Interview craft: research, questions and listening

0:00 Interview craft

Have you ever listened to an interview where the guest simply recited their LinkedIn profile for forty minutes? Every answer polished, every story one you've heard before. Now think of the best interview you've ever heard. The guest said something surprising. Something honest. A story they clearly hadn't told a hundred times. That doesn't happen by accident. It happens because the host prepared deeply and listened even more deeply. In this lesson, you'll learn to research beyond the obvious, design questions that unlock stories, use a follow-up ladder, handle difficult moments, and use AI for research without letting it invent facts.

0:44 Why craft matters

Why does interview craft matter so much? Because listeners come for insight, stories and moments of honesty. Those moments are what people share, clip and remember. And they're what make guests want to recommend your show to their peers, because you made them sound brilliant. Great interviews feel effortless because the host did the work beforehand. Research tells you where the interesting stories are. Question design gives the guest room to tell them. And listening tells you when to go deeper.

1:19 Get into the kitchen

Here's an analogy. Think of a guest's knowledge like a house. Most interviewers stay in the living room, where the guest has greeted a hundred visitors with the same polite conversation. Your job is to be invited into the kitchen, where the real stories are. The failures, the decisions, the specific numbers, the thing they changed their mind about. Research tells you which doors exist. Good questions knock politely. And listening tells you when the door is opening.

1:53 Research

First, research. Read or listen to the guest's work: posts, talks and previous interviews. Note what they've said many times, their standard answers, and plan to go beyond them. Find specific moments: a decision, a failure, a turning point or a surprising opinion. And prepare accurate pronunciations of names, companies and places. Getting a guest's name or company right is basic respect, and getting it wrong is the fastest way to lose their trust in the first minute.

2:27 Question design

Next, question design. Open questions invite stories, like "how did you approach your first hire?" Specific questions get concrete answers, like "in your first year, what did you stop doing?" Story prompts bring emotion: "take me back to the day you lost your biggest client." Process questions give practical value: "walk me through how you plan a campaign, step by step." Contrast questions find fresh angles: "what do most people get wrong about this?" And listener-proxy questions keep focus on the audience. Avoid closed yes-or-no questions, double-barreled questions, leading questions and long preambles.

3:07 Example 1: generic to great

A simple example of turning generic into great. The generic question: "What's your advice for new freelancers?" The generic answer: "Work hard and network." Now a better sequence. "Tell me about your first paying client." Then, "what almost went wrong?" Then, "what did you change in your contracts after that?" And finally, "can you share the exact clause you added?" The guest tells a specific story about a late payment, and shares the payment-terms clause she now uses. That's practical, memorable, and very clippable.

3:44 The follow-up ladder

Here's the follow-up ladder from your lesson, which you can use in any interview. When an answer is interesting but vague, climb one rung at a time. Rung one, example: "can you give me a specific time that happened?" Rung two, detail: "what exactly did you say or do?" Rung three, stakes: "what was at risk for you then?" Rung four, lesson: "what do you do differently now?" And rung five, transfer: "how could a listener apply that this week?" You won't always need every rung. But knowing the ladder means you're never stuck with a vague answer.

4:27 Example 2: a much-interviewed founder

Now a realistic scenario. A host in Karachi interviews a well-known Pakistani e-commerce founder, who has given dozens of interviews. The host's research finds the standard answers: the funding story, the growth numbers, the advice to "focus on customers." So she plans to go elsewhere. Her opening question is about a warehouse fire the founder mentioned once, briefly, in an old talk. The founder pauses, then tells the full story for the first time on a podcast: the night it happened, what they nearly lost, and the insurance lesson every founder should learn. That segment becomes the episode's cold open, and the most shared clip of the season.

5:14 Watch me: AI research prompt

Watch me use AI for research, the safe way. I paste my own research notes and a transcript I'm allowed to use into an assistant, with this prompt. I'm interviewing this guest for a podcast for this audience. One, list the five stories or claims they repeat most often. Two, suggest eight open, specific questions that go beyond those. Three, suggest two story prompts and two process questions. Four, flag any claim in my notes I should verify. And do not add facts that aren't in my notes. The reply flags a revenue figure I'd written down from memory. Good catch. I check it against the company's own announcement, and it was wrong. That's exactly why the last instruction matters.

6:06 Active listening

Now active listening, because the best follow-ups come from listening, not from your list. Listen for energy, emotion and specifics, then dig in there. Allow silence. Guests often add their best thought after a pause. Don't interrupt to show your own knowledge. The episode is about the guest's insight. Summarize and bridge: "so the key was this, which leads to..." And on audio, use verbal encouragement sparingly. Constant "mm-hmm" is distracting. On video, nodding works better. Structure helps too: three or four segments, each with an anchor question and optional follow-ups, plus a signature question at the end.

6:49 Difficult moments

Let's handle difficult moments. A rambling guest: steer gently with "that's fascinating, I want to make sure we get to this." Generic answers: ask for a specific example. A questionable claim: ask politely for evidence or context, and note it for fact-checking. Sensitive topics: agree boundaries beforehand, and don't push for personal details. Nervous guests: start with easy questions and remind them mistakes can be edited. And stay legally aware. Hosts and publishers can be responsible for defamatory statements in episodes. If a guest makes serious allegations about identifiable people or companies, seek advice or edit them out. Frame health, financial and legal advice as general information, and respect off-the-record requests.

7:37 Mistakes + recap

Here are the common mistakes. Reading questions in order without listening. Asking what the guest has answered a hundred times. Talking more than the guest. Letting unverified claims about others go out unchecked. And using AI research output as if it were fact. Let's recap. Research deeply to go beyond standard answers. Use open, specific, story and process questions. Climb the follow-up ladder. Listen actively and follow the energy. Handle difficult moments with care. And use AI to organize research, never to replace verification.

8:14 Try this now

Here's your try this now. Choose a real or imagined guest. Write an interview guide with four segments, two story prompts and two process questions. List the guest's standard answers you'll avoid. Run the AI research prompt on your own notes and verify every flagged claim. Then practice the follow-up ladder in a normal conversation this week. You'll be surprised how much people share. In the next lesson, we focus on you, the host.