How ElevenLabs Builds, Prices, and Grows AI Consumer Apps

ElevenReader removed an entire pricing tier and rewrote plans to be expressed in hours of listening/playback rather than tokens, credits or characters, abstracting away AI cost details for consumers even though real compute costs exist. Separate apps let each app price to its own users (flagship creator app keeps credits).

Web offers upwards of five plans; on mobile the flagship app offers one price because five plans didn't make sense with store limitations and for simplicity. Experiments vary incentives around that single price rather than adding plans. Part of a broader rule to distill the best of web to mobile rather than porting everything.

The team ran many experiments smoothing and optimizing the onboarding flow (quiz, goal setting) following popular blog advice. In hindsight Jack would first test the base case: a signup form straight into the zero-state product, e.g. strip onboarding entirely for 10% of users for five days, to size the incremental retention, engagement and monetization the flow unlocks.

Instead of onboarding quizzes and goal setting, shorten the path to the aha: user brings their own ebook, EPUB or PDF and quickly finds a voice they want to hear it in. Fewer screens, clear about who the app is for.

Run two tracks at once: earned-media launches (about monthly per team, more than once a quarter) targeted to the right channels (all channels, email, LinkedIn or not), plus a steady paid UA engine. Earned attention feeds ElevenLabs branded search, the first paid target, and makes paid ads convert better because people already know the brand.

Launch doc: write the core value promise and secondary value promises tightly, then draft an X/Twitter thread (headline, hook, each sub-tweet leading to the next). Then turn the thread into video, from in-house motion graphics to on-camera content. Head of growth and creative team give feedback on the thread and script. Spend 1-2 weeks refining the story before launch. A feature that can't make a good thread is a red flag.

Run paid UA on Meta, TikTok, Apple Search Ads, Google Play and Google Ads, reviewing results incrementally each week or month per channel rather than paid in aggregate. Because brand launches build awareness, they don't have to optimize for day-0/6/8 conversion or week-one payback and instead test into the right payback or LTV:CAC while testing new creatives.

ElevenLabs hired an internal, on-staff creator to film and create short-form content and has an in-house team making motion graphic videos, so the company owns and tells its launch story rather than relying on App Store featuring.

The two apps share database, MMP and testing tools. UGC made by the same creators/agencies is boosted on paid or posted organically, and what works for one app is picked up by the other. MMP events that carry the most signal are mapped and defined consistently across apps. Retention, engagement and monetization are built separately per app.

From a horizontal platform, pick an ideal persona and form factor and build a dedicated app (creator app vs PDF/ebook reader). Give each app its own brand and look so store searches for 'ElevenLabs' and for 'turn anything into an audiobook' each find the right app, while sharing a design system and tech. More small bets are being incubated.

Both app growth leads test their own paywall hypotheses at the same time and compare notes, roughly doubling or quadrupling the surface area for learning.

About 10-12 pods ('speedboats'), each with its own development, design, growth and product, operating freely; the mobile team is ~12 people. Shared org-wide roles (ops, creative and brand, SEO, web) provide guidelines so pods don't rebrand themselves. Guidance to pods is 'go ship fast'.

Screen engineers for side projects and user-pain obsession by looking at public GitHub profiles; target people from companies with strong product-building cultures; recruit many former founders; referrals from those hires form a flywheel.

Rather than copying Spotify Wrapped, find the product's natural shareable moment. ElevenLabs gave builders on its agents platform a milestone showing how many hours of agent interactions they'd had, similar to YouTube or Stripe milestone awards.

Monitor mentions across Reddit, Discord and X. Found users share screenshots and screen recordings of a chosen voice reading their content (e.g. a famous-sounding voice reading homework). The team is putting more fuel on that; also redesigning the voice orb to feel more emotive.

RevenueCat's in-app share-a-chart features were almost never used. Building a mobile app with an aesthetic, easy-to-screenshot home screen made customers post their numbers on X. Not designed as a sharing feature.

Canva spun out Canva Stories, a separate app focused on Instagram stories (2019-2020). Users wanted Canva itself and saw the spin-out as a subset.

Canva's video product launch had a 12-month run-up, three months of messaging work, a strong video and news articles lined up for one launch day.

After raising B2B prices stepwise ($100, $200, $400 a month up to a couple thousand), churn appeared around $1,000/month and customers cited budget, so the price was dropped to retain them.