Two weeks ago, I hosted the first-ever in-person Creator Spotlight event with our partners at ElevenLabs. It was a blast!
We had room for 50 people and received over 900 RSVPs. Which blew me away. There are more than 430,000 of you on our email list these days, which is an incomprehensible number. Those of you who showed, thank you. We’re looking to throw more get-togethers in the coming year — in New York and beyond.
In today’s newsletter, I’m sharing what I learned from this event and my preparatory research about how creators are using new-gen AI audio tooling to increase reach and revenue.
— Francis Zierer, Lead Editor
P.S. This issue is coming to you outside of our usual Tuesday/Friday schedule because it’s part of our partnership with ElevenLabs. Our partners there asked me to collect my thoughts after our event; they were not otherwise granted editorial control over the specifics of this issue.
Introducing the most emotive voice model ever
From the emotion in a single line to the voice that carries a whole story, ElevenLabs’ latest voice models, Eleven v4, give creators more control over how their work sounds. The result is audio that feels performed, with way more ways to make it your own.
Now you can add direction for pacing, delivery, and reactions right in your script. Clone your voice from just 10 seconds of audio, keep it consistent across longer projects, and create for audiences in 90+ languages. That could mean a narrated newsletter, a character-driven video, or an audiobook ready to find global listeners far beyond your usual reach.
This is an advertisement.
Your voice is an asset
There is always a panel, but the best part of events like the one we threw with ElevenLabs is meeting people, learning about how they see and shape the world.
The panel shapes those conversations, and the title for our panel was The New Reach: How creators are growing with AI audio tooling.
I moderated a group of three:
Dara Denney, a creator, creative strategist, and Meta ads expert.
Ryan Bast, who leads creator education and enablement at ElevenLabs (and previously held similar roles at LinkedIn and TikTok).
Tyler Denk, co-founder and CEO of beehiiv, and creator of the Big Desk Energy newsletter.
And from our conversation, three themes emerged:
The increasing demand for a text-to-speech option on written content
Dubbing videos from one language to another to capture growing international segments of a creator’s audience
Voice cloning as a form of insurance and extension on a creator’s voice (important for anyone working in audio or video)
The panel, mid-conversation. Credit Connie Yao, our partner on the ElevenLabs team.
Your readers want to listen
Have you ever listened to an article or newsletter issue instead of reading it?
Here’s a first: you can listen to this issue in audio form instead of reading it. If you’d like to try, go to the web version. You’ll find the audio player at the top of the article; click play, and you can go about your other tasks as you listen.
I recently read an article in the Columbia Journalism Review about how major news sites have been using AI narrators. Two details stood out:
The New Yorker partnered with ElevenLabs as early as 2023 to offer this service. Some 20% of the magazine’s subscribers now listen to these narrated stories, per CJR.
A spokesperson from the Wall Street Journal said that, in the past year, their audio-article feature was used “about five million times, with a 65 percent completion rate.”
If you write a newsletter, you can give your audience the option of listening instead of reading, and people who may not have been readers may become listeners.
The New Yorker and Journal are proof. You can both grow and better engage your audience by giving them the option:
For writers using beehiiv: Since beehiiv has a native text-to-audio feature (using ElevenLabs on the backend), you can do this all in the editor for any given post. That’s how I set up the audio for this post. Go to the “Web” tab at the top of the editor and scroll down to the “Audio newsletter” section. (Find a step-by-step guide in the beehiiv knowledge base.)
For writers using any other platform: generate your audio directly in ElevenLabs and embed their Audio Native player, which plugs into WordPress, Ghost, Webflow, Squarespace, Wix, and Framer.
Tyler also pointed out a less obvious benefit at our panel: beehiiv uses ElevenLabs to generate transcripts for the podcasts it hosts (example). That transcript is an SEO asset. When I Google an older Creator Spotlight interview, the podcast page, with its higher word count and repeated mentions of the guest’s name, often ranks first.
Localization, localización, localização
What would you do if you found out somebody was taking all of your videos and re-uploading them with Brazilian-Portuguese dubbing? If these bootleg Brazilian versions of your content were receiving thousands of views?
