Lovable's Growth Playbook: Give It Away, Ship Daily and Recapture Product-Market Fit Every Three Months

Verna says Lovable launched in the third week of November 2024, having existed earlier as GPT Engineer, and passed $200 million in annual recurring revenue before its first birthday. It reached $100 million at the end of July and doubled in about four months. She reports more than eight million users who have tried it and hundreds of thousands of paid subscribers, with about 100 people at the company, up from 30 six months earlier.

She warns founders against treating this as a benchmark, calling it a once-in-a-lifetime company in a fast-moving category. The revenue, she says, comes from three groups: non-technical founders building apps from scratch, employees building internal tools, prototypes and landing pages, and people exploring what is possible. She uses John Cutler's framing that software moves through capability, value and scale stages, and says vibe coding is still in the capability stage, where what you can do changes every month or three.

Verna says she can't share exact figures, but paid retention is comparable to the B2B SaaS products she has worked on, including Miro, Dropbox, SurveyMonkey, Netlify and Amplitude. Net dollar retention is good because people who build want to buy more credits. The team is more focused on engagement retention, which she calls the leading indicator, because the north star is getting as much usage as possible and tuning monetisation afterwards.

She adds that Lovable doesn't optimise for revenue at all. Internally they discuss how to give more away and get more paid subscribers and users in the door to win market share. Revenue is an outcome of focusing on inputs, she says, not a number they try to grow directly.

In previous roles, Verna says, she was confident in about 80% of the patterns she brought and felt she was copy-pasting. At Lovable, she estimates only 30 to 40% of what she learned transfers, mainly paid marketing, habitual retention and free-to-paid frameworks. The market is a new category in fast-moving water with many competitors, so staying ahead means reinventing the solution rather than tweaking dials, and she now spends 95% of her time on innovation.

Her growth team therefore ships features that would normally belong to product. It built the Shopify integration to support e-commerce use cases and added voice mode to increase engagement. The team is also moving into agent work, codifying agent instructions so customers activate better, which takes growth deeper into core product functionality.

Activation, usually her main focus, gets little of her time, because users simply prompt an agent and the agent team works on it day and night. An improvement to the agent helps every generation, not only the first. Lenny notes that this doesn't mean ignoring the experience, only that the team building it is already obsessed with it.

Building in public is one of Lovable's biggest strategies, coupled with founder-led and employee social posting. The company still tiers launches, with tier-one launches as big moments meant to change product-market fit. In between, Verna says, the goal is to "maintain noise in the market" by shipping every day or several times a day and talking about it, which also works as a resurrection and re-engagement strategy because users check social to see what has changed.

Velocity is the core value of the development team, and engineers, whom Lovable calls product engineers, announce what they ship rather than funnelling it through marketing. Lenny compares it to Captions' policy of shipping a marketable feature each week. Verna says this requires autonomy and agency, otherwise you would need an enormous marketing team.

She argues that marketing is changing in three ways: channels, marketing's involvement in product, and where marketing hires. Positioning now lasts about three months rather than years, so product marketers focus on big narratives. For organic growth, she says, social has replaced SEO, including CEO posts, employee posts, creators and users sharing their own results.

For founder and employee socials, Verna favours X and LinkedIn, but says you can't have ChatGPT write your posts. People need to see personality and humanity, because with so much competition on functionality, they want to rally behind a team. She adds that customer word of mouth only appears when you blow their socks off and give them stories that make them feel they have superpowers.

The product's name helps: if something isn't lovable, it doesn't ship, and the way to fix a bug is to say it isn't lovable. Lovable has no brand marketing team yet, so brand work goes into product interactions and the "love marks" a designer adds. Influencer marketing is ten times bigger than paid social for them, because a short video shows what is possible in a way a written value proposition can't.

Verna says the first requirement is being in the right place at the right time in a fast-moving category, and the way to stand out in a crowded one is through experiences that feel human. Because the cost of building software is falling, she argues, companies can now invest in the emotional feel of software, and a designer should be one of a startup's first hires.

To avoid a Frankenstein product when everyone ships, Verna says the answer is mainly who you hire. Lovable looks for people for whom the work is a hobby, with fire in their belly, rather than people with big logos on their CV. High agency and high autonomy are the key traits, so people can own work all the way from start to finish with little supervision, and failure is cheap because of the pace.

