I notice that the GIS Stack Exchange Tour page doesn't explicitly state that there should only be one question per post. The closest it comes is:

This site is all about getting answers. It's not a discussion forum. There's no chit-chat.

Can we change this to state that followup questions, which substantially change the original question, require a new post?

An example of why this is necessary is shown in Modifying placement properties for labels using ArcPy?. I can sense the OP's frustration ("just answer the question!") but IMO the original question has been answered ("it's not possible" is a valid answer) so a new question is warranted.

Does the GIS SE site state that a new question is the correct approach in this situation?


4 Answers 4


The Tour is now updated


Thanks for the feedback and improving the GIS-SE site content


I can't find a place where it explicity states "one question per question" (in an official capacity, though we do have Framing (asking) good Questions for GIS Stack Exchange? linked from the help), but I have understood that to mean/be a response to questions that give all of their problems at once. In effect, the question is too broad.

One good example I have from recently is How do I create zones using location features? There are three distinct questions there. And all three have multiple considerations and possible approaches. A single answer to the asker's 'question' as a whole could (would) be very long. Another example that comes to mind is Network Analyst: Location Allocation for Model Builder where the basic question is 'troubleshoot my model'. We solve the first specific issue with the asker then moving to the next issue that came up. That turns into a running dialog rather than a Question and Answer.

In the example you cite, I think (and have understood from site help files) it's perfectly ok to answer 'I don't know how/you can't do it that way, but you could do it this way' with a full answer/explanation, even if it differs from the original question. This stems from the How do I write a good answer? page which states:

What, specifically, is the question asking for? Make sure your answer provides that – or a viable alternative. The answer can be “don’t do that”, but it should also include “try this instead”.

If there were no possible work-around, then by all means the answer should be 'you can't and here's why.' But in your example, you can't but you could do it this way would be a great answer to that question. Search results will be both questions and answers that have the terms - you could solve it without requiring another search. You did give the answer (just not the details), so I can see going either way - a new question or an answer there. I think the latter would be most helpful, or at least linking the current question to the new question (or one that already answers it, should such exist).

In short, I don't think it's a hard-and-fast or black-and-white rule. I think it's a case-by-case sort of thing, and some of them are fuzzier than others.


I think we overuse the "too broad"/>1 question closure reason (actually I came to meta to open a new discussion on that, but found that this germane question was near the top of the active list!).

It is, of course, true that questions which are too broad are not suited to the SE format. And a laundry list of multiple questions which require unrelated answers is a not infrequent example of that.

But we routinely "penalize" (i.e., close) questions which are well thought out, where the questioner has tried to thoughtfully decompose the problem into related subquestions. Rather than creating an unwieldy omnibus, that decomposition indicates thought and prior effort by the questioner. And an answer is likely to meaningfully address all of the subquestions productively.

In particular, I am sad to see 3 close votes at Importance of WGS84 realization for centimeter-level accurate UTM transformation, which I think is a well thought out and structured question that is not "too broad", even if the questioner has helpfully asked 4 inter-related subquestions (and bolded the most important one).

There has been various discussion here over the past year or two how we can broaden our knowledge base to broader GIS topics, not just technical support for GIS software packages. I do think this is related: we seem too often apply the "too broad"/"too many questions" closure reason especially where someone is asking a more conceptual question, and where they are "showing their work" with their logical decomposition into sub-bullets. By doing so they are in effect doing the analogue of showing exactly what they typed and what the error message was in a more tech-support type question.


I think multiple questions in a Question, whether they be posed at the outset, or via the Question scope creep you encountered in this instance, are too broad for the Q&A format we use.

Making this clear in the Tour (if possible) or on the Asking and/or Don't Ask pages seems like a sensible thing to do.

Following up on a suggestion by @ChrisW in a Comment here, another heading could be added to one or other:

Your most important question is important to us

If you ask more than one question within your Question, you may see it closed as being "too broad". It is harder for potential answerers to try and provide multiple solutions, and difficult for later readers to try and follow which answer matches up to which question, so just ask your most important question, and then research/ask the others separately.

Once your most important question has been answered, reward the person who helped by Accept-ing their Answer and, if you have follow up questions, research/ask them separately to avoid re-entering the "too broad" scenario which could lead to it being closed to new Answers.

For Tour wording perhaps something just a little "punchier" ...

Your most important question is important to us

Asking one, and only one, important question within your Question helps attract prompt and clear Answers.

Your other questions are just as easy to research/ask separately!


You must log in to answer this question.

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