Visiting guest Richard Cook demonstrated before an audience how he would approach an interview as part of an incident retrospective. A volunteer agreed to relive a recent incident as a "straight man" in this performance.
Things we have learned from watching Richard / discussing technique with him:
• Start by establishing a connection. Don’t just tell them this is a conversation, make it a conversation. Ask them about something they’re comfortable / proud to talk about.
• Ask a simple question and then STFU. (This is excruciatingly hard for me. I want to shape a question and outline alternatives and show off how much I already know and prompt…)
• “What do you mean by ____” is almost always a good follow-up question. Never assume you know what someone means.
• If you shut up and stay quiet, people will usually keep talking.
• Don’t overplan. (Another hard one for me.) Follow the conversation, not an agenda.
• Stay above the line. Don’t get distracted by technical stuff unless that’s specifically what you’re going for. This is why it’s hard for domain experts to conduct effective interviews.
.
Beth reported these observations in slack. I asked for permission to report them here and call them ours.
> I especially liked it because it didn't try to be a tutorial. I was there and still found it useful. But almost all of it could make sense to anyone, The one exception is your mention of "above the line" but you still explain it and drive home the liability of being an expert.