Discussion Forum Access Methods

In the context of deciding how best to evolve/change the platform underlying Mozilla’s current discussion forums, the question arose as to what percentage of people use each of the 3 available access methods (email, NNTP, Google Groups). I did some research on this a while ago, although I can’t find it now, and fortunately was able to find my script again and re-run it. (Tip: if you need to download a newsgroup as an mbox file on Unix, use the nntp-pull program from the sinntp package.) Here are some statistics, from what I hope is a representative cross-section of groups which are currently active. The stats are for the (approx.) 1000 most recent messages in the group, or all of them if there are fewer than 1000.

mozilla.community.philippines
            Total   Web           Mail           News         
   Posters:    41    18 (43.90%)    20 (48.78%)     3 (7.32%) 
     Posts:   135    34 (25.19%)    96 (71.11%)     5 (3.70%) 

mozilla.dev.planning
            Total   Web           Mail           News         
   Posters:   222    31 (13.96%)   128 (57.66%)    63 (28.38%) 
     Posts:  1001    72 (7.19%)    688 (68.73%)   241 (24.08%) 

mozilla.dev.platform
            Total   Web           Mail           News         
   Posters:   215    51 (23.72%)   107 (49.77%)    57 (26.51%) 
     Posts:   999   147 (14.71%)   590 (59.06%)   262 (26.23%) 

mozilla.dev.quality
            Total   Web           Mail           News         
   Posters:   207    79 (38.16%)    73 (35.27%)    55 (26.57%) 
     Posts:   871   144 (16.53%)   475 (54.54%)   252 (28.93%) 

mozilla.dev.tech.layout
            Total   Web           Mail           News         
   Posters:   172    75 (43.60%)    45 (26.16%)    52 (30.23%) 
     Posts:   860   250 (29.07%)   193 (22.44%)   417 (48.49%) 

mozilla.mozillians
            Total   Web           Mail           News         
   Posters:   171    32 (18.71%)   120 (70.18%)    19 (11.11%) 
     Posts:   887   187 (21.08%)   513 (57.84%)   187 (21.08%) 

TOTAL of all the above:
            Total   Web           Mail           News         
   Posters:   853   274 (32.12%)   404 (47.36%)   175 (20.52%) 
     Posts:  4753   835 (17.57%)  2557 (53.80%)  1361 (28.63%)

Notes and observations:

  • Each access method is used by a significant number of contributors.
  • Mail is the most popular access method. The order of the other two depends on whether you count posts or posters.
  • People using news have a higher rate of posts per poster, suggesting a deeper level of involvement. (That, of course, says nothing about the utility of each post.)
  • People using the web have a lower rate of posts per poster, suggesting more casual involvement (which makes sense, as the web version has a low barrier to entry).

5 thoughts on “Discussion Forum Access Methods

  1. The web results could easily be negatively skewed by the problems we’ve had with Google Groups the last several months, too. Some people who would otherwise use the web interface might be intentionally be avoiding it because they know it’s been broken a lot.

    • Possibly; although the biggest problem with GG has been connecting up new groups, rather than the working of existing ones. There was a problem with that for a bit, but it was a while ago.

  2. > the web version has a low barrier to entry

    This is an important point, IMO. The web interface has a low barrier to entry but is correspondingly unpleasant to use extensively.

    The nntp interface has the highest barrier to entry but allows the use of a real newsreader, which provides a number of very useful and otherwise unavailable features for managing individual messages, threads, groups, and even entire sets of groups. Just for example, the ability to kiboze is not available any other way.

    So I would say that the web and nntp interfaces, provided they can be kept reasonably in sync, complement one another very nicely.

    The mail interface is somewhere in between, depending on exactly how you use it. If you use it with a webmail account, it’s mostly similar to the web interface, with just a few extra features (depending on exactly which webmail service you use). With a good mailreader, it becomes much more feature-rich.

  3. > Just for example, the ability to kiboze is not available any other way.

    I just realized this is technically untrue. With the right software, it would be possible to kiboze mail groups. However, the barrier to entry would be much higher than for just using standard off-the-shelf solutions for kibozing usenet.