Sub Club Podcast Episode Summary

Episode Overview

Title: How ElevenLabs Turns Feature Launches Into a Growth Engine
Date: March 9, 2026
Guest: Luke Harries, Head of Growth at ElevenLabs
Hosts: David Barnard, Jacob Eiting

In this focused episode, David Barnard interviews Luke Harries to explore how ElevenLabs transforms every new feature launch into a powerful growth engine. The conversation delves into the company’s systematic approach to feature launches, their massive paid ad spend, practical uses of AI in both creative production and localization, and the evolving role of marketers as AI creative directors.

Key Discussion Points & Insights

1. The Growth Engine Strategy

[01:40–02:56]

• Building Growth Engines:
  Luke describes growth as a repeatable, optimizable system—essentially an “engineering approach to marketing,” with ambitious goals and constant iteration.

• Feature Launches as Heartbeat:
  Major product releases unlock new capabilities and viral moments. Product and growth teams work closely so that each feature launch gets full marketing support: demo videos, tweet threads, world-class landing pages, and updated ad creative.

• AI's Unique Viral Advantage:
  The constantly evolving capabilities of AI give the company frequent “what’s new” moments that attract attention beyond their existing user base.

  “The heartbeat of the growth team is working hand in hand with our incredible product and research team, such that when they're launching new features...we put it through this growth system and that normally translates into...a really viral moment.”
  —Luke Harries [02:09]

2. Maximizing Launch Moments Across Channels

[02:56–05:19]

• Beyond Just Earned Media:
  David emphasizes that many companies undervalue the broader opportunities in a feature launch—most barely update existing users, much less attract new ones. Luke agrees, noting the importance of leveraging each launch for both earned and paid attention.

• Automating & Optimizing Ad Copy:
  ElevenLabs built a custom GPT tool that analyzes top-performing ad copy and rewrites announcements (e.g., X posts) into high-converting ad variants for different platforms.

• Asset Variants for Conversion:
  Rapid generation and testing of different image, video, and static ad variants help pinpoint what drives the best results.

  “We built a custom GPT where we've analyzed what are all the top performing meta in Google Copy...and then we pass it through this custom GPT which will then iterate on it...”
  —Luke Harries [04:38]

3. Creative Production: AI, UGC, In-House, and Agencies

[05:19–08:33]

• The Creative Mix:

  • Start with agencies for format testing, then optimize and scale in-house.
  • Two full-time “AI creative producers” experiment daily with tools and trends.
  • Most core ads use high-quality motion design with text-to-speech and 11Labs’ music models for rapid hook and variant testing.
  • AI-generated images are integrated to boost volume and quality.

• Use of UGC and Educational Video:

  • A dedicated in-house creator (Alec) produces educational, “person-in-front-of-camera” content for YouTube and other channels.
  • While avatars and advanced AI creative have future potential, quality remains paramount, so studio-style educational videos are still preferred.

  “At the moment, what we found works really well for our core app ...is they create with after effects really great motion design...layer in with like text to speech voiceovers using 11 labs with the music. And that enables you to do like rapid testing…”
  —Luke Harries [06:26]

4. Spending $100M+ in Paid Ads: Strategy & Measurement

[08:33–13:03]

• Channel-Specific Growth Engines:

  • Google Search remains the biggest channel; focus is on creative, localized copy and high-conversion landing pages.
  • Meta and TikTok: Quality and quantity balance, with an increasing focus on segmentation and personalization using AI.

• Localization via AI:

  • ElevenLabs switched from agencies and SaaS tools to a fully custom, LLM-powered localization system, automating and accelerating copy adaptation for 70+ languages.
  • The in-house system parses every PR, autotranslates strings, and localizes ads as well as product content.

• Bottom-Up, Incremental Approach:

  • Teams look for bottlenecks and scale channels stepwise (10–20% at a time).
  • Continuous focus on funnel optimization and localization.

• Incrementality and ROI:

  • Direct channel ROI drives most decisions for self-serve/web products.
  • Regular lift studies (A/B with holdout groups) correct for conversions that would have happened anyway.
  • Platform data is adjusted for true incrementality before allocating new budget.

