You've probably read Q&A intervews where a celeb, politician, business leader--whomever--is asked a series of questions, and answers those questions. It sounds like a plum job for a writer, or a non-writer; all you have to do is chat with the subject, transcribe the answers, and turn in the manuscript. You could do that, nothing to it!
But it's not quite that simple. Before you do anything else, you have to figure out the interview slant, and whether the questions will be issue- or event-driven, or subject-driven. (Examples: "So, what made you hate spinach so much?" or "Why did you destroy 30 acres of spinach with a flamethrower?" or "What attracted to becoming a green-vegetable expert?") Which of course means you need to research your subject, maybe find some earlier interviews with her.
And yes, you really need to do the research, no matter how big a fan you are of your subject, and no matter how much you think you know about him. There's always something new you can dig up. (One of the joys of doing such research is finding some fact that no other interviewer has discussed with the subject. Sometimes the subject likes it, too. Answering the same questions from interviewer after interviewer gets old.)
Armed with your research and a list of questions based on that research, you sit down with the subject, in-person or by telephone. After you're set up recording equipment, of course; you weren't planning on just scribbling shorthand, were you? You'll need a recording for all sorts of things--the nuances of meaning conveyed by voice inflection, the demands of your editor to prove that your subject really said that, and the claim by the subject that she said no such thing.
Okay, so you sit down and ask questions. Or do you?