I think when I was doing my very first interviews, I probably brought a notepad and did ask people my first fifteen questions while sitting in a Starbucks or something horrible like that. And I found that, oftentimes, the most important thing at the very first interview is just establishing a personal connection and developing some sort of rapport so that I can go back to them again, and then maybe again, and maybe again after that.

Often, I'm spending months with a person in a very intimate context, getting to know the ins and outs of what they ate for breakfast, not to mention dredging up the most traumatic experiences of their lives, digging through their documents and photographs from difficult times, all of that. And that process, I think, can be extraordinarily strange for subjects who've never been interviewed before, especially if you don't acquaint them from the get-go with what you're trying to do, what it entails, and why you care.

I think when I was doing my very first interviews, I probably brought a notepad and did ask people my first fifteen questions while sitting in a Starbucks or something horrible like that. And I found that, oftentimes, the most important thing at the very first interview is just establishing a personal connection and developing some sort of rapport so that I can go back to them again, and then maybe again, and maybe again after that.

The tremendous challenge of narrative journalism about subjects that are underreported is, how do you make people care about something they think they already know about, or think they don't need to know about?

I tend to gravitate toward the "act two," or "act three," or "act four" stories - either things that are underreported, where we think we already know the common narrative, or things that are at the margins of an over-reported story, where we're all so focused in one direction that we're missing something crucial that's unfolding off to the side.

One thing I've discovered is that I never think of something that didn't work out as just "something that didn't work out." I think so often with investigative work, things that initially look like failures wind up leading to your biggest stories.

One thing I've discovered is that if you remain in contact with people, if you build longitudinal relationships, if you invest in sources who seem at first like they're uncomfortable or unwilling to talk, if you keep in touch with them, a year later that might yield something much more powerful.