35

I often hear the argument that the end goal of Stack Exchange is not to provide individual answers to questions, but to act as a knowledge base for people coming from Google:

You are just completely missing the point of SO. It is not to give an individual user help. It is to give the legion of future googlers a good hit. The one he couldn't find. [...] – Uphill Luge (source)


Stack Overflow does not exist to help someone when they have problems. Oh, we do that, but only as a means to an end: to help the next person who has the same problem. Helping you in a way that doesn't help anyone else is an anathema to what we do here. – Nicol Bolas (source)

I recognize that we shouldn't just be serving help vampires, but this argument doesn't make much sense to me. It is my understanding that Stack Exchange is intended to act as a venue where users can reliably get expert answers to questions and contribute their knowledge to their Internet community, not solely the latter. I feel we should focus on the former and deliver the best possible experience for users of all levels of expertise. We should welcome all content as long as they are within our guidelines and show reasonable effort, even if the underlying goal is just to get help for, say, a code issue specific to a personal programming project.

Is my viewpoint acceptable to the community? Is it best to focus primarily on building a knowledge base for Googlers, or should we try to provide a venue for Q&A that everyone can participate in and rely on? Should the latter really be sacrificed in favor of the former?

  • 1
    To a certain extent it's always going to be a balancing act, one tempered by each individuals compassion/heartlessness/irritation and a companies desire to have a lot of people visiting so that they can sell more advertising. The company doesn't want to annoy the people who provide the answers (and had noble intentions from the beginning) but at the same time can't afford to push too many people away or there won't be answers, so no one will come, so no advertising revenue... – ben is uǝq backwards Jan 22 '14 at 17:15
  • It's to make it possible to select an answer that has a probability of truth, even if you don't know enough yourself to evaluate the reliability of the tons of garbage elsewhere online. – WGroleau Jan 22 '14 at 17:28
  • 13
    If people are asking good questions then there is no difference between the two goals. – Bill the Lizard Jan 22 '14 at 17:29
  • 4
    The point of that statement to me is not so much "we don't like you all that much and we don't care about your problems" but rather that if your question will only ever help you and nobody else, then the question is not good for the site. Perhaps if it were phrased like "we love helping you out, but only if that's of value to others as well", you wouldn't find it so problematic? – Bart Jan 22 '14 at 18:27
27

Not Exactly...

Our main goal is to help the most people we can.

So, we really, really love re-usable questions.

A question that is likely to viewed by, and helpful to, tens of thousands of people, is very high on the list of things we want.

But helping 100 people is pretty good, too, and helping 1 is better than 0

It makes me a tiny bit crazy when I hear someone say, "you can't ask that, because it'll only help you".

Most people - even nice people - who ask a question are extremely focused on the help they need, not the approximate estimate of others in a similar jam.

So, if a question can help one person, someone is willing to answer it, and it has no other impact on a site, it should be encouraged.

The complexity comes in when one-offs interfere with the ability to help more broadly.

  • If the front page is full of "find my typo" questions, it could reduce the visibility of the more beneficial methodology questions, and reduce the net people helped.

  • If the front page is full of increasingly tangential questions, eventually experts won't recognize it as a site about their subject, and you'll have fewer people who can help.

Questions should NOT be closed simply because they only help one, or too few people, but patterns of questions that threaten the greater good should eventually be made off-limits.

  • 2
    I think you hit it right on the nail. We need to strike the right balance between helping solve the user's particular problem and helping the Internet community as a whole. The check goes to you. – bwDraco Jan 22 '14 at 22:05
  • 10
    This deviates from what Stack Overflow and Stack Exchange are (were?) all about — “build a library of detailed answers”. There's no point in building a library that no one will read. A question always has an impact on the site — the time I spend answering that chien means dog is time I haven't spent elsewhere. I didn't sign up to help the asker, I signed up to build a repository of answers. – Gilles Jan 23 '14 at 1:00
  • 6
    helping 1 is better than 0 That's naive, IMHO. This should not happen at simply any cost. There are more factors involved in the equation. That said, I generally agree with your answer. – Lightness Races in Orbit Jan 23 '14 at 2:04
  • 15
    I kind of strongly disagree with this "if it only helps 1 person that is OK" advice. That is setting aspirational goals far too low, and sends the wrong message about participating in our network. If you do not reach the goal of helping multiple people with your question, that is fine, but let's not move the goalposts underground in the process. – Jeff Atwood Jan 23 '14 at 6:52
  • 1
    @JeffAtwood, I'm not sure we're disagreeing that much. 1. Helps more > Helps less. 2. Askers needn't prove how many they think can be helped. MOST questions can help others. (And the ones that usually can't AND tend to become significant clutter, are already OT for other reasons.) 3. Answerers needn't overthink trying to test questions to ensure they help enough to be allowed, but they should by all means focus their energies on the questions that can help the most people. – Jaydles Jan 23 '14 at 14:22
  • 11
    @Jaydles "So, if a question can help one person, someone is willing to answer it, and it has no other impact on a site, it should be encouraged." based on this sentence, we are disagreeing quite a bit, I find this sentence highly objectionable and out of tune with network goals. Selfish "but I got mine" justifications are not something I consider acceptable. Everyone is a part of the greater community of peers and should always be thinking about how their question moves the community forward -- not just themselves. – Jeff Atwood Jan 24 '14 at 1:10
  • 3
    I like the ideas espoused in this answer. It seems an undue burden to expect one to estimate impact either as an asker or answerer. When I help an old lady across the road, I don't stop to think "well, I could help more people with the additional tax revenue I would generate by not stopping to help, so I'll just carry on". Sometimes helping one is good enough. – copper.hat Dec 6 '14 at 23:55
  • @Jaydles, Why do you write "not exactly"? That's synonymous with "yes, partly". – Pacerier Aug 28 '15 at 13:21
  • 1
    @Pacerier, apologies - not sure I follow your question. That line in my response was meant to convey that the asker is right that we try to optimize for helping more, vs. helping fewer. But that I disagree that we should avoid helping when we're not sure if the answer will serve "enough people". So it is partial agreement, I think - I'm basically agreeing with optimizing for a sort like the OP implies, but not a filter. – Jaydles Aug 28 '15 at 13:26
18

