StackOverFlow/Reddit/Quora — a non-exhaustive list of platforms that have popularised the exchange of ideas over the internet. Promoting collective participation around topics organized in threads has provided answers to some of the most pressing questions from their communities.

In turn, it has allowed its members to build an identity online and grow a solid reputation.

Different badges of honor help identify the most respected members from the nOObs trying out the internet for the first time. They also come with a straightforward mechanism to identify the best answers or comments — upvotes and downvotes.

Twitter recently announced they are trying the feature for research purposes. Interesting, replied Reddit.

The simple head nod or headshake pioneered an era of gamification in user experience and was extremely beneficial at driving retention. Numerous psychological studies have shown the effect of dopamine release on the members’ brains while using social platforms. Hooked, they stay active in the community, participating in as much debate as possible, driving engagement up.

It was also very effective at guiding a reader towards the interesting conversations happening on the platform. For most platforms, the conversion of non-members is crucial to revenue. By keeping traffic and user engagement metrics up, they could sell advertisements much higher than their competitors.

Therefore, the rules governing the game have been pretty simple:

1. Do not mess up with the user’s attention.
2. Make something visually pleasant so that readers find the answers they were looking for in the shortest time span. If the experience is satisfying, they will come back, for sure.

The result? The enhanced UI we are all subconsciously familiar with; an indented comments section, sometimes placed at the top for more clarity. A stylized plus sign, rendered with a different color, usually green and bolded for the most praised comments, or a negative sign, always in red to shed some light on the most shameful and disgusting parts of human nature.

But how helpful is this for the reader to find what she came for in the first place?

In terms of user engagement, the success of the upvote/downvote speaks for itself. How many times have you gone down the rabbit hole of impassioned debates and never-ending threads tainted with hilarious replies?

More often than not, we only partially uncover what we wanted and leave the site thinking that the last half an hour could have been more productive.

On the way, we also often encounter some questionable sources of information, trolls, and even downright cases of harassment. And what do they all have in common? A high degree of engagement, reinforced by … upvotes or downvotes.

Fake news, plain and simple abuses and all the worst-case scenarios have triggered a welcome crisis in public opinion. Regulators and politicians expect more from social platforms to protect their communities. Lawful actions can certainly help to go in the right direction, but UX researchers have a unique opportunity to fix what seems to be broken by design.

Expecting more insights from a system built to assess everything on a one-dimensional scale is almost unfair.

I can find a comment funny but misleading or irrelevant to the conversation, but as a reader, what choice do I have? Conflicted between binary options, I may end up influenced by what the community thinks of it.

Psychological research has shown on multiple occasions the frightening impact of group dynamics on our own behavior. If people think that a borderline racist joke is funny, our brain is wired to mirror them.

And protected by the anonymity of the screen … you could very well upvote it, like 999 people before you.

To fight psychological bias, designers need to rethink the upvotes/downvotes system and let collective intelligence better qualify what characterizes good or bad answers.

Collective intelligence or the wisdom of the crowd is a powerful concept with tremendous potential when applied correctly. By gathering independent opinions from a wide set of individuals, we end up painting a much better version of the truth.

We could, for instance, find a way to break the one-dimensional view into three. The reader could be asked to define an answer by:

• Its relevance: is the answer on point or completely off-topic?
• Its precision: are the arguments used factual and validated by trustworthy sources? Or does it look like a bag of garbage and no one knows what it is, but it’s provocative, and gets the people going?
• Its clarity: Is it hard to read and understand? A feisty reader might be interested in something lighter or easier to grasp in 5 minutes.

How could we change the upvote or downvote system? Let the reader pick from a list of adjectives the main attributes defining an answer, for instance. And the little tags could then be rendered next to a comment and describe how relevant, precise, and clear the community thinks it is.

Therefore, a reader could know exactly the nature of what he is about to read and decide if he wants to engage down this lane or study the most precise answers from the thread.

The benefit could also serve the author.

I love receiving claps or likes. However, I would love to know that 70% of the reader base thought it was not concise enough. It would remind me to get to the point quicker in my next article.

Is it too much friction for the reader? Well, it depends on how it is designed.

Are we taking the risk to drive engagement down? Maybe; some A/B testing could help validate the benefit of the approach.

But the extreme nature of the content produced and constantly validated by the masses forces us to be honest with our priorities. We are left with the choice of a user experience, over-optimized, dull through, and sometimes dangerous to the most vulnerable or try to bring back meaning in how the internet spreads ideas and knowledge amongst communities.

I have made my choice.

Please share and leave feedback in the comment section — I’ll take the time to answer :)