Introducing Discourse Discover

Big picture: Seems like an excellent idea and I’ve opted my public sites in. If nothing else, having another way to find potential community members is helpful.


  1. Do you have an idea of what criteria might cause CDCK, Inc. to include/exclude a site? Is the goal to maximize the number of sites included or build a curated list similar to the customer list?
  2. Potentially related to #1, do you plan to use nofollow links?
  3. Is there a possibility of reciprocal sharing? For instance it could be interesting to know what the median DAU/MAU rate is for all sites in the sample that have more than X topics. If we are giving up some of our data, it seems reasonable to get data about the universe of Discourse sites in return.
  4. Would there be some sort of ranking or categorizing system?
  5. Is there any plan to showcase interesting topics within a site?

General observations on . . . this sort of thing

I recently submitted my site to Product Hunt. It’s not an ideal fit but I thought it would be worth a try. From what I can tell, all my engagement came from what I classify as hustlers. I got some reports of website vulnerabilities from very helpful people willing to help me fix them (for a fee) and some offers of more upvotes on Product Hunt to help my project get more visibility. I did get a lot of anonymous visitors and web crawler traffic, but the whole thing seems like a waste of time.[1]

Over on Stack Exchange there’s a page for all the communities on the network. When I was working there I used it a lot as a community manager, but I don’t suspect it was very useful for people just looking for a place to find out more about, say, 3D printing. I suspect it was exclusively used by people who were interested in Stack Exchange first and something in the list caught their eye.

The bit that really worked was the hot network questions list that’s the default view of the page. You might not care about a site called “Retro Computing”, but really be curious about What did corporations use for long-distance networks in the 1980s? It’s the same sort of thing as Hacker News or reddit’s home page. Nobody starts by being interested in a site or community. Everyone starts getting interested in specific content first and only later in the group of people that create that content.

It should be noted that the connecting thread for this feature is the software that is being used to create the content. That’s not an ideal hook for the general public.[2] For people like me who want to understand how Discourse is used, this could be a great resource. If I could find out how other sites use Form Templates it could give me some models to copy. But there’d need to be some way to find the sites that are most effectively using them. Not an easy problem to solve without some sort of curation.

  1. If I could have done the same thing by checking a box in my site settings it would be a lot smaller investment, but still pointless. ↩︎

  2. Again, content is king. People like reddit because of all the cool things people post there, not because it happens to be on reddit. ↩︎