic. The worst thing we can do is just accept comments as they exist now as “standard design”.

Instead, let’s do what we always do: iterate and improve.