Is "liking" a post too intimidating?

On one of the Discourse forums I participate in, someone posted a feature request. Three posts below the original one, someone posted a poll with the choices “Yes add this feature” and “no don’t add this feature”. The original post was only liked 18 times, but the poll was voted on 60 times and had 57 votes for “yes”. You’d think everyone who wanted the feature would have liked the first post in the thread, but only 18 out of the 57 people who wanted the feature liked the original post. Why is this?

I’m extremely active in this forum and read almost every post, so I can say for absolute certain that nobody is hitting the daily like limit. Did the poll get more votes because it was anonymous as opposed to publicly displaying that you supported the addition of a particular feature? The feature wasn’t anything embarrassing that you wouldn’t want people to know you liked – it was just about adding Trello links to the list of allowed links on the parent site. Is it because those 39 people reserve liking posts for ones that are extremely noteworthy? In that case, does there need to be another way to vote on feature requests in place of liking the original post in the topic (not necessarily a Discourse feature, but a plugin)? Features aren’t decided on based on votes, so it doesn’t seem appropriate to add a poll to each one. But how many people want the feature does have an impact on its priority, so it is important that how many people want a feature is known.

Any thoughts on why the poll had more votes than the OP had likes? Thoughts on how the number of people who want a feature is gauged? Any similar experiences you’ve encountered previously?

The forum the feature request is on is private, but the Discourse staff have an account there, so here’s a link to the topic mentioned in this post for them.


I’m curious. How many of the poll voters that did not Like the OP have used Like before on other posts?

i.e. Might it be that the Like feature is less obvious than a poll and these members don’t know it exists?


Possibly. Our community is pretty small (every member hand-picked from the parent site to join the forums) though, so 39 is a larger amount of people than it’d normally be considered – we get less than 10 new members a month, so there’s a good chance most of the people who voted have been around a while. I find it hard to believe 39 people didn’t know the like button existed. Maybe a couple newbies didn’t like the OP because of that, but I don’t think it’s the reason for the majority of the 39.

1 Like

Hey. I think… It is because the latest post under that topic gets auto scrolled to the view and more people see it?

And also poll interface is attractive with large yes, no … which is visually big than small heart button…

Also, in polls - there are just 2 options - so human mentality to choose one among the 2 is easier than simply liking something…


I’m thinking there may some kind of psychological difference.

Liking a post is kind of a one-to-one thing.
But voting in a poll is more like group participation.


Also can you share how many views did that topic get? So we can estimate likelyhood of someone liking a post against voting in a poll?

1 Like

The topic has 164 views and 29 replies (the last one being half an hour ago). It was posted a day ago, if that makes a difference.

Yeah, that was one thing that crossed my mind when drafting this topic which is why I titled it what I did.

It scrolls down to the latest unread post. If there are 1,000 posts but you haven’t read a single one, you start at the beginning.

1 Like

You are correct.

Two levels exist as to how a Human relates to everyone else.

As an individual, and as a member of a group or tribe.

Cognitive dissonance is the psychological mechanism showing this “low likes” and “high votes” behavior described here.

It is also the reason someone may suddenly fall onto a sidewalk pavement in cardiac arrest, many people around during rush hour walking-traffic, and not a single person will go, race to this person and offer assistance.

But, if this person dropped onto the same sidewalk during the nearly-empty lazy-afternoon hours with only a few individuals to witness, a much greater chance be someone of those few will drop everything and run to help.

In a group-oriented species, no one wants to be the odd one out. No one is willing to “be themselves” when it also means they no longer have a group to witness it, because they just gave that unique snowflake the FAIL hail, sometimes involving a lot of image memes (depending on said group).

To exert individuality so openly means one risks negative feedback from the group, and expulsion. That risk, right there, is a manifestation of cognitive dissonance.

“This…not “feel” right to me, to hit this like button, and no one else doing it either. Oh, a poll; that does not mention any sort of strong individualistic word-choices such as “like”. *casts vote* Did my duty there, using the vague mask of “general community citizen” to cast my very individualized opinion. That feel goood to exert my individuality but feeling safe at the same time. Dissonance itch scratched.”


