How to seperate the "drive by" topics from the "deeper" discussions?

(Jan P.) #1

I moderate a technical community that is used by many as a support channel: “I have this problem, help me fix it”. This is totally fine, that is how it is supposed to be used. Unfortunately many of these “drive by” users and topic are of very low quality: Didn’t read the docs, didn’t read the error message he got, underlying problem is not related to the software at all or similar. That’S also fine and we are working on things to improve this.

Problem is: The few “deeper” or more advanced discussions get lost in the noise.

Right now the forum is grouped by “component”, so which part of the software a user has a problem with. Most of the topics get posted in the “main software” category, only some into one of the “child component” categories.

The initial idea was to create a “beginner”, “newbie” or “support” category (or whatever name is not too pejorative), but of course these users won’t think of their question as this and will still use the main category and this could also mean we create a “ghetto” of questions nobody bothers to answer.

The other plan was to create an “advanced” category, but this would detain well written and researched posts of new users as they won’t feel welcome in this category. Also some would of course try to abuse this and post their lower quality questions in there to get the “advanced” people to reply.

Any thoughts?

(Evgeny) #2

We created a category: sandbox. In which the rights to publish with the level of trust 0-1. In all other categories, we have similarly changed access rights. You can view all categories, but beginners can write only in one category. Topics from which are not visible on the central page of the site.

Or it can be done with the help of groups.

(Jan P.) #3

That is an interesting technical variation of the idea. Can these users only create topics and reply to posts in this category or also reply to all the others topics?

(Evgeny) #4

All users can answer in all categories. I think it’s not right to limit them in this. But they start writing topics, i.e. They can create the first posts in only one category.

It is possible to establish the rights to categories in different ways, and to achieve as the experience grows give the opportunity to start the topics in more and more categories.

The options for combinations of categories / user groups can be huge. The platform allows this.

There is a resource: Лучшие публикации за сутки / Хабрахабр
Where this idea is tested by time. There, all new users can post new messages only from the “category” that is not displayed on The central page of the site… There are no “stupid” publications, questions, etc.

(Jeff Atwood) #5

Yes, we’ve known about this since the genesis of Discourse – for example, @howtogeek was very adamant that his community not turn into a support wasteland where there are no interesting discussions, just people desperately seeking help. And he’s right!

I call this the “mole” problem. You’re throwing a party for doctors, and instead of having fun at the party and talking to each other, the doctors spend all their time answering questions from random people at the party like “hey doc, I got this weird mole on my arm, can you take a look at it?”

Would be interested to hear @hawk’s thoughts on this, and @erlend_sh have you written about this in the past at all?


The core of the issue is that you’re attempting to hybridise as support community with a community of practice – and they both have very different objectives and value propositions.

Support communities are (obviously) designed to provide support in the quickest way possible with the lowest barrier to entry. Ideally people do a quick search and then post their problem. They get an answer and they don’t bother coming back until they have another problem. These communities aren’t designed to be browsed through. The ROI here is clear – you’re supporting your customers.

CoPs are all about browsing and reading. Their objective is to collectively share knowledge so that everyone learns more. No one wants crappy questions and a higher level of autonomy is expected because the idea is to keep people coming back or the ecosystem dies. The ROI here is much more vague – usually social return or exposure to other products (consultancy, training resources etc).

So my question to you is why do you want the higher level/deeper discussions?
What value do you get from repeat visitors?

(Jeff Atwood) #7

I question this assumption; Stack Overflow is exactly that, and it seems to work. But do note that it is tightly focused on Q&A and has nothing to do with discussions… which is a common misconception about Stack Overflow.

Huge, huge difference between discussion and focused Q&A.

(Lowell Heddings) #8

We never solved it.

But lessons learned:

  • Putting support into a buried category turns it into a graveyard.
  • Since we launched so many years ago before the ability to hide categories from the homepage was added, our users just were flooded with tech support for the first 6 months or whatever, and no amount of trying to change the focus mattered. I eventually gave up trying to change the focus to tech discussion vs tech support.
  • If I could do it over I would have just embraced the tech support angle, realized that my forum would never be fun at parties, and eventually made an optional category for more interesting discussions for the more dedicated users.

(Jeff Atwood) #9

Each howtogeek article has an associated Discourse discussion topic … so one thing I personally think you could have pursued much more aggressively is requiring article authors to engage in the comments on their articles. The rule would be something like “you must reply to at least two comments (of your choice) on each of your articles”, and since comments auto-close after a few days, there’s no following up weeks later anyways.

I’ve noted that Gizmodo and other places do this, and it makes discussions a whole heck of a lot better!

(Of course, this advice is of zero use on sites that don’t publish articles and comment topics alongside their normal discussion content.)

