Lovable

📌 Snapshot: 0-1 · AI app builder · Stockholm · Open-source repo (Jun 2023) → two failed launches (2024) → rebrand and relaunch (Nov–Dec 2024) → ~$10M ARR in 60 days, $100M ARR in ~8 months, ~$400M ARR by late 2025 with ~100 people (company-reported).

Problem statement

Software creation was gated by engineering skill. Anton Osika's bet: the 99% of people who can't code should be able to describe an app and ship it. The harder GTM problem was proving that LLMs could actually do this, in public, before anyone believed it.

GTM stack & growth engine

• Open source as proof and audience. gpt-engineer (June 2023) became one of the fastest-growing GitHub repos of the year (50K+ stars). It proved the idea and built a developer audience before a company existed.
• Fail fast, relaunch loudly. Two commercial launches in 2024 under the GPT Engineer name stalled. The team fixed reliability (agent getting "stuck", large codebases), then rebranded to Lovable to signal a new product and audience. The relaunch hit #1 on Product Hunt and Hacker News on the same day.
• Generous free tier. Elena Verna's rule: give AI away. Credits go to free users, hackathons and events; it is treated as a marketing line, not an OpEx problem.
• Performance marketing is not a meaningful lever. Category creation means demand has to be created, not captured.

Content loops

1. Ship → post → users build → users share builds → more users. Shipping velocity is the #1 dev value, and building in public is how all that shipping gets distributed.
2. Every employee is a channel. Founders, growth, design and engineers all post product decisions, milestones and failures.
3. User builds as social proof. What people make with Lovable (and the traffic those apps get) becomes the brand's story.

Distribution system

• Social over search: founder and team accounts on X and LinkedIn, plus creators and users sharing unprompted.
• Discord as the community home, actively staffed by the team, with helpers rewarded with early access and visibility.
• Structured programmes later: SheBuilds 48-hour buildathons (3,000 applicants, 200 selected), Ambassadors and City Leads hosting events, Campus Leaders at universities.
• Enterprise came after self-serve (Zendesk, Klarna, Deutsche Telekom, Uber cited).

AI angle

The product is AI-native, and the AI output itself is the distribution: every generated app is a shareable artefact. Fast model upgrades become frequent launch moments. Teams vibe-coding internal GTM tools on Lovable became their own content theme.

Key people & public artefacts

• Anton Osika (co-founder, CEO): built the audience via the open-source repo; podcasts with 20VC and Lenny.
• Fabian Hedin (co-founder).
• Elena Verna (Head of Growth, joined early 2025): publishes the playbook openly on her Substack and Lenny's Podcast.
• Felix Haas (design): reposts team and user builds.

Key insights

• Steal: Release a free, open version of your core idea before the company; it validates and builds distribution at once.
• Steal: If a launch fails, fix the product and relaunch under a name that signals the change, rather than repeating the same launch.
• Steal: Treat free usage as marketing spend in a new category, where the "aha" lives inside the product.
• Caution: Elena Verna is explicit that this worked because of timing and a fast-expanding category; it won't transfer wholesale to slower markets.

Sources

• Growth Unhinged: Lovable growth story (Kyle Poyar)
• Elena Verna: Growth lessons behind Lovable's $6.6B valuation
• N.Rich Challenger Brand library: Lovable
• Lovable community programmes