How do you prove you're the expert in a podcast interview?

Why do smart experts sound like everyone else on podcasts?

Because they talk about the concept and stop there. They explain what's wrong, share what they've found, and never get specific enough for a listener to think, oh, this person has actually done this.

Here's the thing. Your competition can talk about the concept too. Anybody can read the research and repeat it into a microphone. What they can't do is tell the story of the client they walked through it, with the numbers to back it up.

I said this to a client in a prep session this week, and I'll say it to you: "Too many of these conversations happen where folks like to talk about what is wrong, what they found, the research, all the things, and they don't give any real, tangible next steps or solutions."

Listen, I get why it happens. You know your material so well that the proof feels obvious to you. The listener doesn't know you yet, though. If you don't show them the evidence, they have no reason to believe you're different from the last guest who said something similar.

For consultants who built the business on referrals, that gap is expensive. The people who already know you never needed the proof. Strangers do, and strangers are who you're trying to reach.

Why does proof matter more now that buyers ask AI?

Because getting mentioned and getting recommended are two different results, and AI tools are making the gap easy to see. A study Omniscient Digital published on November 6, 2026 looked at 8,392 AI responses to 300 buying prompts across ChatGPT, Claude, Google AI Mode and Google AI Overviews.

The finding that stopped me: "Models named at least one brand in 97.0% of answers, but directly recommended one in only 47.6%." And even when a model described a brand positively, 72.9% of those positive mentions still stopped short of a direct recommendation.

Read that again. Getting mentioned is the floor. Getting recommended is where the money is.

So when AI is looking for the top five, top 10 in your niche, are you in there? And when you are, has it seen enough proof to actually recommend you?

Every interview you do turns into written material: a transcript, show notes, an episode page, plus the clips and articles you make from it. If what you said on air was vague, the written version is vague too. If you told a specific client story with a real number in it, that specificity lives on in text, in places buyers and AI tools can read. (I wrote more about how that works in why ChatGPT cites your website without saying your name.)

That's what I mean when I call podcast guesting authority architecture. Every episode becomes searchable expertise, and the quality of what you say decides how much that episode is worth.

What should you prove in the first two minutes?

Prove that you're worth staying for. A podcast is a long format, and you still have to earn the first 30 seconds.

Here's how I put it to a client getting ready for two interviews: "podcasts, yes, it's 40 minutes, 30 minutes, an hour, whatever the case is, but you have the same responsibility as someone who is scrolling the internet to hook them in that first 30 seconds, or they're gone."

Not everybody makes it to minute 30. So the things you most want a buyer to remember have to show up early, in your introduction and in your answer to the first question (which is almost always some version of "how did you get into this work?").

When we practice an intro in prep, I put a two-minute timer on, because after two minutes people stop listening. Inside those two minutes, hit three beats:

1. A personal moment the listener can picture. Skip the resume and tell them about the day something had to change.
2. The name of your idea. If you built a framework, a process or a program, say its name out loud from the very beginning. My note to that client was simple: you came up with a particular thing, so name it.
3. Who you serve now, framed for this audience. Describe the people you help in words this host's listeners would use about themselves.

Once it's written, practice until you can perform it instead of reading it. The words can be perfect and still fall flat if they sound recited. Pick your spots to slow down, to lift your voice, to pause.

How do you tell a story that proves you know your stuff?

Tell it in three parts: the problem, what you did, and the payoff. That's the shape I coach clients on before interviews, because it keeps a story from wandering and makes sure it lands on a result.

Here's how I explained it in a prep session. The problem is "what was the thing that they were struggling with... that's like one or two sentences and we're done." Then you're "talking about the ways in which you went about helping the situation, what did you actually do, and then talk about the transformation, which is the payoff."

One or two sentences on the problem. That's it. Keep the setup short so you have time for the part that proves something.

Then make the people in your story specific. When a client told me a story about a colleague, my note was to give it more meat: say she's a senior leader, say she runs a department across several locations. Specificity is what lets the right listener recognize themselves.

This is the Story in the Bankable Authority™ framework (Stand, Story, Sell). Your Stand is the belief you own. Your Story is the evidence that it works. Sell is the bridge from that evidence to your offer, and it only feels natural when the evidence came first.

Stories are also what people hold on to after they stop listening. As I said in that same session, "what they'll remember are those stories." Your job is to tie the memorable moment to a transformation they want.

Should you quote statistics in a podcast interview?

Yes, as long as they're real and you can say them naturally. A story makes people feel the problem. A statistic tells them your story is part of a pattern.

When I was prepping a client who knows her research cold, I told her: "as you're telling me stories, and you're incorporating the knowledge that you have in terms of like, literal statistics, and studies... it is going to be, oh, this person gets it."

That blend is what separates the expert from someone who simply figured something out for themselves. And honestly, most guests skip it. In my words to her: "most people cannot do that, and most people do not do that. And that's how you stand out."

A few ground rules we use:

• Use numbers from studies you can name, or from your own client work. If you can't say where a number came from, leave it out.
• Match the depth to the show. On a broad, general-audience show you don't need the ROI math you'd bring to a business podcast, and the studies are still relevant.
• One or two numbers per answer is plenty. More than that starts to sound like a lecture.
• Pair every number with a story, so the listener has a person to attach it to.

Then practice. Blending stories and stats is hard. As I told her, "this is why you have to practice to make sure that it actually comes off non-salesy."

What should listeners walk away with?

One thing they can do. A tip, an exercise, a question to ask themselves. Here's how I described the listener's side of it in prep: "I'm pulling out my notebook. Right? And now I'm engaged."

When a client finished a strong story in prep, my note was simple: "give a clear tip, something that someone can try... because that would be the clip." Some hosts even recap the takeaways at the end of the episode, so hand them good ones to repeat to the audience that already trusts them.

Then answer the question that isn't being asked. The host may never say, "what would you tell someone who's doing this right now?" Answer it regardless.

Last, make your next step one step. One link, one thing. In prep I put it this way: "It needs to be, hey, go to this link and get the thing. That's it." Use a unique short link for each show, so you and your team can see which interviews sent people your way.

What happens when an expert's message carries proof?

They get chosen faster, and the conversations they get are better. Hosts notice too. A client of ours prepped exactly this way for a large show, and afterward the host told her she doesn't usually ask guests back, then invited her back next quarter. Hosts of that size don't do that just because.

Two more examples from our client results:

• Dr. Erin Wilson closed multiple five-figure contracts within her first three months with us, plus more book sales and a spike in organic email signups, directly from podcast guesting.
• Kisha Wynter: we revamped her messaging and repositioned her in B2B consulting. Prospects stopped lumping her in with leadership coaches, her pitches got faster responses, and she was featured on the SHRM podcast.

What can you do about it this week?

Pick your next interview (or the one you most want to land) and build one proof story before Friday:

1. Write the client's problem in two sentences.
2. Write what you did, in plain steps.
3. Write the payoff, with a real number from that client or a study you can name.
4. End with one tip a listener could try tomorrow.
5. Say it out loud with a two-minute timer running. If it runs long, cut the setup first.

Do that once and you have a story you can bring to every show on your calendar, adjusted for each audience.

Frequently asked questions

How do you show expertise on a podcast without sounding salesy?

Lead with a specific client story, back it with a real statistic, and end with a tip listeners can use. When the evidence does the work, you never have to tell people you're an expert. Practice the blend out loud so it sounds natural.

How long should a podcast guest introduction be?

Keep it under two minutes. Cover a personal moment the listener can picture, the name of your framework or process, and who you serve, described in this audience's words.

Should I use statistics on a podcast?

Yes, one or two per answer, from studies you can name or from your own client results. Pair each number with a story so it sticks, and match the depth to the show's audience.

Does what I say on a podcast affect AI search?

Episodes become transcripts, show notes and episode pages that buyers and AI tools can read. A November 2026 Omniscient Digital study of 8,392 AI responses found models named a brand in 97.0% of answers but directly recommended one in only 47.6%, so specific, provable expertise matters.

Keep reading

• Why you need a signature story as a podcast guest
• Why ChatGPT cites your website without saying your name
• How The Citable Expert™ Engine works