(Lowell Heddings) #10

Sure. But since most of our articles are how to do things, most of the comments are about how to do things. It’s all tech support which I should have simply embraced.

Lesson learned. Which is the only reason I’m sharing this now.

If support is the primary focus of your forum, embrace that fact and maybe put the outlier advanced topics into a category if necessary.

(Jeff Atwood) #11

Point taken, but surely not all the comments on an article are of the form

Halp me plzzz omg!

and some are

Why did you choose (x) instead of (y)? What happens if we (z)?

And you could selectively choose to reply to the latter comments, maybe even delete the former…


Totally agree that a hybrid can work in some circumstances but that’s rare.

Also – SO isn’t a product support community aimed at reducing support costs, right?
Its revenue model relies on eyes on pages.

So much this. I fought that same fight at SitePoint and continue to do so at UXMastery. I think it’s key to survival of that particular model.

(Angus McLeod) #13

I’ve been thinking about this recently. In fact I was just reading this thread on a related issue with SO.

See in particular the answer from Shog9.

The thrust of the thread is a little different - and is generally a bit more nebulous / hand wavy - however the subject matter is similar: i.e. the rise of poor questions on SO and maintaining the ‘community’.

One of the thoughts that sticks with me is something that @HAWK touched on. It’s important to distinguish between what the goal of your website is and some vague sense of creating a ‘good’ community.

Another way of expressing this is “It depends”. More specifically to @Sujan’s case, I guess the first question that arises is: What do you mean by “deeper … discussions get lost in the noise”? What qualifies as deep? Lost by who? Is it that they are getting lost or not enough ‘deeper’ discussions are getting posted in the first place?

Another theme here is something Shirky touches on in A Group Is Its Own Worst Enemy:

Now you could ask whether or not the founders’ inability to defend themselves from this onslaught, from being overrun, was a technical or a social problem. Did the software not allow the problem to be solved? Or was it the social configuration of the group that founded it, where they simply couldn’t stomach the idea of adding censorship to protect their system. But in a way, it doesn’t matter, because technical and social issues are deeply intertwined. There’s no way to completely separate them.

Applying this to @Sujan’s case, the question arises: Is there a way (existing or imagined) to preference “deeper” discussions (whatever that means) in the discovery topic lists, either in the UI or the algos, so they naturally get more visibility?

(Erlend Sogge Heggen) #14

I’d say community culture & norms is the most common approach here. While on BoingBoing you can rack up crazy high Like scores with a killer joke or political burn, here on Meta the more sure-fire ways of getting your content Liked (and consequently appearing in Top) is to:

  • Help someone (the more the better)
  • Fix something
  • Create something
  • etc.

…because we the staffers have been actively and publicly applauding this behaviour. The algorithms are already in place; the rest is good old fashioned community management.

One fairly low-hanging fruit we could explore here though would be to have a setting for “send me an Activity Summary even when I’m logged in”, which in effect would be more like a “cool stuff you might have missed” email. That would give the most popular topics an additional boost.

(Angus McLeod) #15

Right. There are proxies for deep / interesting already, the primary ones being “Likes”, “Views” and “Replies”. Discourse uses these proxies to highlight ‘interesting’ posts via specific topic lists - “Top” (core) and “Hot” (plugin) - and via the design of the topic list itself (separate column for likes; heat colors);

One way of thinking about this is: What other proxies for deep / interesting are there? Here are some possible candidates:

  • Reading time (this is particularly underutilized I think)
  • Writing time (interesting posts normally take a little longer to write)
  • Number of links
  • Frequency of posts quoting other posts

This data already exists in Discourse, it’s a matter of using it in an interesting way.

It is perhaps worth mentioning in this context the potential for machine learning to assist as well; if only to plug my existing (unfinished) machine learning plugin, which I would be happy to work with somebody on.

Another way of thinking about this, and this gets back to the OP, is “How do I handle different types of users, with different goals, on the same platform?”.

Say, for the sake of argument, we divide the ‘user experience’ management of Discourse into:

  • Defaults: Site settings. In particular top_menu and category page style.
  • Trust: The trust system.
  • Taxonomy: Categorization / tagging / grouping.
  • Visibility: This discovery list design and user control over notifications, emails and theme.
  • Curation: mods & staff

In addition to improving each system independently, some interesting avenues arise when you think about the connections between them, e.g.

  • Taxonomy + Trust + Visibility: e.g. A topic list that ranked based taxonomic variables of the topic and trust variables of the posters, in addition to likes. This could split the difference between hiding the ‘sandbox’ / support category and letting it pollute your homepage.
  • Trust + Defaults + Visibility: e.g. Give trusted users greater control over the Discovery UI to set their personal preferences; or
  • Defaults + Visibility: Show different default topic lists to different users based on things like how new they are or how long since their last visit. It seems that this approach is already used (I can’t find it in the current codebase though).

(Dave McClure) #16

:+1: This is an area of general interest that I have as well:


I keep coming back to this because it is in essence what we’re planning to do here.

Most of Meta is based on a support model and the ROI is reducing support costs, however our plans to build out this #community category are focused on a very different goal – product retention. We will essentially have a CoP living within the wider ecosystem of a support community.

And perhaps that is the key. Create space for the deeper conversations to happen and nurture the kind of behaviour you want within that space. Sub communities work fine in other places – think about feedback type ecosystems where there are rules around reciprocation (i.e. you have to do x reviews before requesting your own) that don’t exist in the wider space.

Add to that @angus’s vision of a machine learning driven UX and I think it could work very successfully.

(Angus McLeod) #18

A threshold question of sorts, is the distinction between ‘forums’ and ‘social media’. I feel like that question needs to be at least acknowledged before tackling the nitty gritty of ‘Different Users / Same Platform’. Partly because some people dislike social media and like forums, so they naturally resist attempts to employ what are perceived as social-media-like approaches for a forum.

The distinction is historical. Our idea of a ‘forum’ is rooted in a time when it was not possible to curate the experience of a product to individual users. In this narrative, Discourse is the latest iteration of ‘forum’ software improvements going back decades. In this narrative, social media exists in a different track. In this narrative, targeting content on your platform to users based on user preference is part of the the ‘social media’ track. In this narrative, while social media targets content to users to drive engagement which drives advertising revenue, forums give the user more independence and respect and are not concerned with the profit motive. In this narrative, forums are ‘communities’ and social media is narcissistic / solipsistic. In this narrative, forums are egalitarian and/or libertarian, while social media is driven by consumerism and celebrity. This narrative is reflected somewhat in Discourse the product and here on meta.

From a ‘product’ perspective, I think a better way of thinking about the forum / social media distinction is simply that forums are topic-centric and social media is relationship-centric. That’s it. If you just focus on this distinction and put the narratives we have about forums v social media to one side (for the purposes of the product discussion at least), it clarifies the product question of Different Users / Same Platform. It clarifies it, because it allows us to think of the curation of content employed so successfully by social media simply as a product strategy, rather than a practice intertwined with social and ideological narratives.

How does this help us? Well, I think it allows us to more squarely face the fact that the ‘Different Users / Same Platform’ problem has already been solved. Twitter, Facebook etc have a vast array of different users with different goals. They tackle this by heavily targeting content to users based on perceived user interest. Imagine if the default user homepage for Twitter or Facebook was a ‘category’ list, rather than the user’s feed. Or a list of the latest content across the entire platform.

It doesn’t follow from this that Discourse should adopt concepts like ‘following’, ‘friending’ or other social media features. Those are just proxies that social media uses for user interest.

It does follow that targeting content to users based on their individual interests is a product strategy worthy of consideration for Discourse. The proxies for ‘user interest’ on a forum are different from social media, but they still exist.

postscript. @HAWK, @mcwumbly, @erlend_sh there’s a theme in your approaches (it’s true of my instincts as well, but I’m trying to question my / our instincts). When we think about ‘Product’ for a forum we instinctively think about ‘manual’ control. Manual control / curation via moderation, manual control of the topic list by individual users (@mcwumbly’s interesting spec). This contrasts somewhat with the ‘automated’ approach of social media. This may be partly because doing ‘automation’ well is harder, but I think it’s deeper than that.


Your historical narrative is interesting. I agree that those are widely held perceptions, which I think is unfortunate. I hope we are maturing to a point where we aren’t so disparaging of social media.

Ah – so here I think we’re back to the original conundrum. I believe that ‘product/support/branded/B2C’ communities are very much topic-centric but ‘CoP/B2C’ communities rely on being relationship-centric for their survival.

Both audiences might be drawn to a place because of the topic, but the former will leave and the latter will stay for the connections. So how to you combine those two behaviour models in one environment?

I don’t think this is about types of content… it’s about types of behaviour. Your model and mine are the same (I think) but you are trying to solve the ‘how do you get people into the part of the forum where the behaviour they want to experience is happening?’.

As an aside, but one that I think is important to raise –

I think it’s been a long time since this was the case. CM has matured a lot in the last 3 or 4 years to a point where we’ve realised that we very much need to be concerned with the profit motive if a community is going to have longevity. Most of the long term communities that still exist are either monetized or ROI in another way. I think it’s naive (as a user) to expect something for nothing – most often the nothing just isn’t obvious.

(Christoph) #20

As for “other social media features” I’d like to also mention the ability of social media to integrate an individuals interest in a set of different topics as well as different communities about the same topic. I think there is potential here in finding ways of linking different communities together