0

In the sidebar of SU there are various stats on the different pages about Questions - questions without answers, count of questions.

According to this the time to answer a question monthly average is around a day.

If I am a new user, coming to SU and haven't found an answer to my question, I have to make a decision about whether I should go to the trouble of creating an account and asking a question, or I should keep searching and perhaps ask elsewhere.

So putting the average time to answer as a statistic in the sidebar might help a new visitor make the decision to stick around. In chat, the main objection raised was that this might set an expectation for a quick answer to a low quality question.

This could be mitigated by having a text link that directs people the "how to ask a good question" section of help. "What kinds of questions get answered quickly?" for example.

Let Me Paint You a Picture

The Unknown User arrives at SU. They have the best question in the world. They have the type of question that once answered will help people for generations. It is the type of question that will get Ramhound salivating at the opus he might write in response. But we don't know what the question is, because they haven't asked it yet.

"Is this for me?" they think, as they look at the home page. "Wow, 300,000 questions asked, lots happens here". So they proceed to click around.

"Oh look, unanswered questions. Hmm, nearly 100,000 unanswered questions. So one in three questions does not get answered. Perhaps the site is in decay. Perhaps they can only answer questions about topics other than my question. Perhaps I will look for an alternative resource that focuses on my specific topic."

And the Unknown User departs, and takes with them the best question ever to be asked.

My first thought when considering this is, perhaps an indication of activity would be better than an indication of inactivity. So perhaps an average time to an answer might tell the Unknown User to not go elsewhere, that here is the place for their extraordinary question.

It doesn't have to be average time to answer. Or median or mode. It could be anything. Numbers can tell a story very quickly, which is why the site is peppered with them.

So let me rephrase this question - what kind of statistic do you think would be the most compelling for the Unknown User to invest their time in asking their incredible question at SU rather than elsewhere?

Here I am trying to shift the consideration away from those users that will ask crappy questions and demand fast answers. We will deal with those like we always do. They will come no matter what.

I am trying to fix your attention on those that have the brilliant questions, and I am asking if a question count and an unanswered question count are the ideal metrics to show these amazing question holders to send the message that this is the best place for their question.

14
  • Do we really want to give people any reason not to ask a question? We have never been the type of community, you should seek out, if you expect an answer "right away". What does it matter if its a week or an hour, if the question is good, and somebody knows the answer you still get help eventually. People already don't read the help center, so having a link, isn't going to prevent the concerns people have.
    – Ramhound
    Apr 19, 2016 at 14:22
  • This might get skewed by unanswerable questions. I'd love this as a high rep user privilege tho .
    – Journeyman Geek Mod
    Apr 19, 2016 at 14:35
  • Average time doesn't really tell you much, and it doesn't relate time to answerability. A more useful indicator would be a frequency distribution. Cutting it off at something like 4 days or so would provide detail covering most answerable questions. If you want a single number, maybe something like percentage taking more than 1 day.
    – fixer1234
    Apr 19, 2016 at 15:42
  • @Ramhound My thinking is that this would encourage new visitors to ask a question, giving an indication of timeframe. My feeling is that most people that have a question like it answered sooner rather than later, and understand that some questions are answered sooner and some take longer. I am talking about getting someone across the line in making the investment to ask the question in the first place.
    – Paul
    Apr 19, 2016 at 23:34
  • @fixer1234 Average time does tell you the average time. It might not be immediately apparent to a new visitor how vital the community is at SU. "I could ask this question, but if no one is around to answer it for weeks then I might go elsewhere"
    – Paul
    Apr 19, 2016 at 23:36
  • 1
    I am more inclined to ignore questions from people who expect us to help and/or expect it on their timeline.
    – Ramhound
    Apr 20, 2016 at 1:04
  • @Ramhound I don't think I have explained my premise very well. This is about setting an expectation to new users that this is a vibrant and active community of people willing and wanting to help others.
    – Paul
    Apr 20, 2016 at 1:13
  • I still believe it's a mistake to setup an expectation that an answer will come quick, I spend days writing answers, my best answer took 2 weeks to write.
    – Ramhound
    Apr 20, 2016 at 1:20
  • @Ramhound The average timeframe for an answer is a day, which suggests that many questions do not need days to process. This doesn't imply anything about the right answer or the best answer, just that if you ask a question, it will be attended to. Managing expectations is straightforward, it is even included in the question in paragraph 4.
    – Paul
    Apr 20, 2016 at 1:39
  • @Paul: The average is somewhat misleading because the distribution is extremely skewed. The fastest answers come within minutes but the first answer sometimes shows up months, even a year later. The majority that get an answer probably get one within a day or so. The median or mode would be a better indicator. Rather than a number, though, it might be better to have a sentence that explains the nature of the process and not to expect an immediate response.
    – fixer1234
    Apr 20, 2016 at 5:07
  • @fixer1234 There are numbers already, and they are looked at ahead of anything else. I have updated my question to better explain the scenario I am thinking about.
    – Paul
    Apr 20, 2016 at 6:07
  • @Paul: Your revision targets people with great questions who want quality answers. The quality of the existing knowledge base tells them what they want to know about whether to ask here. Stats on answer time aren't what will make the difference. They can also look at existing threads to get an idea of typical response time. Much of the value is in comments, which wouldn't be covered. For someone interested in expected response time, they want to know about "now", not some averaged period. For that, stats about the current online audience may be more relevant.
    – fixer1234
    Apr 20, 2016 at 6:53
  • @fixer1234 That assumes that just because someone has a great question, that they know that it is great, and know that they want quality answers. They may have the best question in the world but to them, something is just broken and they want an answer and don't want to waste time explaining it to a low-effort audience. The idea that people will do a whole bunch of research before committing time to a website is counter to what actually happens.
    – Paul
    Apr 20, 2016 at 7:51
  • You are concerned about the first-time visitor (not even registered yet; not even a lurker) going elsewhere because she doesn't perceive SU/SE/etc. as a vital community. For her, it might be persuasive to display numbers like number of questions asked and number of answers posted in the past day or week. I have a different concern — the user who asks a question and then gives up in disgust (never to return) when it hasn't been answered within four hours. It might be particularly useful to tell him something about how quickly questions get answered — perhaps after he's posted his question(?).
    – Scott - Слава Україні
    Apr 22, 2016 at 17:19

0

Reset to default

You must log in to answer this question.

Browse other questions tagged .