  “We treat each channel as its own growth engine...we're obsessed of, okay, how do we make really great ad copy, how do we localize it...and then it's like, okay, week over week, how do you scale it 10%, 20%...”
  —Luke Harries [09:13]

  “We initially went down the route of ...an agency to then who have people to localize it. But I kept on having our team just sending screenshots of ChatGPT...So we actually ripped out all the localization infrastructure...just sends it with like a prompt per language about the localization guide.”
  —Luke Harries [10:25]

  “The important thing to do is to run the lift studies within the platform which actually do an A B test with holdout groups to tell you... actually there was only an incremental 200 because 300 people converted anyway.”
  —Luke Harries [12:20]

5. The Future: Marketers as AI Creative Directors

[13:03–16:16]

• Emergence of the AI Growth Partner:
  Luke predicts that future marketers will operate as “AI creative directors,” leveraging agents that understand launches, competitive landscapes, and personas, and autonomously generate core messaging, storyboards, and localized creatives.

• Dream Workflow:
  Set direction, taste, and priorities—then AI agents handle storyboarding, motion design, language adaptation, and channel-specific variants. The marketer/“growth person” edits, approves, and adapts creative, much as a director would.

• Solo Builder Revolution:
  Even solo founders could orchestrate full launch engines—shipping code, triggering AI-powered announcement processes, and generating multi-channel, multilingual campaigns with minimal overhead.

  “Probably each product will have a growth person who has really great taste and really great ideas. And then you work with the engineering team and you go, okay, I'm going to launch your product...you then chat with them agent ...It reads through all your internal docs and knowledge base and your Personas, helps you craft like the core of the messaging... helps storyboard out what does this launch video look like...”
  —Luke Harries [13:24]

  “The ideal is probably you wake up every single day and you have 10 to 20 high performing ads…you're that creative director of where we're going and apply your taste. But then you're like, yes, I like these. No, I don't like these...Or these need to be tweets...So this is, yeah, ideally what we're going to build end-to-end…”
  —Luke Harries [14:13]

  “Potentially the first billion dollar star suck could be built by that.”
  —Luke Harries [15:36]

Notable Quotes & Memorable Moments

• "The heartbeat of the growth team is working hand in hand with our incredible product and research team…" —Luke Harries [02:09]
• "You don't know until you try. And I think more apps should be trying." —David Barnard [03:09]
• "We built a custom GPT... to iterate on it into like the high performing meta and Google Copy." —Luke Harries [04:38]
• "At the moment what we found works really well [is] after effects motion design... layer in with like text to speech voiceovers using 11 labs." —Luke Harries [06:26]
• "We treat each channel as its own growth engine... how do you scale it 10%, 20% for Meta, it's a similar one where yes, you want great amount of creative, but you also want really high quality." —Luke Harries [09:13]
• "If the LLMs are way better at translation than these agencies and expensive tooling...we actually ripped out all the localization infrastructure." —Luke Harries [10:25]
• "The ideal is probably you wake up every single day and you have 10 to 20 high performing ads..." —Luke Harries [14:13]
• "Potentially the first billion dollar star suck could be built by that." —Luke Harries [15:36]

Key Timestamps

• [01:40] Growth as a repeatable, engineering-driven system
• [02:56] Turning feature launches into marketing events for users and the wider public
• [04:38] Automating and optimizing ad creative with a custom GPT
• [05:57] In-house AI creative production, evolution from agency to internal team
• [09:01] Strategic deployment of $100M+ ad spend; bottoms-up channel optimization
• [10:25] Radical shift to full AI-powered localization
• [12:20] ROI, lift studies, and measurement approaches
• [13:24] Marketer as AI creative director: the future of growth
• [15:36] Solo builder possibilities and the next billion-dollar start-up

Hiring and Call to Action

[16:16]

• ElevenLabs is hiring lots of mobile engineers and encourages listeners to try their 11Creative products and mobile apps for their own growth and performance marketing needs.

Conclusion

This episode delivers a masterclass in how ElevenLabs turns feature launches into sustainable growth engines by tightly integrating product and marketing, leveraging AI for creative production and localization, and constantly experimenting and measuring for incrementality and true ROI. Luke Harries leaves listeners with a vision of the near-future marketer—an orchestrator of AI agents, able to execute launches at a speed and scale never seen before.