YouTube Music Is About to Pitch Your Podcast in a Synthetic Voice: Write a Show Description That Survives Being Read Aloud (2026)
YouTube Music announced Your Podcast Lineup, a weekly AI-voiced preview that tells listeners which shows they'd like and why. A machine will soon describe your podcast to strangers out loud. Here's what's confirmed, why most show descriptions fall apart when spoken, a rewrite pattern and title audit you can do in 20 minutes, and a spoken pitch you control.
1. What Your Podcast Lineup is (and what we don't know)
| Item | What's been said |
|---|---|
| What it does | A weekly spoken AI preview of recommended shows, why you'd like them, then suggested episodes |
| Who gets it | YouTube Music Premium and YouTube Premium subscribers, worldwide |
| When | "Coming soon." No date |
| Announced alongside | Ask Music, a conversational experience across podcasts and more than 300 million songs |
| What the AI reads to write its pitch | Not published |
The last row matters most. YouTube hasn't said whether the preview draws on your description, your titles, your transcripts, your chapters, viewer comments, or all of these. Joycast's reading is that your show description and episode titles are the raw material. That's a sensible assumption, because they're the text you control that describes the whole show. Transcripts may play a part too, which is one more reason to state your show's premise out loud in your episodes, not just in the notes.
It also fits a wider pattern. Spotify, YouTube and others are adding AI layers that summarise, preview and recommend audio for listeners. Each time, the system describes your show in its words, built from what you gave it.
2. Why show descriptions break when spoken
Read your podcast description out loud right now. Most fail in one of these ways:
| Problem | Typical example | What happens when it's spoken |
|---|---|---|
| Keyword stuffing | "Business, entrepreneurship, startups, marketing, leadership, growth, mindset, AI..." | Sounds like a list. Gives the AI nothing specific to say about why someone would like it |
| Generic praise | "The ultimate podcast for curious minds" | Says nothing that's true of your show and not of every other show |
| Host credentials first | "Hosted by award-winning author and keynote speaker..." | The listener hears who you are before why they'd care |
| Symbols and formatting | "🎙️ New eps every Tues | S3 out now → link in bio" | Emojis, pipes and arrows get dropped or read oddly |
| Inside references | "From the team behind TBP, now with more DMs" | Acronyms and in-jokes mean nothing to a stranger |
| Calls to action | "Rate, review, subscribe and share!" | Useless in a preview. The listener isn't subscribed yet |
A recommender that wants to say "you might like this because..." needs a because. If your description doesn't give one, the AI will fill the gap with something generic, or with whatever it can infer from your episode titles.
3. The ear test: hear your metadata before the AI does
The quickest way to know what a synthetic voice will make of your text is to have a synthetic voice read it. It takes five minutes:
- Paste your show description into a free text-to-speech tool and play it in a neutral voice.
- Paste your last ten episode titles and play them one after another.
- Listen without looking at the screen. Note every place you lose the thread, every word that's read oddly, and every line that sounds like any other show.
- Ask a friend who doesn't know the show: "After hearing that, who is this for?" If they can't answer in one sentence, neither can a recommender.
You'll hear problems you can't see. Abbreviations get spelled out. Numbers come out wrong. Clever wordplay that works on a screen turns to mush when spoken. A description that seemed fine reveals it never says what an episode is like.
Run the ear test free
Paste your show description and episode titles into a free SpeakSay tool, listen with the screen off, and fix what sounds wrong before a recommender reads it.
4. A rewrite pattern for the show description
Write the first two sentences as if someone will read them aloud to a stranger who likes something similar. Use four parts, in this order:
- Who it's for: a specific listener, not "everyone curious."
- What they get: the change or payoff, in plain words.
- What an episode is like: format, length and tone, so the listener can picture it.
- One proof: a specific episode, guest type, result or number.