I have always felt that those who express this sentiment are taking a good idea to an extreme. It is true that on balance, the needs of future visitors outweigh the needs of any specific asker. But it is not true that this means we can simply disregard the needs of askers for the needs of those who might later benefit.

After all, if we did so, we would only be duplicating Wikipedia, since if we ignore people's problems, we only have answers, and Wikipedia already does a good enough job of that. What SE does, to a large part, is provide teaching, context, and solutions. Further to that, there is little reason to distinguish someone who couldn't find the answer they needed on Google and asked a question from those who later benefit simply because they arrived later.

In general, "good answers to good questions" resolves this conflict, as it typically results in a win-win situation. Beyond that, a little kindness never hurt anyone; while we should try to avoid creating broken windows, blind enforcement of the letter of the rule is also something to be avoided.

  • 6
    Not exactly. There is one article on asphalt on Wikipedia, but thousands of possible questions about asphalt. We provide fractal levels of detail compared to Wikipedia. Also consider that you don't have to read a giant complex deep page full of dense text to get to that one nugget of specific info that you are looking for. – Jeff Atwood Jan 23 '14 at 6:54
11

Is the ultimate goal of Stack Exchange really to build a knowledge base for Googlers, not to help individual users (even if they ask good questions)?

Yes.

  • For those questions which are well-posed and well-researched, the two are the same.
  • For those questions which are not, they are not ready to be posted.

Stack Overflow is a programming Q&A, not a helpdesk. Alas, due to the inevitable increase in scale that it has experienced over the last few years, this is incredibly difficult to maintain, since the vast majority of human beings only care about getting help, not about populating a Q&A database.

For more information:

10

Most of the time, there is no difference between these two goals. I ask a question about a code problem I'm having, you answer it, then a year from now someone else has the same problem and googles it and is happy to find the answer.

Where the difference commonly arises is two places, which are covered easily with Close reasons, and one issue with answers.

First, it is in questions that are either too large in scope ("do my project for me"), which have both the problem that it's unlikely anyone would ever have the same issue, and that it isn't really one single issue (thus not easy to search for or identify). These questions are "Too Broad".

Second, is in questions that are too specific to users' issues: typos or other issues that do not arise from a lack of understanding or a poor approach to a problem, but from mostly-unique circumstances (misspelling a word, etc.), or issues that require understanding an entire (large) project to comprehend and thus would be not useful for similar reasons to the above. The former have a new Off Topic reason, the latter is again Too Broad ("too long for this format").

Finally, with answers, we ask that answers not be solely links to off-site resources, even if that sufficiently answers the question. This is the one time these two goals might be seen to conflict: the fastest answer is pasting such-and-such link to documentation, and actually explaining will be inferior in the space available. However, this isn't really a conflict either; it simply asks that you explain the link in some detail prior to posting it as an answer. It may delay the 'help' by a few minutes, or even a day or two if you're busy, but hopefully nobody is coming here with a question so urgent that this is truly a problem.

  • 2
    That's not true- link-only answers don't serve the Google group either because the link may well be dead by the time they arrive. – DeadMG Jan 22 '14 at 18:45
  • 1
    @DeadMG Link only answers don't serve the Google group, no. That's the point. Link only answers serve the asker. – Joe Jan 22 '14 at 18:57
  • Yes, but you don't describe it as a conflict, when it certainly is a conflict. – DeadMG Jan 22 '14 at 19:19
  • That's my final paragraph... – Joe Jan 22 '14 at 19:48
5

We should be simultaneously helping individual users and building a resource for the googlers of the future. Good help is sustainable help that will benefit once and future seekers.

If you give good answers to the current askers, then those answers are bound to help someone searching in the future.

  • Agreed - except if the questions are so localized they're unlikely to help anyone else. That's why we don't want those. – Pëkka Jan 22 '14 at 18:45
  • 2
    @Pekka, yes, I'm all for closing too localized syntax error and typo questions. – Lance Roberts Jan 22 '14 at 18:46
  • 3
    Strangely enough, I've found typo questions that happen to contain extremely odd compiler warnings / errors to be extremely helpful at times - probably saving at least an hour by knowing what to look for. I do agree in general, though, the majority of these would likely be more pollution than help in search engine results, which does make them a tad harmful with a few exceptions. Also like how simply you put this Lance, it really is just common sense. – Tim Post Jan 23 '14 at 5:22

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .