Your podcast intro is a contract between you and your listener.
In the first thirty seconds—sometimes less—they decide whether to stay or skip. Whether to listen to the whole episode or delete it. Whether they'll subscribe or forget you existed by next Wednesday. That's not hyperbole. That's listener psychology. Studies consistently show that podcasts lose 50% of their audience in the first five minutes, with the biggest drop happening right at the beginning.
Here's the good news: you can change that. The intros that work—the ones that keep people listening—aren't complicated. They're built on three simple techniques that you can learn and deploy immediately. And AI voices? They actually excel at this, because you have absolute control over pacing, emphasis, and pause placement.
Let me show you how to build an intro that stops people from hitting skip.
The Anatomy of an Intro That Works
A great podcast intro has four distinct parts. Skip any of them, and you lose impact. Structure them correctly, and you're golden.
Part 1: The Hook (0-5 seconds)
This is the promise. What's the benefit of listening to this specific episode? Not the topic—the benefit. Why should they care?
"We're about to reveal why every productivity system you've tried has failed you."
"I spent six months undercover documenting how AI is changing your job, and it's not what you think."
Not: "Welcome to the podcast. Today we're talking about productivity." That gets skipped.
Part 2: The Context (5-15 seconds)
Now you give a little more information. Just enough so listeners understand what they're getting into. This is where you establish authority and credibility. You're basically saying, "I know what I'm talking about."
"I've built three companies, failed twice, and what I learned in the rubble is..."
"My research team spent eighteen months analyzing salary data from seventy thousand employees..."
Short, specific, credible. Move on.
Part 3: The Value Promise (15-20 seconds)
This is the moment where you tell listeners exactly what they'll get from this episode. Not vague value. Concrete value. You'll hear about X, Y, and Z. That's it. Simple. Clear.
Part 4: The Transition (20-30 seconds)
The final beat. You're moving from intro into content. This is where energy picks up slightly. You're saying, "Here we go." Time to listen.
Now, here's where most people go wrong: they read all four parts at the same pace. Same speed, same tone, same everything. Listeners check out because there's no dynamic. No rhythm.
That's where these voice techniques come in.
The Pacing Blueprint: When to Speed Up, When to Slow Down
The secret isn't fancy voices or perfect pronunciation. It's variation.
Part 1 (Hook): Start with energy
Fast pace, high energy. You're grabbing attention. Think 1.1x to 1.2x your normal speed. Not rushed—just engaged. If you're generating with an AI voice, this is where you sound like you actually care about what you're saying.
Recording at 1.15x speed creates immediacy. You sound like someone
with important information, not someone reading the phone book.
Before the Hook Lands: Dramatic Pause
Here's the magic moment. Right after you deliver your hook, you pause. Not a tiny pause. A real pause. 800-1000 milliseconds. Long enough that listeners actually absorb what you just said.
We're about to reveal why every productivity system you've tried has failed you. <break strength="x-strong"/>
This pause does two things. First, it gives your hook time to land. Second, it creates tension. Listeners lean forward. They think, "Okay, I'm listening."
Part 2 (Context): Return to normal speed
Drop back to 1.0x speed. You're providing context now, not selling. Listeners are already hooked. You can relax slightly. Deliver the information clearly, but without the urgency.
Part 3 (Value Promise): Slow down slightly
This is where listeners decide whether they're actually invested. Slow to 0.95x speed. You're being deliberate. You're showing respect for their time by being clear about what they're getting.
SSML helps here. Use emphasis to highlight the actual benefits:
You'll learn how to identify a toxic project before you join it.
You'll discover the one question that separates good bosses from bad ones.
And you'll hear directly from people who've lived this and made the jump.
Each sentence gets delivered with weight. Not rushed.
Part 4 (Transition): Energy back up
Pick the pace back up to 1.05x as you're moving into the actual episode content. Not quite the initial energy of the hook, but faster than the context and value sections. You're saying, "Let's go."
Voice Selection: The Trick Most People Miss
You have 54 voices to choose from in Vois. You'd think everyone would sound good for intros. They don't.
The best intro voices have these qualities:
- Bright and energetic without being annoying — People respond to energy in the first few seconds. But annoying kills you.
- Clear articulation — Technical content or a fast-paced intro will lose listeners if they can't understand you.
- Just a touch of warmth — You want trust. Pure authority sounds cold. Pure warmth sounds fake.
For AI voices, that sweet spot is usually voice blends. Single voices work, but blends give you character + clarity + warmth all at once.
The "Perfect Podcast Host" blend:
af_sarah(65)+af_nova(35)
af_sarah is warm and approachable. af_nova adds brightness and clarity. Together, you sound like someone genuinely excited about sharing something important. Not sales-y. Just... genuinely interested.
Use this blend for your intro at 1.1x speed and watch engagement jump.
If you need authority instead:
am_michael(60)+bm_lewis(40)
American warmth meets British precision. You sound knowledgeable and personable. Perfect for intros about serious topics where you need credibility.
Before/After: What This Actually Sounds Like
Let's get concrete. Here's a podcast intro without any of these techniques:
Welcome to the podcast. Today we're discussing AI in the workplace.
I'm joined by experts who work in the field. We talk about how AI
is changing job responsibilities and what that means for your career.
We also cover salary negotiations, staying relevant, and the future
of work. Let's get started.
Single voice, single pace. Technically correct. Emotionally flat. Listeners skip it.
