13

I have a good "problem": a fairly new Stack Exchange site I participate on1 took off running. Our users are active, we get lots of engaging questions and answers, and our metrics are in good shape.

On some sites (with SO as the poster child) it's impossible to keep up with what's going on; you carve out your little niche and pay attention to that, but you're unaware of the vast majority of the doings on the site. At that scale, it can't be helped. But on smaller sites it's often possible to keep up, and for a young site it's especially important to do so because site norms are still fluid -- scope, guidelines for behavior, and the results of experimentation are all somewhat up in the air compared to a mature site.

As a user who wants to be part of bringing all that together, I want to read everything -- but there's enough that I can't. And our community is young, so we're still training people to bring things up in our discussion areas. On other sites I also might not read everything but I'm reasonably confident that we aren't forming pockets of users who might need more attention than they're getting, or developing interesting and novel trends out of view. Imagine if SO were new and you found out that the Perl users were all off in a corner asking and answering obfuscated-regex questions for the fun of it -- not what the site is meant for but they didn't know any better and nobody noticed.2

What are good ways for me to maintain a good "aggregate" view of doings on the site, in addition to the questions I read out of my own personal interest? I can imagine a few approaches; I'm interested in (a) more input and especially (b) reports of what has actually worked for people in this situation:

  • I could pick a couple tags, read everything on them for a week or two, and then move on to a couple more tags -- taking deep samples in narrow areas, in other words.

  • I could do the same with users -- pick a few people to stalk for a while, then move on. Maybe we wouldn't actually call it "stalking".

  • I could (wave my hands and) use a script to read every Nth question and all its answers.

  • I could use queries or on-site searches to find, um, something. (What might I be looking for -- high votes? High views? New users?)

Let's assume that whatever I do is on top of my regular use of the site -- reading, asking, and answering questions that are personally interesting, using the review queues, and keeping an eye on meta and somewhat on chat.

1 And also moderate, but that's not the point of this question. I'm aware of this question, which asks more from the perspective of moderators, but I'm asking about the user perspective.

2 I hope that's not real; I don't hang out on the Perl tag on SO. :-)

  • did you consider tracking site meta / chat? – gnat Mar 18 '15 at 10:01
  • 1
    I do track meta closely. (People don't always know to use it, but we do have a healthy meta.) Chat is...voluminous...and I do my best but it's hard. – Monica Cellio Mar 18 '15 at 13:00
7

As Jan Doggen already mentioned, you could actually use a RSS feed. However, this problem seems to be more complex. The real problem is: You have too much content for too less time.

You already showed four possible solutions of your problem. I will try to make it clear to you which solution is the best in your case.

  • "I could pick a couple tags, read everything on them for a week or two, and then move on to a couple more tags -- taking deep samples in narrow areas, in other words."

    1. As you already noted, your site is "fairly new". So, now you want to take some deep samples of tags still evolving. You actually can read everything concerning some tags, however, topics will change by a short time due to the developing process of your community and as a result of that, the whole tag environment (tags, users, questions, answers) will change. After you will have left those tags then, they will change to something different.
    2. You lose time. If you keep track of specific tags, you can't keep track of other tags. Therefore you can't take action against things going wrong. You also can't benefit users for doing things right. Imagine this: You have taken care of your particular tags, and are ready to see the next tags. However, in the time you were busy with the tags before, the tags you're concentrating on now aren't like you want them to be. So you can start off from the scratch, again - yay!
  • "I could do the same with users -- pick a few people to stalk for a while, then move on. Maybe we wouldn't actually call it 'stalking'."

    1. The advantage is: User behaviour doesn't change that quickly or unexpectedly like tags can. Tags change because of the different ideas gathered by users contributing. Users change because they are users. Usually, they come to your site to answer this one question and everything related to it.
    2. It is routine. The user you "stalk" may be interesting for one or two weeks but after that time you can usually spot a specific behaviour, e.g. this users is a good contributor, this user answers question as quickly as possible.
  • "I could (wave my hands and) use a script to read every Nth question and all its answers."

    1. Chance won't help you. You most likely will end up seeing the most popular tag way more often than the other tags since there are less questions.
    2. You may come across content that isn't readable for you (because you are not that much interested in that topic).
  • "I could use queries or on-site searches to find, um, something. (What might I be looking for -- high votes? High views? New users?)"

Your last approach is a good one. Nevertheless, it needs to be more elaborated to work. You want "to read everything" - that's impossible. To maximize your power of reading (I just call it like this now), you must search for specific questions and users, but not for tags.

  • Users are a good way to keep track of the content. You will come across a few users who are really encouraged to answer every unanswered question. Look up their profile to get led to some questions and answers.
  • Don't search for unanswered questions. In order to see the most content, this is not helpful. An unanswered question just has the half of content it could have - it simply misses an answer, in other words, content. (Of course you must decide this on your own, maybe you want to answer as much questions as possible, too!)
  • Search for high rated and low rated questions but not the average because average is average. High rated content is high rated because the content is really high valued. Low rated content is low rated because the content is low valued, and here you come. Just because some content is low rated, it doesn't mean that it's bad. Often, there are answers which are just different and not accepted. If you look out for them, you might extend your view of sight in a specific topic.
  • Use your time efficiently. Don't "waste" too much time on Meta. Meta is content as everything else, but it normally is content which is meant to discuss other content. That would be interesting as moderator but not as your wish as user.
  • Never search for tags. If you search for a tag, you automatically exclude everything else beside this one tag. Since you want to read everything (which I interpret as "everything" - every tag, every question, every answer, no restrictions on one tag), this would be really negative.
  • Collaborate with a user having the same intents as you have. You can split the content then. User A searches for this content, User B searches for this content, and the most interesting content is shared by providing links to it.

You clearly can't read everything but you can maximize it by looking up particular questions or collaborating with users.

To show an artistic comparison: Imagine a huge, empty cuboid. It's stuffed with an adhesive mass, your content. Also, it's stuffed with your users, a huge mass, too. And then there are tags breaking up this mass into little pieces of mass. However, if you look at a piece of mass, you can't see the rest of it. To see the mass at its whole size, you must look at the quader, and try to consume (eat) as much of it as possible. Inhale it!

2

Maybe following posts using RSS is an option?
I really like the ability of RSS to give me a list of titles which I can pick from - it significantly speeds up my handling of streams of data where I don't want to read them all.

The linked meta post mentions the open source StackToRSS converter.

Note: I found this searching for RSS on meta and plan to use it for an upcoming project where I need to monitor at least 2 StackExchange sites.

2

Use the meta. If there is something going on in a tag you don't frequent, the hope is that someone will alert to that on the meta. If some tags need to be split or reshuffled, that should happen on meta. If a new custom close reason is needed, again you're likely to see that on the meta.

At the same time, model what you want by using meta yourself to work out whatever decisions and conclusions you're coming to in your corner of the site - about what is on and off topic, what needs spoiler markup, etc.

If you have time on top of that then the review queues will show you the problematic parts of the site in a cross-tag way. This won't help you learn what the other parts of the site are supposed to be like, but it will show you what problems the site as a whole is dealing with (eg attracting certain kinds of offtopicness.)

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.