Europe's fastest growing startup?
Lovable’s path to $30M ARR in 4 months
I’m probably not the first to tell you about Lovable.
The Swedish AI coding startup grew from $0 to $10M ARR in only 60 days. Then they hit $17M ARR and 30,000 customers in 90 days – giving Lovable the reported distinction of being perhaps Europe’s fastest-growing startup. And now? $30M ARR in 120 days.
The company told me they’ve burned a mere $2M to get there. They did it with a team of 18 people, which translates into more than $1M ARR per employee (🤯).
It would be an understatement to say that this traction is impressive. Not only is it impressive, it’s practically unheard of in my 15 years of working with startups.
Even still, I was conflicted about whether to write about Lovable’s growth in this newsletter. A cool product meets viral word-of-mouth growth is great for Lovable, but that’s only so helpful for the rest of us (newsletter writers included!).
Spoiler: the story was so fascinating that I eventually caved. Keep reading for my interview with Lovable co-founder and CEO Anton Osika – along with what the ‘rest of us’ can learn from their viral growth.
👋 Hi, it’s Kyle Poyar and welcome to Growth Unhinged, my weekly newsletter exploring the hidden playbooks behind the fastest-growing startups. Subscribe to join 69,000+ readers who get posts like this delivered straight to their inbox.
The backstory: 18 months of building before breakout success
Lovable’s growth has seemingly come from out of nowhere. The reality is that Anton built an early proto-version of Lovable, called gpt-engineer, back in June 2023 while he was CTO at another startup. The project was a hit, blowing up on GitHub with “hundreds of thousands” of users and more than 50,000 GitHub stars.
Seeing the demand, Anton decided to build a company around gpt-engineer. His ambitions grew, too, as Anton believed he could have so much more impact driving change not just for developers, but for anyone with a great idea (no coding background required).
This led to a dedicated GPT Engineer app, which initially launched in December 2023 without much fanfare. Anton and team kept iterating.
They launched a better version in August 2024; it saw modest success, but growth quickly flatlined. GPT Engineer turned into Lovable in late November 2024, which is when growth started to truly explode.
What happened over those twelve months?
Lovable found a way to stop its AI from getting ‘stuck’
The team made improvements to allow Lovable to perform well on large codebases, unlocking the potential for ‘real’ use cases and greater reliability
They rebranded from GPT Engineer to Lovable to raise awareness about how much better the product had become
The TL;DR: Nail product-market fit with early customers before shifting attention to growth.
How Lovable keeps going viral online
Lovable’s early growth strategy sounds deceptively simple.
They launched on Product Hunt as The first AI full stack engineer, reaching #1 product of the day on November 21, 2024. They posted about product improvements on social media, mostly X and LinkedIn. And they ran a competition online, offering prizes for folks building with Lovable.
All of this is pretty standard fare for today’s startups. What made it work so well for Lovable: the AI was so good that people wanted to talk about it and the Lovable crew is exceptional at social media.
If you follow Anton, you might notice behind-the-scenes storytelling on Twitter threads. Regular awkward selfies on LinkedIn. Oh, and constant updates about product improvements and user successes. He comes across as authentic and his personality shines through.
“People think that when you post on LinkedIn, it should be formal and in a professional tone,” Anton told me. “It should actually be more like Twitter. You should be more authentic.”
The TL;DR: To standout on LinkedIn, keep it authentic and informal.
An exceptionally small team
For a software company with $20-$50M ARR, a “good” ARR per employee is $200,000. “Great” is $275,000. Lovable’s $1M ARR per employee is literally off-the-charts.
The team today includes 18 people. Of these, about 10 are writing code and a few folks are in operations.
“We believe in hiring super high agency people and very technically focused talent,” Anton emphasized. “Most of the team are former founders.”
Only three people work in growth, including a community lead, a growth engineer, and a content creator. There’s nobody in sales, although Anton mentioned that Lovable is hiring someone to onboard agencies (combination of sales and customer success).
This corresponds to five trends I’ve noticed around GTM hiring:
Fewer management layers. Even more senior hires are player-coaches who are asked to do their share of individual contributor (IC) work.
Keeping in-house teams small for as long as possible. Folks are hiring fewer, but high agency people. There's a premium on former founders or folks with side hustles — people who are used to tackling unfamiliar challenges. In-house teams are being supplemented with next-gen GTM tools, expert agencies and functional advisors who can fill gaps.
Generalists over specialists. A marketer might own content and paid. Or community and partnerships. Or product launches and demand gen. A GTM hire might look more like a 'GTM architect' doing a mix of pipeline generation (BDR/SDR), RevOps and growth.
Emphasis on velocity of experimentation. The idea: more experiments → more learnings → able to find what works faster than everyone else.
Premium on product, domain and/or technical knowledge. This is coming even to GTM roles where it didn't used to matter. The rationale: these skills allow folks to a) deeply understand users, b) build community and content that resonates, c) automate what they're doing so they can take on more.
The TL;DR: Lovable is part of a trend toward hiring fewer, but higher agency, people with a focus on generalists over specialists.
Lovable’s best (and worst) growth bets
When I talked to Anton, one thing became immediately clear: he’s big on activities.