Now, same content with structure and voice techniques applied:
Script with SSML and pacing notes:
<prosody rate="1.15">Your job is about to change. Probably this year.</prosody>
<break strength="x-strong"/>
I've spent the last six months talking to hundreds of people in tech,
finance, and creative industries. <break strength="medium"/> And what
I learned might actually surprise you.
<prosody rate="0.95"><emphasis level="strong">You'll hear exactly what's
happening at companies hiring AI right now.</emphasis></prosody>
<break strength="weak"/>
You'll learn the skills that make you indispensable.
<break strength="weak"/>
And you'll hear from people who've already navigated this shift.
<prosody rate="1.08">Let's dive in.</prosody>
Same information. Completely different impact. The first version sounds like someone reading notes. The second sounds like someone who knows something you need to hear.
The pacing variation creates rhythm. The strategic pauses let ideas land. The speed changes match your emotional intent. The blend voice carries warmth and authority simultaneously.
The Technique: SSML for Intros
Here's the SSML you'll actually use:
For speed/energy control:
<prosody rate="1.15">Fast part here</prosody>
Rate values:
1.2= 20% faster (high energy)1.1= 10% faster (slightly energetic)1.0= normal speed (neutral)0.95= 5% slower (deliberate)0.9= 10% slower (serious)
For emphasis:
<emphasis level="strong">This matters.</emphasis>
Emphasis levels: weak, moderate, strong
For strategic pauses:
Breaking news. <break strength="x-strong"/> Here's why it matters.
Pause strengths for intros:
weak= ~250ms (between list items)medium= ~500ms (after sentences)strong= ~750ms (emphasizing importance)x-strong= ~1000ms (dramatic moments, topic shifts)
Pro tip: Put pauses after your hook, before major value statements, and after the context. Skip them in the middle of sentences—that's where speech sounds choppy instead of intentional.
Real Intros: How the Pros Do It
Here's a non-fiction podcast intro structure that actually works:
<prosody rate="1.12">There's a decision you're going to make this year
that shapes the next decade of your life.</prosody> <break strength="x-strong"/>
I'm Sarah, and this is the career podcast. <break strength="medium"/>
Over the last five years, I've interviewed hundreds of people who made
major career transitions. <break strength="strong"/>
<prosody rate="0.95"><emphasis level="strong">Today, you'll hear from three people
who left corporate jobs to start companies, got fired, and why they're
grateful it happened.</emphasis></prosody> <break strength="weak"/>
You'll also learn the one thing they all did before making the jump
that saved them from disaster.
<prosody rate="1.08">Let's start with Marcus.</prosody>
This intro:
- Hooks immediately (fast, energetic)
- Pauses to let the hook land
- Drops to normal speed for context
- Slows down for clarity on the actual episode content
- Picks speed back up as it transitions
Result: listeners are in. They're curious. They're going to hear the episode.
Here's a true-crime podcast intro:
<prosody rate="1.15">The witness testified that she saw someone leaving
the building at 11:47 PM.</prosody> <break strength="x-strong"/>
Everyone believed her. The jury believed her. Even the defendant
believed her. <break strength="strong"/>
<prosody rate="0.93">But she was wrong. Completely wrong. And that mistake
cost an innocent man eleven years of his life.</prosody> <break strength="medium"/>
<emphasis level="strong">This is the story of how we convicted the wrong person,
and how DNA evidence finally proved it.</emphasis>
<prosody rate="1.05">The evidence was sitting in a warehouse the entire time.</prosody>
The pacing creates tension. Listeners are hooked because you've established a mystery in the first fifteen seconds. They have to know what happens next.
The Voice Blend Recommendations for Different Genres
Not all intros need the same voice.
News/Current Events: bf_emma(55)+bf_alice(45) — professional, neutral, credible
True Crime: af_heart(60)+bf_lily(40) — intimate but authoritative (draws listeners in)
Business/Finance: am_michael(70)+bm_lewis(30) — serious, knowledgeable, trustworthy
Comedy/Light: af_nova(65)+af_sky(35) — bright, energetic, fun
Self-Help: af_sarah(60)+am_adam(40) — warm, approachable, credible
Storytelling: af_heart(55)+bf_lucy(45) — emotional, compelling, intimate
The pattern is always the same: pick a voice that carries the emotion you want, then add a second voice that adds clarity. The blend does the heavy lifting for you.
One More Thing: The Autosave Technique
Once you've built an intro that works, save it as a template. Don't rebuild it for every episode.
Create a script in Vois with your intro structure already in place. Your hook at 1.15x speed, your pause in the right place, your SSML already written. Then for each episode, you just swap out the specific content while keeping the structure.
Consistency matters for podcasts. Listeners recognize your intro because it's the same format every time. But the content changes, so it doesn't get stale. That's the sweet spot.
The Truth About Intros
The best podcast intros aren't about fancy production or expensive voices. They're about understanding that the first thirty seconds are everything.
Structure your intro in four parts. Vary the pacing—energy at the start, pause to let it land, deliberate speed for important information, quick pace as you transition. Choose a voice blend that matches the emotional tone of your show. Use SSML strategically for pauses and emphasis.
Do that, and you're not just keeping listeners past the thirty-second mark. You're making them want to stay. You're making them curious. You're building the kind of show people actually look forward to.
That intro is your contract with your listener. Make it count.