Here's a before and after for a made-up show:
| Before | After |
|---|---|
| "The ultimate podcast for small business owners! 🚀 Marketing, sales, growth, mindset and more. Hosted by serial entrepreneur Dana Reyes. New eps every Tuesday. Rate, review and subscribe!" | "For people running a shop or service business with fewer than ten staff. Each 25-minute episode takes one real owner's problem, like a slow January or a bad review, and walks through what they did and what it cost. Recent guests include a bakery that rebuilt its weekday trade and a plumber who stopped answering the phone at night." |
The "after" version gives a recommender a ready-made reason ("because you run a small shop and like practical stories") and gives a listener a picture of the episode. The host's name can come later in the description, where people reading the full page will see it.
Keep the full description for search and show pages if you like, but make sure the first two sentences can stand alone, because those are most likely to be quoted or paraphrased.
5. Auditing your last ten episode titles
When the preview suggests an episode, the title is what it has to work with. Score your last ten:
| Title pattern | Spoken result | Fix |
|---|---|---|
| "Ep. 142 | w/ @guesthandle" | A number and a handle. Nothing to want | Lead with what the guest knows: "How a nurse turned night shifts into a side business" |
| "Mindset Monday #37" | A slot name, not an episode | Name the idea: "Why your to-do list makes you slower" |
| Inside jokes | Means nothing to strangers | Keep the joke in the episode. Put the topic in the title |
| Very long SEO titles | Too long to say naturally | Under about 12 words, with the payoff near the front |
| Questions you never answer | Feels like bait when spoken | Only use a question if the episode answers it |
A good test is whether a friend could say the title out loud as a recommendation: "you should listen to the one about how a nurse turned night shifts into a side business." If it sounds natural in that sentence, it will sound natural in an AI preview.
6. Make the spoken pitch you control
You can't control what YouTube Music says about your show. You can control what your own channels say. A 30 to 60 second spoken pitch is worth having for:
- the top of your feed as a trailer episode, which is often the first thing a new listener plays
- a pinned Short or Reel
- your website and newsletter
- guest swaps on other shows, as a ready-to-play promo
Use the same four-part pattern as the description, spoken. If you have strong episodes and little time, an AI podcast tool can turn your best episode summaries into a short two-voice pitch that you edit and approve. Our podcast trailer and audiogram playbook covers the trailer itself, and sonic identity covers making it sound like your show.
A word of caution. If your show uses AI voices, say so in your own pitch. Listeners react differently to AI voices in audio and in video. We looked at the data in AI voices on audio vs video podcasts. A recommendation that leads to a surprise erodes trust.
Turn your best episodes into a 45-second pitch
Give SpeakSay's AI podcast tool your strongest episode summaries and get a short spoken pitch you can edit, voice and place at the top of your feed.
7. A 20-minute checklist
- Play your current description through TTS with the screen off. (3 minutes)
- Rewrite the first two sentences: who it's for, what they get, what an episode is like, one proof. (7 minutes)
- Remove emojis, pipes, arrows, acronyms and calls to action from those two sentences. (1 minute)
- Play the last ten titles through TTS and rewrite the three weakest. (6 minutes)
- Play the new description again. If a stranger could say who it's for after one listen, you're done. (3 minutes)
Then do the same for future episodes as part of publishing. It costs a minute per episode.
8. FAQ
What is YouTube Music's Your Podcast Lineup?A personalised AI-audio guide announced on September 23, 2026. It gives a quick weekly spoken preview of recommended podcasts, why the listener might like them, and suggested episodes. It's coming soon for YouTube Music Premium and YouTube Premium subscribers.
What information will the AI use to describe my podcast?YouTube hasn't published this. Your show description and episode titles are the most likely raw material because they're the text that describes the show. Transcripts may also be used, so state your premise out loud in episodes as well.
How should I write a podcast description in 2026?Start with two sentences that work when spoken: who it's for, what they get, what an episode is like, and one specific proof. Keep keywords, host credentials and calls to action for later in the description.
Do emojis in my podcast description matter?On a screen they're decoration. Read aloud by a synthetic voice, they're dropped or read oddly. Keep them out of the first two sentences.
How can I hear what an AI voice will make of my description?Paste it into a text-to-speech tool and listen with the screen off. Anything you stumble over, or any line that could describe any show, needs rewriting.
Should I make a trailer episode?Yes. A 30 to 60 second spoken pitch at the top of your feed is the version of your show's description that you fully control. It's also useful for Shorts, your website and guest swaps.
