I have used multiple techniques to do this, depending upon the amount of time available on the design project. At the very least, you need to have focus as to what you are designing, and know the target audience for what you are doing. Shakespeare wrote "the readiness is all" and before you are ready to dive into the UI "layer" you have to understand those two core things
There are best practices that can be borrowed and applied from other industries, and I've borrowed two key processes from journalism and film pre-production to define my "foundation"
The first is the Journalism 101 principle of Who What When Where and Why:
Who is this design for?
What will they be doing?
When will they use it (how often)?
Where will they use it?
Why would they use it (as oppossed to another process or application)?
This last question gets to the heart of the value proposition of what is being designed/developed. In my opinion, if you can answer all these questions to YOUR satisfaction, then you are ready to get to the next step in designing the interface.
So, how do you get the answers? Well, some of these answers can be provided by the business analysts you are working with, or the key stakeholders… But I've found that the best way to get the answers is to talk to the users themselves, to understand thier workflows, thier needs and frustrations… by doing it yourself the answers will be more "personal" to you and will remain "top of mind" as you do your design work.
The second process is storyboarding or user stories. I think you don't need both, as they both service the same need. You should do whatever you are most comfortable doing. if you're a person who sketches a lot, do storyboards. If you enjoy writing, then do user stories. The point of this is to define a "flow" of the tasks that the users do in and with what you are designing – always keep in mind this is NOT UI design, you are capturing and defining the user experience. You can share these artifacts with key stakeholders to review and get concensus and THEN, after all are agreed, start doing the interface design work.
There's more things that you can do (persona definition, ethnographic research, competitive analysis, focus groups) but I consider the above to be the bare minimum.