What information is important to include in a question (and when)?

Question

A problem that's often encountered on UX is that often you don't quite know enough about the situation to answer. Or maybe you think you do, you answer, and then the asker notes that your idea is great...but they can't use it because of this major design constraint they didn't mention.

So what makes a good question? What sort of information do you need to make a reasonable answer? Specifically, what is often missing? Things like mockups, what you've tried and what you're trying to do are always relevant. Some other things like audience and specific design constraints may be important. What are the common things you see (or don't see) that keep you from being able to confidently answer a question?

I'm hoping to use the feedback here to make an faq post detailing what information your post should have and when, so we can improve the site for both asker and answerer by making our questions clearer and more answerable.

Answer 1

• Tell me what solutions and patterns you've already considered, then tell me why they didn't work for you. This doesn't just stop me giving you an answer you can't use - it helps expose the constraints.
• Describe the use case to me. Tell me why users are interacting with your interface, and what point they're at in their offline workflow. Tell me what users are already expecting; what experience they have; how much they know about the domain and the way the application works. If you're lucky, this process may give you the answer before you have to ask for it.
• Describe the nature of the content; what sort of data, in what sort of form, with what points of difference.
• Accept that research isn't always forthcoming. HCI is a scientific practice that can only evaluate design options with discrete, measurable variables. Some web design elements are highly context specific in their effects, so they might not easily lend themselves to that sort of study. You might find other webmasters' own studies and A/B stats, but this data can be unreliable or site-specific.