2

What is the best way to go about posting an informational question when there are existing near-duplicate questions but they do not contain keywords essential to people searching for the information?

Stackoverflow (and to a lesser degree, the rest of stack exchange) have stunning SEO. When I have a problem and a stackoverflow post is the first result, there is a good chance there is a solution I am seeking in concise form. Thus, when I have a problem but find no stack overflow posts in the Google results, I sometimes create a question even when I've already found the answer so that future Googlers (often myself several months later) are not left with the same desert of search results.

Recently, one of these questions got down-voted for being a near duplicate of another question. Gaping dissimilarities between the two questions aside, nothing appeared in my searches for this problem. My question now appears as the first Google result for my old search, but the question and answer are littered with discussion of duplication and this detracts from the essential point. Thus the question above. What does one do when existing questions on a topic are bad enough in SEO that they they may as well be different questions even if they deal in the same content? In particular, what if these questions have bad enough SEO that you don't even know they exist until people rail on your question for "redundancy"?

As an added thought, what if there are several disparate questions on a similar theme and you want to have a single question that addresses them all? For example, a question about databases where there are answered questions for specific databases that are fairly similar but none asking the question in general.

  • If there are enough similar duplicates, make a canonical question/answer. – Qantas 94 Heavy Oct 14 '13 at 2:50
  • @Qantas94Heavy How does one go about doing that? – Zags Oct 14 '13 at 2:51
  • You don't really need to; just pick the best of what's there and improve it enough, or generalize it a bit if needed, such that other similar questions can and should be closed as duplicates of it. – Servy Oct 14 '13 at 2:53
  • @Servy Last time I tried that, the OP rolled it back. I'd argue it's easier to just start a new question. – Mysticial Oct 14 '13 at 3:40
  • @Mysticial That depends on how much needs to change. Generally for questions asked a lot you can find one that's already very close to what you need, and that requires very little, if any, changes. If it needs a lot of rework then that's likely not a good approach. Most people are pretty bad at asking questions on SO when they know the answer to them; they don't end up as good questions, so for most people that's not a good first step. – Servy Oct 14 '13 at 14:15
4

Rather than intentionally posting a duplicate, just improve the original page. You could post an answer that uses different terminology, or if you're able to do it properly (i.e. without changing the underlying content or adding inaccuracies) you can potentially edit the question or an existing answer to use the proper terminology, thus improving searchability.

  • In this particular case, keywords on the existing questions were so bad, I couldn't even find them. Also, new answers tend to get lost under existing vote weight. – Zags Oct 14 '13 at 2:55
  • 3
    @Zags Well, if your purpose is to improve SEO, not get votes, why does it matter? If your goal is just to get rep, and not to improve the SEO of some question, then obviously you'd spend your time differently. If you don't see obviously correct ways to improve the existing content, adding a new answer is much "safer" and easier than trying to edit, if you don't feel comfortable with it. – Servy Oct 14 '13 at 2:58
  • 1
    It matters because votes (at least in a relative context) are signals of quality. Whenever I see a question with negative votes, I am usually skeptical. Also, I rarely read past the accepted answer and the one with the most votes. It's a heuristic that I suspect many people also use. – Zags Oct 14 '13 at 3:28
  • 2
    @Zags Your original premise was that the question already had valuable information, it just wasn't possible to find it when searching because it didn't use the right termanology. If your answer doesn't get votes, but it has they keywords that Google finds to direct a user to that page then it doesn't matter if they even read your answer; reading the older answer should still solve their problem. Dealing with a question that has incorrect or inappropriate information in an answer is somewhat different, again, than an answer that's good, but not discoverable. – Servy Oct 14 '13 at 14:09

You must log in to answer this question.

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