I’ve been participating in online communities for over 30 years (Pre-Internet in other words) and I’m surprised ‘we’ haven’t yet managed to evolve beyond the like button. It’s a very blunt tool and in cynical (undignified?) hands can be used to express negative sentiment. Something I rather clumsily refer to as negative likes.

Negative likes are given in a number of ways. The most obvious is when a comment attacks another member/comment and, in turn, attracts one or more likes from folks that share that negative sentiment. Some might say this is democracy at play but I see it differently and will be instructing my mods, when I go live, to actively dissuade this type of behaviour.

The second less obvious, but arguably more destructive use of negative liking happens when an original post or comment is effectively ignored, only to see a follow-up comment within the same thread, often rehashing the topic itself, attracting a disproportionate number of likes. I see this a lot and it’s a way of saying I have issues with the OP, but I agree with the sentiment of the topic itself.

Both behaviours mentioned above are rife in quite a large car forum I use and the moderators are either blind to it, or tacitly promote it, given they seek traffic to feed their Adsense revenues, not dignified discussion as we see here.

In saying that @Drew_Warwick, I am not suggesting that either of the above are behind the behaviour you describe given the poll might have attracted more likes simply because it was recognised as being a more a more appropriate device for gathering feedback in that scenario.

Either way, I would like to see Discourse look at other ways of promoting dignified rapport. Yes, we should be allowed to express sentiment. But a simple like button, on an otherwise mostly brilliantly executed platform, falls short of what’s needed. In my Sunday morning opinion, of course.

Keep it up team. Great job.


That is Human Nature. Give a Human a tool, and their ability to think independently outside its pre-established context defined by the group will show in time. That is what separates Man from Ape.

I do not say that to throw out your concerns. They are valid; they become tiny grains of dust and sand flung at mach-speeds onto a windshield’s front, breaking integrity over time, little by little until only another single tiny pebble shatters the entire piece.

I suggest looking at how likes mitigate this “nature’s subconscious deviance” when compared against other tools and techniques employed in other modern forum software available; tools to express approval and build rapport.

As to the car forum mentioned, those moderators should have those behaviors pointed out to them.

Sometimes, all it takes is a little bit of open, honest communication with users, to steer them from unwanted behaviors, now that said behaviors are so prominent that others have become cognitively aware and can now define them, as you did above well.

I said this in another topic on Meta; same concept.

Example: if user A opens a topic, is ignored; user B replies with same topic, gets attention. I would be the one to link to the OP as a reply and gently steer it, meld it into the current discussion so that user A now has an opening to reply and join the topic fully with their words. That I call a “gentle steer” drawn with utmost positive gain, sans negativity as much as possible to keep things smooth. I do it here on Meta when answering support topics, in many various ways. (And upon further reflection, my steer will give others that “oh damn; we got caught not being nice and he just hinted to it” feeling. That refers to my first post in this topic: being the odd one out of the group dynamic that is expected to be harmonious, and not silently berating.)

1 Like

My thoughts on a very active Discourse forum I participate:

  • People use “like” a lot. It seems it’s used as an “acknowledge” button some times. Usually when it’s something funny.
  • I heard there is much more lurkers than posters on the internet, but lurkers don’t like much stuff on said forum.

One hypothesis is that many lurkers would vote on your poll, but won’t click on like because their name will would appear on the like list, and the vote is anonymous.


It’s an interesting UX story. My personal theory:

Likes are really personal, so when someone Likes a feature request, they feel like they’re not only expressing their feelings for that feature request, they’re expressing their feelings for that person. It doesn’t have to be about genuinely liking or disliking another person, it’s just that not everyone wants to explicitly “favour” anyone. Likes is also a form of social currency, since it affects Top list rankings, Trust Levels and Bagdes.

This is why I theorised a “Likes Archetype” that could be detached from regular Likes and be made less personal. When you’re strictly voting on a thing and not also on a person, it’s suddenly not a loaded decision any longer.

Reading your topic I also saw that your forum’s old equivalent of “Likes” was "Thank You"s, which will certainly leave an imprint in older user’s minds.

