One of the most important processes of designing a game is to concept what you are going to create. It helps to develop an initial idea into something that you can work with regardless of which area of the team you are.
The most obvious is 3D modellers using artwork from the concept artists to create the models. However the concept artists have to get their ideas from somewhere else, they will collect images and photographs and create mood boards before drawing out their ideas, unless they just doodle ideas from their head. I also feel that games with sequels can be used as concept for the newer games, a good example of this is Skyrim. Most of the Elder Scrolls games have references to the others, Skyrim has books in-game detailing the history of Oblivion and Oblivion has books about Morrowind.
In a way the old models from Morrowind can be described as concept for Skyrim, as some of the same NPCs and props are used but have been remodelled in Skyrim. Also the different races have been greatly improved.
Personally I am not great at concepting, I generally prefer to draw things from life and find it difficult to create an idea from nothing. I have to do a great deal of planning and collecting imagery before I can create something “unique”.
Planning is the first stage in the creative process of tackling a brief, and the initial brief and final outcome can get messed up if sufficient planning has not been put into place. An important part of planning is knowing what you're going to do and how you're going to do it. Making lists of what I need to do helps me get my work done, although I am not very good at time keeping, and I hear that it's a good skill to have, so I will try and focus on that this year! Some people use timetables to help with their planning, but I feel that the time spent on creating the timetable might overweight the actual project. The final outcome of a project should be obtained by the development of initial ideas. Planning and concepting will help to focus on the initial idea and to not stray off into a random conclusion.