Building Lovable: $10M ARR in 60 days with 15 people | Anton Osika (co-founder and CEO)
Anton OCK, co-founder and CEO of Lovable, discusses their AI engineer product, which hit 10M ARR in two months with 15 people. He shares insights on building lovable products, hiring generalists, and leveraging AI for rapid development.
Deep Dive Analysis
19 Topic Outline
Introduction to Lovable: Your AI Software Engineer
Lovable's Rapid Growth and Scale
Live Demo: Building an Airbnb Clone with Lovable
Tips for Mastering Lovable and AI Tools
Origin Story of Lovable and GPT Engineer
Overcoming AI's 'Getting Stuck' Problem
Lovable's Reliability and Differentiators
The Vision and Future of Lovable
Evolving Skills and Product Teams in the AI Era
Lovable's Hiring Philosophy and Team Culture
Advantages of Building a Startup in Europe
Product Prioritization and Roadmap Strategy
Tools and Work Environment for Fast Growth
Tactics for Moving Fast with a Small Team
Advice for Building Future-Proof Product Teams
Empowering Non-Technical Founders with AI
Lovable's Upcoming Features and User Support
Product Lesson from a Past Failure
Final Advice: Mastering AI Tools
3 Key Concepts
Minimum Lovable Product (MLP)
A concept emphasizing the goal of building products that users genuinely love, beyond just being functional or viable. It suggests striving for a 'lovable product' and eventually an 'absolutely lovable product' as a measure of success.
AI 'Getting Stuck'
This refers to a common challenge in AI systems where they perform well initially but then encounter bugs or an inability to progress further. The AI isn't smart enough to debug itself or find a way out of the problem, requiring human intervention.
Agentic Behavior (in AI)
This describes giving an AI system more autonomy to make decisions and take actions independently. For example, an agentic AI might decide to write a test, run it, identify failures, and then autonomously fix those issues to achieve a goal faster.
9 Questions Answered
Lovable is an AI software engineer that takes an English prompt and codes a fully working product in minutes, enabling non-technical individuals to turn their ideas into real businesses and helping developers build software much faster.
Lovable launched less than three months ago, hit $4 million ARR in the first four weeks, $10 million ARR in the first two months with just 15 people, and is the fastest-growing startup in Europe ever.
Anton Osika was inspired by ChatGPT's ability to generate code from human instructions, leading him to create the open-source tool GPT Engineer. He then co-founded Lovable to empower the 99% of the population who don't write code to build software using AI.
Lovable painstakingly identifies specific areas where AI agents get stuck (e.g., adding login, data persistence, payments) and then quantitatively tunes the entire system with a fast feedback loop to improve performance in those critical areas, making it more reliable.
Being a generalist with diverse skill sets (architecture, design, product taste, user communication) will be crucial. The ability to figure out 'what to build' and having good taste to refine products will become more valuable than pure engineering execution.
Lovable looks for individuals who deeply care, preferably obsess, about the product, users, and team collaboration. They seek people with absolute superpower in one dimension (like AI optimization) combined with a generalist mindset, often using work simulations to test candidates.
Lovable uses a very simple algorithm: identify the biggest bottleneck or problem, solve it quickly, then pick the next one, without overthinking a long roadmap. They are engineering-led, with weekly planning and demos, and a clear roadmap for the coming month that is flexible.
Anton learned that retrofitting AI into an existing product (like an API for personalized learning) is very difficult. Instead, it's crucial to start with the end-to-end user experience and the big picture of how the product should work, then strategically integrate AI to solve specific problems.
To become a top 1% user, one should find a problem they want to solve, dedicate a full week to using AI tools (like Lovable) to achieve that outcome, and constantly ask the AI (or other tools like Claude/ChatGPT) questions to understand how it works and what to do.
15 Actionable Insights
1. Become an AI Tool Expert
Dedicate a full week to solving a problem using AI tools to become proficient, as being in the top 1% of users will significantly differentiate you in the coming years.
2. Build Lovable Products
Strive to create products that users genuinely love, starting with a “minimum lovable product” and continuously improving, as this drives organic growth.
3. Practice Clear AI Prompting
When interacting with AI, be extremely specific about expectations and what is or isn’t working, as precise communication is even more critical with AI than with humans.
4. Prioritize by Biggest Bottleneck
Identify the single largest problem or bottleneck, solve it quickly, and then move to the next, avoiding overthinking long-term roadmaps for rapid progress.
5. Start with User Problem, Then Add AI
Begin product development by understanding the end-to-end user experience and core problem, then strategically integrate AI to solve specific aspects, rather than retrofitting AI into existing solutions.
6. Hire for Deep Care/Obsession
Seek candidates who demonstrate profound care or obsession for the product, users, and team, as this commitment is a strong predictor of long-term contribution and success.
7. Conduct Work Trials for Hiring
Implement work simulations, having candidates join the team for at least a day or a full week, to assess their thinking, reasoning, and collaborative abilities in a real-world setting.
8. Create a Clear Hiring Filter
Use job descriptions to explicitly communicate the intense pace and ambitious mission, attracting candidates who thrive in such environments and filtering out those seeking comfortable work.
9. Embrace Generalist Skills
Prioritize hiring individuals with diverse skill sets across architecture, design, product taste, and user interaction, as being a generalist is increasingly valuable in AI-driven product teams.
10. Leverage AI for Post-Launch Growth
Consider how AI can assist founders not just with building, but also with critical post-launch activities like user acquisition, feedback collection, and overall go-to-market strategies.
11. Build in Public for Awareness
Regularly share product updates and achievements on social media to generate awareness and foster organic word-of-mouth growth.
12. Embrace Engineering-Led Product Decisions
For AI-first products, allow engineering to lead product decisions, as optimal solutions often involve technical details and larger initiatives intertwined with problem-solving.
13. Foster In-Office Collaboration
Encourage in-office work and informal interactions, like shared lunches, to facilitate high-bandwidth communication, cross-pollination of ideas, and continuous subconscious problem-solving.
14. Use AI Chat Mode for Learning
Utilize AI tools’ chat modes to ask questions, understand functionality, and get unstuck, which also serves as an effective way to learn about software engineering principles.
15. Use Linear for Talent Tracking
Employ versatile tools like Linear for various company operations, including talent application tracking, due to its effectiveness and simplicity.
8 Key Quotes
Lovable is your personal AI software engineer. You describe an idea and then you get a fully working product.
Anton Osika
The best word for a great product is that it's lovable.
Anton Osika
Being in the top 10% in using them is going to be to absolutely set you apart in the coming months and years.
Anton Osika
Explaining exactly what you expect and what you're not getting is even more important with AI than with humans.
Anton Osika
We're seeing the biggest change humanity will ever see, I think, where like before you had manual labor being taken over by machines, but now it's actually cognitive labor being done better than humans by machines.
Anton Osika
If I'm putting together a product team today, I would re-obsess about getting as much of as many skill sets as possible for each person I hire.
Anton Osika
The biggest bottleneck for most products these days is not going to be as much on engineering, but having good taste, good intuition about your users.
Anton Osika
If you spend a full week on trying to reach an outcome... you're in the top 1% in the global population.
Anton Osika
1 Protocols
Lovable's Product Prioritization Algorithm
Anton Osika- Identify the biggest bottleneck or problem facing the product or users.
- Focus intensely on solving that specific problem.
- Iterate quickly on the solution.
- Once the problem is solved, pick the next biggest problem and repeat the process.
- Avoid overthinking or dreaming out a very long roadmap, focusing on immediate, impactful solutions.