Statistics and Stuff

Over on SFN, we’ve been interested in figuring out how to stimulate growth and post volume to allow SFN to expand. It’s an interesting challenge — we have four large competitors (that I know of), and there is no simple “how to get more posts” procedure that we can follow. It’s a seat-of-the-pants venture.

To get a better handle on what we’re dealing with, I’ve compiled a graph and some handy statistics.

Weekly Posts on SFN
(Click for the full version.)

That’s the number of posts per week on SFN since the beginning of SFN time (July 2002). You can see the sudden drop-off toward the end of 2006 after a peak in May 2005. We began recovering in 2007 and have maintained a (slow) growth since. My goal is to reach 2005 posting levels at least.

To aid us, I wrote a script that does some analysis on a per-forum basis and figures out which forums on SFN are the most popular, by posting volume. You can see the results (automatically updated daily) here. (If you’re interested, you can also see a breakdown of your posts by forum as well.)

We’re still trying to work out how to achieve the growth we want. It’ll be an interesting challenge. If you have any ideas, feel free to voice them.

14 Comments

  1. It would be interesting to see how some of the currently available statistics vary over time.

    For example: could you graph posts per week only in the physics section? And plot that against other sections? And rather than sections could it be done with users (although I see less point in this)?

    This doesn’t aid our growth (obviously, well actually, we might get some crazy statisticians!) so it’s purely on an interest level. Although if a clear pattern showed one section continually decreasing in popularity we could guess as to why that was.

    Also the stats might be a bit iffy, as the structure of the forums have varied slightly over the past few years, but I guess if we (and by “we” I mean you!) just give it a go and see what happens, some interesting pattern might be found!

  2. I’ve often heard newbs talk about how intimidated they are since there are SO many people who know SO much. Perhaps create a “shallow end” of the site while people teach themselves to “swim.” Then, they can venture into the “deep end” with more confidence. How you might do this, you ask? Yeah… I’m not sure, so I’ll just move on now. šŸ™‚

    I think on a broader level, the issue you have is that veteran members tend to be more “reactive” posters, addressing questions raised by others, as opposed to “proactive” posters, creating their own threads and stimulating new discussion. So, we rely on newbs to bring up a topic, which then gets tossed about and answered, but then goes dormant. If we could find a way to have veteran members stir up new discussions, this would greatly help.

    The difficulty with this is that veteran members tend to have already read and posted on a vast array of topics, so struggle tremendously to think of somthing new… something not yet done.

    Finally, finding ways to increase your google strength is always good, but I don’t know much about that either.

    If you truly want to increase post count, you need something that gets people agitated and posting hurriedly… my first thought being something like a religion forum. Nothing gets threads rockin’ and rollin’ like the hot button issues of religion… just look at the post count they see over on the Dawkins site… It’s boggling (and, admittedly, part of the reason I don’t post there really anymore… too hard to keep up).

    Can you comment at all on how the post count changed when you archived P&R? I wouldn’t be surprised if there is a correlation between the deactivation of such discussions and a drop in posts, but I may be wrong.

    All the best, and good luck. The membership on this site is very strong, and we’re all fortunate to be able to participate in such a capable and diverse community. Cheers.

  3. Well, if I’m reading the two graphs correctly, your big drop at the end of ’06 precisely coincided with the shutting down of P&R.

    You’ll notice also that the spikes tend to occur at the start of new school semesters (Jan/Feb and Aug/Sept), but then trail during the semester. Your max in July ’05 is interesting, and appears most closely tied to a combination of physics, chemistry, and P&R. I’d guess there were some big discoveries that summer that people were trying to better understand…

  4. P&R at the time of closing only accounted for 15% of all posts, so it can’t account for the entire ’06 drop. Physics had already dropped to its low when P&R was closed, so it doesn’t look like P&R’s closure drove slowdowns elsewhere.

    I think you’re right about the reliance on new users to post interesting new topics. I think part of what needs to be done is encourage more *discussion* — not arguing — about various topics.

    Alternately, we can attract more new users. Initiatives like tutorials (I’m partway through a calculus tutorial, for example) can help us attract more new users. Blogs will draw traffic to SFN, although we need a good way of bringing people from the blogs to the forums.

  5. P&R = philosophy and religion. It was shut down because it was slightly non-scientific and got a bit out of control with long winded and often repetitive arguments. Still it was a popular section. Wasn’t a similar forum made just for P&R? I seem to recall a kind of half-sister forum which was for P&R? Maybe if we considered opening a philosophy section, but not religion?

    I also agree with iNow’s comment on how “veteran” members don’t create enough discussion topics, and there’s a minor sense of: question > answer > end of thread.

    Cap’n: you do remember Dave’s calculus tutorials:
    http://www.scienceforums.net/forum/showthread.php?t=4624
    http://www.scienceforums.net/forum/showthread.php?t=4706
    http://www.scienceforums.net/forum/showthread.php?p=145977
    (I still had them in my Favourites!!)
    although I suppose even if you just repeated those they would be popular, as Dave’s were posted a long time ago, so go for it.

  6. Yes, I found the Mathematics Tutorial forum a little after I made my previous comment.

    I don’t recall any tutorials, other than Dave’s, in the past (although I do recall talks of them), but maybe Dave’s threads (and any others that do exist and I don’t remember, or were made whilst I was inactive, or maybe even before I joined!) could/should be moved in to that section. It would give it a nice little starting kick-off, IMHO.

  7. I’ve been considering the opening question. If your primary concern is post count, then you can do something simple like start a bunch of threads that are for the sheer purpose of post padding.

    On other fora, I’ve seen things like “Great Words,” where literally every post is a single word that the poster just happened to think of and decide to share.

    Or, 3 word story telling, where each poster posts only 3 words, then the next poster continues the story with their own 3 words… It’s sort of like mad libs.
    Poster 1: In the beginning
    Poster 2: there was a
    Poster 3: giant kitten farm
    Poster 4: where aliens injected
    Poster 5: jesus juice using…

    It drives up posts quickly. You can also have threads like “Words that start with the letter Q.” It doesn’t matter, really.

    The challenge, which I believe is obvious by this point, is that you sacrifice quality for quantity, and it’s often counter productive to your desired goal.

    Last, I think you might increase your google strength if you put a link to the forums on EVERY blog page hosted here. This would also encourage blog readers to enter the fora.

    I’d also be curious to see how many visits the site gets. Often, people will rush to the site to read posts, but not post themselves. It may be useful to consider those numbers as opposed to simply looking at post counts. For example, the GUTs will be more “read” than “responded,” so it’d be advantagous to measure “views” in the fora.

  8. I wouldn’t be satisfied if we increased posting volume by increasing the activity in General Discussion. Volume needs to increase across the board.

    I’m trying to work out a good way to make each blog link to SFN. There’s got to be a better way to do it than to edit each and every theme used by blogs… I’ll find out.

    I can keep track of the number of “hits” we get daily. If we count every hit the server gets (image requests, CSS, and hits from search engines and so on), we get 2 per second, for a total of around 170,000 hits. As for page views, I’d have to check.

  9. ” As for page views, Iā€™d have to check.”

    It shows up on the searchdaily page, and also on the main fora pages, so it’s there somewhere for tracking. Granted, you probably already know this, considering all of the cool stuff you’ve done as a site admin, it wouldn’t be beyond the realm of possibility for your knowledge of the system to FAR exceed mine. šŸ˜‰

    FYI – It looks like Hypo just opened up a blog too. They’ve also implemented a FaceBook type system, which has certainly caused user activity to spike this week. It will be interesting to see if that activity remains consistent, or if it’s a temporary surge due to novelty. Right now, the content of SFN is definitely of higher quality, as are many of it’s members, but Hypo seems to be focussing on user experience more. While that’s helpful, it definitely draws a different group, many of whom would put the scientific method behind the social experience, so it truly is a trade-off.

    Cheers.

  10. Google Analytics shows us at around 7,000 pageviews/day.

    It looks like upgrading to vBulletin 3.7 (only a release candidate now) would give us the same features Hypography is offering, and I intend to look at the offerings to see what they might give us. Perhaps a staff discussion is in order…

    So I think my to-do list would be:

    1. Blog links back to SFN
    2. GUTs
    3. Whatever vB 3.7 offers
    4. Unique content: we’re thinking about interviews with scientists. (We’ve got one pending now, actually.)

Leave a Comment

Your email address will not be published. Required fields are marked *