Lovable’s philosophy is just launch – they’ll try things quickly and most of those things won’t be repeated a second time.
“We’ve tried a lot of different things,” Anton said. “They haven’t hurt us. The more things you do, the faster you learn what works.”
Among the growth activities Lovable has done in the last three months:
Product Hunt launches
Social media
Hackathons
Competitions
A builder hall of fame
A Product Hunt clone for Lovable products
An instant website builder
Partnership program for agencies
Website improvements
Figma import feature
Affiliate program
Referral program
Company blog content
One growth activity that has stuck is Launched, a Product Hunt clone but for products created with Lovable. “We built that with Lovable, obviously.”
Builders submit the apps they’ve made with Lovable and then the community upvotes what they love. The top five projects each week get 100 free Lovable credits, encouraging creators to keep building.
This is one of the few 🔥 examples of vibecoding I’ve seen for GTM, where a non-technical user can tap into tools like Lovable to build their own products specifically for growth. Why I’m digging it:
Everybody seems to be talking about vibecoding, but most folks don’t know what to build. This is a fantastic way to educate users with real-life examples.
Each of the apps has an “Edit with Lovable” button, which is clickable and takes people to see exactly how the app was built. This brings the community back into the product and allows them to create a similar app in seconds. It’s like a Template Gallery on steroids.
It’s built with and for the community. Launched helps Lovable app builders get users. It’s gamified to drive engagement. And it’s continuously updated — 316 projects were submitted just the last week I checked.
Another promising vibecoding example: Linkable, Lovable’s instant personal website builder. It was created by a “half technical” Lovable team member with Lovable. In a week and boosted by a single Twitter post, Linkable has blown up to create 20,000 websites near-instantly (like this one, made in about 60 seconds).
The growth loop is simple yet effective:
Someone lands on the website builder.
They fill out two fields, entering their LinkedIn URL and email address.
Lovable auto-generates a free personal website. While it’s not perfect, it’s surprisingly professional-looking and, more importantly, instant.
They click “Edit with Lovable” and they’re now vibecoding with Lovable.
One of Lovable’s next big growth bets is, not surprisingly, B2B. This is classic for product-led companies that start self-service and then move upmarket in search of larger deals and stickier customers.
The potential directions for a Lovable B2B offering are nearly limitless. In addition to enterprise-specific pricing or compliance capabilities, it’s not hard to imagine an app building experience at-scale with a pre-set look and feel, access to internal systems and data, and with bespoke protocols. While Lovable hasn’t marketed a B2B offering yet, they do have a ‘Contact Us’ form on the pricing page that’s already attracted “tons of inbound demand from enterprises.”
The TL;DR: The downside of most GTM experiments is low. Shipping something is the best way to get learnings quickly (and tools like Lovable can accelerate this process).
Pricing is still a work in progress
True to form, Anton’s initial pricing philosophy was just ship. Lovable decided to give some free access to everyone and then charge a monthly subscription for more usage (starting at $20 per month).
What he quickly noticed was that Lovable “lost a sh**-ton of money on the super active users.”
Lovable then updated pricing to aim to at least make a small profit when a user maxed out the usage quota. And the team has continued to evolve pricing from there.
Lovable’s pricing model is what I’d categorize as hybrid pricing – where there’s a combination of subscription pricing with usage limits or paywalls. This type of pricing is increasingly coming to an AI product near you including ChatGPT, Replit, monday.com, Clay, Zapier and Asana.
I asked why they didn’t pursue usage-based pricing like many other AI-native startups.
Anton believes the price of AI is going to go down over time, allowing Lovable (and others) to achieve better margins over time with the current pricing. And he views changing pricing as having an opportunity cost.
“We could improve the product or change the pricing. We choose to improve the product.”
The TL;DR: AI products don’t necessarily mark the downfall of subscriptions. Hybrid pricing models are an increasingly popular way of generating recurring revenue and offering customers predictable pricing while protecting margins.
Is Lovable the “last piece of software”?
In Anton’s view, company building is changing rapidly. “We’re going to have AI creating entire companies soon,” he predicted.
In the near term, Anton expects that AI will take over the first phase of company building – the v1 SaaS product or landing page. “You still need an engineering org to go further,” he said. “But that will change quickly. And, if you’re technical, you won’t need to hire engineers.”
As the barriers to creating software fall, Anton thinks the best way to stand out is to tap into the algorithms of social media. “The best way to get awareness and adoption is to create the best content that people want to consume. And then algorithms will get this interesting content in the hands of your users.”
It’s a fascinating vision. Stay tuned to this newsletter to see how it plays out.
I absolutely love what lovable had achieved so far. It's stretching the imagination for many to set new benchmarks for early stage potential - when done right and also with a sprinkle of luck to get vitality.
Would have also loved it there could be some mention of churn. Since it's only 4 months, the ARR is a projection and the biggest next mountain to climb would be to cut the churn due to single/limited time use case offering around building something quick etc. I am confident Anton and team will crack that too.. but would be useful for learning to see forward looking and then come back look at it the hindsight as well..
A great story. And we will see more and more of these as AI based companies surge and the growth models become better understood. Looking forward to Kyle's contributions to shaping our understanding!