Surely those views should be yours, right?
Back in January of this year, this is exactly what happened to Kate Bast, wife of our panelist Ryan Bast, many months before he joined ElevenLabs.
Kate is a DIY and parenting creator with over 1.3 million followers across Instagram and TikTok. Ryan discovered Brazilian and Mexican pages taking her videos and dubbing them — to successful performance — and began investigating ways to capture that attention back.
ElevenLabs’ Dubbing feature would enable Kate to create Portuguese and Spanish translations of her content in a voice that is recognizably hers, without speaking the languages herself.
I took the intro from one of our recent podcast episodes and ran it through the Dubbing feature, in Brazilian Portuguese, of course. To my English-language ears, it sure sounds like me speaking the language:
Panelist Dara Denney had a similar discovery to Kate — her videos were being taken and dubbed in Russian. I’d expect any video creator with a sizable audience and valuable content to run into this issue.
Of course, it’s not an issue — it’s demand for your content in markets other than those you’re able to reach on your own. Here’s the solution:
Identify in which languages your stolen, dubbed content is getting the most traction.
Study how your content is being ported over — straight translation in both the caption and video? Great. Is the caption written differently, using specific hashtags? Great, take note.
Take note of which videos are most popular; these may be different than your most popular English-language videos.
Create new channels for each language (e.g., @creatorspotlightbrasil).
Where possible, work with the platforms you’re using to get the unlicensed translated channels taken down.
Use ElevenLabs’ Dubbing feature, which keeps the distinct qualities of the original speaker’s voice in an audio file while porting it to another language. (Dubbing translates into 90+ languages.)
If you succeed in building up a sizable audience, you now have another asset to bring to specific advertisers looking to reach that country or language group.
(Dara has also put this to work at her agency: for one client, she used ElevenLabs to give an English voiceover an Italian accent for ads aimed at a European audience. It outperformed the plain-English version.)
Voice cloning is uncannily effective
I don’t remember the last time I was spooked as much by the efficacy of a tech product as I was by ElevenLabs’ voice-cloning tool.
After spending maybe 30 seconds reading the script they provide to initiate the cloning, and just a few more seconds sitting and waiting, I had a ready-to-go Francis voice.
It freaked me out. I deleted the clone and tried again, this time altering my voice as I read the script — low pitch, closing my throat so my voice would come out less recognizable — so I could listen back to a new clone without having to confront the uncanny nature of listening to myself-who-is-not-myself speak back to me.
Didn’t work. The diction was a bit different, but the rhythm, the breath — unmistakably me.
There’s a marketplace for this, by the way. Ryan mentioned the ElevenLabs Voice Library program, which has paid out over $22 million on over 17,500 submitted voices.
During our panel, I learned how creators use their voice clones. Two examples:
Recording ad reads when sick. Let’s say you’ve come down with bronchitis while on deadline for a sponsored partnership that requires you to record a segment on your podcast. You do not have time to wait until you get better — use your voice clone. (I had to record an ad earlier this week while under the weather; I ended up having to re-record the next day due to a scratchy voice producing unacceptable outputs.)
Making edits to pre-recorded audio when you do not have access to high-quality recording equipment. I do not bring my podcast mic with me on vacation.
These functions are what I would call voice insurance.
Then there’s what I’d call voice extension.
Former Spotlight guest Matt Kiser is a solo creator with an almost-daily aggregation newsletter. For years, he paid a voice actor to create a podcast version each day. Ultimately, he couldn’t afford it; his is a donations-funded business, and he has a family to take care of.
Instead of getting rid of the podcast entirely, he cloned his voice so that “it presented as me and I was responsible for it,” to quote my conversation with him earlier this year.
One part of Matt’s story I often think about — some of his subscribers were vocally angry about his using AI to produce the podcast. Long story short, a block of readers unsubscribed and were loud about it on BlueSky.
Nobody who unsubscribed had ever donated a cent to his business — again, the only way the business makes money. And the podcast listenership? Didn’t drop at all.