I’d rather say that Likes are largely unappreciated and underestimated as a communication tool. It’s a style of communication that needs to be lifted up by healthy norms just as much as the general discourse on a forum. I believe Meta is quite an excellent example of Likes Done Right. From very early on the co-founders themselves Liked a large amount of posts, indiscriminately yet righteously. Likes are not a rare commodity; it’s a non-zero-sum social currency. Likes begets more Likes begets more Likes.

(p.s., cool feature of Likes: In my reply to you I’m expressing some disagreement, but I still Liked your post, because I think you added value to this discussion. Granted, a user more personally & competitively invested in a discussion might not use Likes that way, which brings us back to healthy norms and setting good examples.)


As I understand, the original intent of the Discourse developers with the like button is to remove all the “+ 1”, “me too”, “this!” and “ok” posts.

They have a wide range of meanings, but they do their job.


I like (haha) “thank you” as it focuses more on an action than a person. Similar to your suggested feature.

And with a heart icon, that also speaks more about deeper emotions. A hand making a thumbs up speaks of a gesture you give a coworker, a team you work with to forward a group’s goal. A heart you give to a bestie or more; something more personal for individuals to share privately. This may be Facebook’s line of logic to have likes represented with a thumb-up.

Changing the heart to the thumbs up may be the only change needed to turn this behavior around. As the universe dictates that the most easiest solution is usually the most appropriate and most useful (much to my chagrin sometimes…).

This whole thing reminds me of a very specific logical fallacy.


I’m not sure we understanding differentiation fully at any level, probably least of all between ape and man - although I offer this input having followed both Dawkins and Fukuyama and consequently sit somewhere between knowledge and confusion. Not wishing to deviate off topic, perhaps a discussion for another time?

Yes, it has been.

That’s all very well, but isn’t the aim to develop a mechanism that guides the user’s behaviour rather than relying on human intervention?

Indeed, but the same functionality running within a different, perhaps more cynical environment, may not fair so well. The car forum I quoted being an example. It runs VB, but the like functionality is (almost?) identical.

Yes, I saw and it was appreciated. But again, that’s an example of a multi-use tool being used well in the hands of a skilled, uncynical, operator.

1 Like

A couple observations; I find that who likes the post matters to me more than how many likes there are.

Have you ever found yourself more inclined to like a post that’s already got a bunch of likes, especially from your friends? I know I have.

I see some users that like almost every post. It annoys me just a little bit, because I can never tell if their likes mean anything. “Oh, I really like your post! And that one, and that one, and that one…”


Absolutely true. Some users like all the replies in topics that have opened themselves…even if a user answers A and another responds Z. Very annoying because it loses its usefulness.


Yes, this is a huge factor. Likes are akin to co-signing someone’s post. It is a public statement of support.

Edit: if changing the glyph is what you care about (from heart to brofist to whatever) that is a few lines of CSS.


Sure. I be game for that. I like to learn and listen. Despite how I may come across. :grin:

Ahh. I think there be a differentiating definition here, unspoken between us.

Let me attempt to explain mine.

It sounds like your experience with Discourse’s likes has failed, devastating kind of fail. Say, if you were to quantify this as a percentage on average “work rate” that satisfies you, with 0% working not at all, at all and 100% would mean you would not have entered this topic with your concerns about Discourse’s likes to begin with. Where does it fall for you right now? Based on your observations? What Discourse-based communities have you seen, that as a sum, would equal this average percentage I ask to know from your POV. Does not have to be accurate. This is all based on “feels”.

Form what I have seen, my brutally honest percentage, based on all Discourse-based communities I have taken at least more than ten minutes to study, would be 80-85%. Up to 90% if the community is very cordial, and the user-base vetted in some fashion. Maybe 70% if the community in question has a reputation for a lot of “individuals” and not so much “team players” (Geek/Gamer/Fandom culture comes to mind for this as example).

I think the likes are great. That 20 or so percent, I can work with. That is Human Nature no one can truly control.

1 Like

There have been some rather interesting thoughts in this discussion, but I wonder if the bit I quoted above is significant. If:

  • The first couple of posts are fairly short AND
  • The Poll got added very early in the discussion

Is it possible that the majority of the users only saw the topic after the poll was added? In that case, they may have been responding to the topic as a whole, and then choosing the more “official” looking response…

Just something that floated through my mind… YMMV.

1 Like