Recruiting in Stockholm means asking people to relocate, but the hype around the product helps, and people reach out themselves. The company runs paid work trials and probation periods because the pace is not for everyone. Verna says she felt she needed to re-onboard after ten days of vacation, and the company prioritises people who do not need clarity and can create it from chaos.

Lenny's summary is to build something lovable in a growing market, and Verna prefers the label minimum lovable product, since viability belongs to the 2010s. She describes the feedback cycle as collapsed, with ideas going to working product and user feedback within a day. She also has a full-time vibe coder on her team, a former chief of staff, who built Shopify templates and prototypes, and she thinks vibe coding will become a skill listed on many roles.

Beyond product and noise, Verna points to community as another lever. Lovable's community runs on Discord, has hundreds of thousands of members, community managers and an ambassador programme, and it amplifies word of mouth, social posting and retention. She says not everyone can build a community, but you can at least plug into someone else's.

The other lever is giving the product away, even though every AI interaction has an LLM cost and traditional companies tend to gate AI behind a paywall to protect margins. Verna says freemium is only the baseline: in a new category you need a wow moment, which for Lovable is the first preview after a prompt. Credits go free to every event and hackathon, and the cost is tracked as marketing spend rather than something to reduce.

AI companies have lower margins, around 40% by her estimate, so treating AI costs as a cost centre is a trap. Lovable has about 100 people, little paid marketing, a small sales team and influencer spend in the low double digits as a share of growth, so the money shifts from ads to product. She calls it product-led growth supercharged, and says VC funding is for future development and hardening the business rather than survival.

Verna says product-market fit used to evolve over years, with long blitz-scaling phases. Now it is every three months, because LLM capabilities jump with each new model and consumer expectations shift within a month. Companies must build ahead of a model release, bet on what it will do, and recapture fit with both technology and customers changing.

She says this is confusing for a $200 million company that must keep reinventing as well as scaling, since the team that finds product-market fit differs from the team that scales. Lenny cites OpenAI's reported "code red" after Gemini 3 launched as evidence that nobody's position is bulletproof. Verna agrees that a decade ago she would have called it crazy to say a company growing 10% month over month could lose product-market fit within three months.

She also worries about adjacent users. Lovable's core users are mostly pioneers excited by capabilities, while the latent majority sits among adjacent users, and the category is so busy recapturing pioneers that it may alienate the majority. She admits she doesn't know the right answer, and notes that OpenAI has arguably begun reaching broader audiences.

Verna advises people to consider whether they are comfortable in the messy middle. AI companies are hectic and unstable because of the product-market fit treadmill, but they let you absorb new skills quickly: at Lovable she uses AI far more than she did at Dropbox. She describes AI as "average intelligence" that gets the base work done so she can add human thinking and creativity, and she says people who prefer structure and deep specialty may be better off elsewhere for now.

On work-life balance, she says she has two kids and a good balance by setting firm boundaries, and that work is also her passion. She doesn't believe in balance as such: she prioritises family or work depending on the moment and asks whether she will resent a choice later. She protects sleep, health, workouts and family time, because she is most creative after a break from work.

Inside Lovable, internal tools are built on Lovable, specs come with prototypes people can click through, and Verna mocks up pricing page changes before handing them to engineers. She also uses ChatGPT's deep thinking mode for brainstorming, Granola for meeting summaries and Wispr Flow for voice input. She says trying to vibe code an idea calibrates her quickly, because it shows which parts of an idea matter and gives engineers a clearer brief.

She also raised concerns about women in tech: reports show a gap in AI adoption between women and men, and at Lovable, despite a pink and purple heart brand, third-party data suggested women were at most 20% of users. She says she doesn't blame men, but worries women are stuck in the latent majority and that products should be built by a representative group. Lovable's response is SheBuilds, a women-only hackathon with 48 hours of unlimited access, where participants often build hyper-local tools for family and community needs.

On hiring, Verna thinks the story that new graduates have no jobs is false for AI-native graduates, who bring fresh perspective and whom she learns from. She adds that failed startup founders are also in hot demand because of their agency and autonomy. She urges schools to teach AI, and says graduates should lead with what they can achieve with AI.