Hey fellow Lemmings,

I’ve been thinking about how we measure the liveliness of our communities, and I believe we’re missing the mark with Monthly Active Users (MAU). Here’s why I think Posts + Comments per Month (PCM) would be a superior metric:

Why PCM is Better Than MAU

  1. Quality over Quantity: MAU counts lurkers equally with active participants. PCM focuses on actual engagement.

  2. Spam Resistance: Creating multiple accounts to inflate MAU is easy. Generating meaningful posts and comments is harder.

  3. True Reflection of Activity: A community with 1000 MAU but only 10 posts/comments is less vibrant than one with 100 MAU and 500 posts/comments.

  4. Encourages Participation: Displaying PCM could motivate users to contribute more actively.

  5. Easier to Track: No need for complex user tracking. Just count posts and comments.

Implementation Ideas

  • Show PCM in the community list alongside subscriber count
  • Display PCM in each community’s sidebar
  • Use PCM for sorting “hot” communities

What do you think? Should we petition the Lemmy devs to consider implementing this? Let’s discuss!

  • KoboldCoterie@pawb.social
    link
    fedilink
    English
    arrow-up
    6
    ·
    8 days ago

    Spam Resistance: Creating multiple accounts to inflate MAU is easy. Generating meaningful posts and comments is harder.

    Isn’t this actually just spam encouragement? A community with a bot that posts 50 low-value posts every day will have a much higher PCM as a result, and that behavior is more obnoxious to users and moderators who have to see it and deal with it, vs. someone creating a bunch of accounts, which is largely invisible to everyone else.

    • CoderSupreme@programming.devOP
      link
      fedilink
      arrow-up
      1
      arrow-down
      1
      ·
      edit-2
      8 days ago

      I appreciate your perspective, but my focus is on enhancing our measurement of community activity; if you have a more effective metric in mind, I’d love to hear it instead of just pointing out flaws.

      • KoboldCoterie@pawb.social
        link
        fedilink
        English
        arrow-up
        5
        ·
        8 days ago

        if you have a more effective metric in mind, I’d love to hear it instead of just pointing out flaws

        I mean, isn’t the whole point of this comment section to discuss the merits and flaws of the proposal you’ve made? If we’re not discussing the downsides, too, what’s even the point?

        That said, an ideal system would be a measure of the quality of content, not the quantity of content so, as another user has suggested, some measure involving net upvotes might be more effective. Yes, obviously a user can create multiple accounts to upvote everything and fuck with that metric, but I kind of doubt many folks would go to the trouble.

        Maybe some combination of PCM and the average number of votes divided by the number of active users could generate some sort of quality metric. At the very least it might be a measure of engagement.

        • CoderSupreme@programming.devOP
          link
          fedilink
          arrow-up
          1
          ·
          edit-2
          4 days ago

          I agree, but that might complicate things. Instead of votes we could also use time spent reading posts as the engagement metric.

          How about something like this?

          Quality Engagement Score (QES)

          QES = (PCM * AVU) / MAU, where:

          • PCM = Posts + Comments per Month
          • AVU = Average Votes per User (total monthly upvotes / MAU)
          • MAU = Monthly Active Users

          PCM measures raw activity, while AVU factors in community approval.

      • Otter@lemmy.ca
        link
        fedilink
        English
        arrow-up
        1
        ·
        edit-2
        8 days ago

        Creating a bunch of accounts possibly to manipulate votes

        Looking from the admin level, doesn’t happen that often. Vote manipulation is already something we keep an eye out for, and usually it’s done to highlight certain content (ex. pushing some political angle) rather than boosting one community over another.

        you can stop seeing by adjusting your “Show Bot Accounts” setting

        I like some bots, but I only subscribe to a bot-only community if the volume of posts is reasonable.

  • Otter@lemmy.ca
    link
    fedilink
    English
    arrow-up
    4
    ·
    edit-2
    8 days ago

    Alternatively, I think both metrics are helpful in different ways

    1. Quality over Quantity: MAU counts lurkers equally with active participants. PCM focuses on actual engagement.
    1. True Reflection of Activity: A community with 1000 MAU but only 10 posts/comments is less vibrant than one with 100 MAU and 500 posts/comments.

    I’d say votes are also an important part of engagement. It helps differentiate between good and bad content. I’m more likely to join a community with a few good posts a day (or even a week) than a bot community with many posts a day. Going by how the subscribers counts change over time, I’d say this is a common experience.

    1. Spam Resistance: Creating multiple accounts to inflate MAU is easy. Generating meaningful posts and comments is harder.

    While any abuse is bad, spam posts and comments are a bigger concern right now. AI generated spam / link spam is obnoxious and we deal with it often (as admins/mods). While someone could make lots of accounts to inflate MAU, it only really affects the community ranking against other communities and not day to day usage. Lemmy is already considering removing the trending section, and admins usually step in if a bunch of similar accounts are created at once.

    1. Easier to Track: No need for complex user tracking. Just count posts and comments.

    I’m not sure I understand this point. Are the vote/comment/post calculations very resource intensive?

    All that being said

    • I don’t see any downside to listing the info in the places you mentioned
    • If people want such a sorting option, then sure why not. Give people options
    • Otter@lemmy.ca
      link
      fedilink
      English
      arrow-up
      3
      ·
      edit-2
      8 days ago

      Taking a look at the current sidebar, it might be nice to reorganize the stats section completely

      What I’m thinking is:

      By default it will only show some stats, where users can select what stats they want displayed in the settings. This way I can hide the stuff I don’t care about, instead of having to look through the already busy list.

      **Statistics:**                       [✏️edit]
      
      - Monthly Active Users: 4,000
      - Total Subscribers: 30,000
      
      [ v see all v ]
      

      Then expanding the box will give the full list of stats:

      [ ^ collapse ^ ]
      
      
      **Statistics:**
      
      Active Users: 
      
      - By day: 800
      - By week: 1,200
      - By month: 4,000
      - By year: 24,100
      
      Subscribers:
      
      - Total: 30,000
      - Local: 12,000
      
      Comments: 
      - Total: 81,000
      - Today: 510
      - This week: 1,315
      
      [... etc]
      

      It opens up the possibility of including more items in that list. We could also replace the expand option with a link to a full statistics dashboard page.