ElevenLabs Pitch Deck

The deck and the round

Round

Pre-seed

Raised in the round

$2M

Year

2022

Investors

Credo Ventures, Concept Ventures

Slides

15

Category

AI

Headquarters

London, UK

Founded

2022

Team today

1,001–5,000 employees

Total funding to date

$792M

ElevenLabs’s pitch deck, slide by slide

1. 1

   Eleven

   Cover

   The company's first name, Eleven, and the line "Powering content in any language with automatic dubbing": dubbing, not text to speech, is the product.

2. 2

   Introduction

   Problem

   People want to watch and listen in their own language, and dubbing, where a human voice actor re-records the dialogue, is how that happens today: about $100 per minute once actors, post-production and studio time are paid, and at least two weeks for a 10-minute video.

3. 3

   Problem

   Problem

   One sentence: there are no affordable tools to make content watchable in any language with high quality.

4. 4

   Solution

   Solution

   Human-quality automated dubbing as a SaaS, on three points: dubbing in the speaker's own voice, from a deep-learning model that preserves it across languages; output that keeps the original emotions and intonation, trained on thousands of hours of professional dubbing; and a one-click, end-to-end product with an optional human in the loop.

5. 5

   Prototype deep-dive

   Product

   The prototype already built, step by step: an English video in, subtitles from speech recognition, translation, separating dialogue from background noise, generating the voice in the new language, and the dubbed video out. A demo video sits beside it, and the headline number: a 10-minute video dubbed in 2 minutes.

6. 6

   Team

   Team

   Two founders who "have studied, lived and worked together" and have been best friends since high school: Piotr Dabkowski, CTO, ex machine learning at Google, Cambridge and Oxford, with a NeurIPS paper and the Js2Py open-source library; and Mati Staniszewski, CEO, ex deployment strategist at Palantir, Imperial College mathematics, with experience at BlackRock and Opera Software.

7. 7

   Vision

   Vision

   "Eleven's automatic dubbing will power seamless communication and content across any language", with the expansion path from the initial focus on creators: offline voice generation for games, audiobooks and podcasts, professional dubbing for feature films, localization and advertising, then real-time voice conversion and real-time dubbing of calls and video.

8. 8

   Expanding total available market

   Market

   Three circles: $4.6B a year as the TAM for professional creators across podcasts and video, $2B spent a year on game localization and movie dubbing, an industry Eleven expects to disrupt, and $24B for localization, translation and interpreting overall.

9. 9

   Market size deep-dive: content creators

   Market

   A bottom-up funnel: 50M+ content creators worldwide, 2M professional creators, 100K YouTube channels with over 500K subscribers, and 10K creators who already upload captions as the immediate market. At three 10-minute videos a month dubbed into three languages, that's 9M minutes a month, or $110M a year at about $1 per minute, with the real pricing a subscription with a base set of minutes.

10. 10

    Our start: content creators

    Go-to-market

    Why creators first: MrBeast's English channel had 96M subscribers, and the Spanish channel he started in 2021 with professionally dubbed videos had 19M, with one video earning about $50K. The takeaways: creators will copy that model, they need speed more than perfect quality, and their volume of data builds the speech datasets that make Eleven defensible.

11. 11

    Traction & feedback

    Traction

    Redacted in the public version.

12. 12

    Competition

    Competition

    A two-axis map of dubbing quality against accessibility and speed: traditional dubbing with voice actors is high quality but slow, text-to-speech tools are fast but not human quality, semi-automated tools still need heavy manual work, and Eleven sits alone in the high-quality, quick, low-intervention corner.

13. 13

    Competitive advantage: research

    Competition

    The technical edge: instead of plain text to speech, the model takes both the speech and the text and generates speech in the new language, with prosody (emotion and intonation, mapped across languages) and the speaker's voice modelled separately. The claim is that it's quick, affordable and easy to extend to new languages.

14. 14

    Timeline

    Other

    Redacted in the public version.

15. 15

    Closing cover

    Other

    The cover again: Eleven, powering content in any language with automatic dubbing.

The numbers in the deck

Cost of human dubbing · slide 2

~$100 per minute

Time to dub a 10-minute video by hand · slide 2

2+ weeks

Time to dub it with the prototype · slide 5

2 minutes

TAM for professional creators · slide 8

$4.6B a year

Localization, translation and interpreting market · slide 8

$24B

Creators worldwide → immediate market · slide 9

50M+ → 10K

Revenue at ~$1 per minute · slide 9

$110M a year

What stands out

1. 1

   The first product was narrower than the company. The deck pitches automatic dubbing for YouTube creators; the beta launched with the round was text to speech, and ElevenLabs became a voice platform. The vision slide had already listed audiobooks, games, real-time voice and calls as the expansion path.

2. 2

   The problem slide is one sentence. Slide 2 does the work with two numbers, about $100 a minute and two weeks for a 10-minute video, and slide 5 answers them with a prototype that dubs the same video in 2 minutes.

3. 3

   The market is sized from the bottom up. Slide 9 goes from 50M creators to the 10K who already upload captions, then to minutes a month and $110M a year at $1 a minute, so the reader can check every step.

4. 4

   One example carries the go-to-market. MrBeast's Spanish channel, 19M subscribers from professionally dubbed videos, shows creators already pay for dubbing and earn from it.

5. 5

   A two-person team, sold on depth. Machine learning research at Google and a NeurIPS paper on one side, Palantir deployment on the other, plus a decade of friendship, is the whole team slide at pre-seed.

Sources

• ElevenLabs: ElevenLabs raises $2M pre-seed and announces AI speech platform (Jan 2023)
• VoiceBot.ai: Synthetic speech startup ElevenLabs raises $2M (Jan 23, 2023)
• Best Pitch Deck: ElevenLabs pitch deck, where the PDF was published

Frequently asked questions