The first episode of Working (a new podcast) features Stephen Colbert explaining in great detail the process of creating The Colbert Report. The entire episode (embedded above) is really good process stuff (creativity, collaboration, finding the story, media firehose, working under pressure) but I want to call out the section about how he prepares and uses the questions for interviewing his guest, as it’s is quite consistent with what I wrote in Interviewing Users.
And then I read the two sheets of questions that the writers have come up, what their ideas are. I usually pick 10 or 15 of those. But I don’t look at them. I don’t look at them until right before I go over [to the set], and then I read them over once again in front of my producers to get a sense of, oh, this is how my character feels about this person.
Come show time…I take them out and I go, oh, yes, these are the questions I chose. And then I try to forget them and I try to never look at the cards. I just have a sense in my head of how I feel. And the cards are in front of me, but I try not to look at them at all. I’m pretty good. Maybe I look once a week at the cards. I put my hand on them, so I know I have them if something terrible happens, but as long as I know what my first question is for the guest I kind of know what every other question is, because I really want to react to what their reaction to my first question is.
And I usually end up using four of the 15, and the rest of it is, what is the person just saying to me? Which makes that the most enjoyable part of the show for me. Because I started off as an improviser. I’m not a standup. I didn’t start off as a writer, I learned to write through improvisation, and so that’s the part of the show that can most surprise me. The written part of the show, I know I can get wrong. You can’t really get the interview “wrong.”