AI Startups Must Re-Find PMF Every 3 Months

I was struck by an interview with Elena Verna, Head of Growth at Lovable. She’s a 20-year veteran who led growth at Miro, Dropbox, and Amplitude - and she says only 30 - 40% of what she learned over two decades still works today.

Anyone running an AI startup has probably felt this firsthand, but her interview lays it out in a way worth revisiting.

The Traditional Growth Playbook Is Dead

In the old world, once you found product-market fit, you could scale for years on that foundation. AI is completely different. The technology moves so fast that customer expectations shift every three months. Even companies hitting $200M ARR have to keep proving themselves all over again.

The most surprising takeaway: funnel optimization contributes only about 5% of growth. The remaining 95% comes from shipping new features. Small, incremental improvements are meaningless. When everything is changing rapidly, obsessing over micro-optimizations gets you nowhere.

What Lovable Actually Does

Ship constantly, and talk about it constantly. Engineers announce their own updates. The founder posts progress on Twitter every single day. The product feels alive and breathing. Customers stay curious, and competitors scramble to keep up.

Give the product away for free - even when it’s expensive to run. AI products have high operating costs, but Lovable leans into free distribution anyway. They fund hackathons, sponsor events, and hand out credits. They treat this not as a cost center but as marketing. Word-of-mouth compounds like interest.

Influencer marketing outperforms paid ads by 10x. This one I felt deeply. Short videos showing what the product can actually do spread faster and convert better than any ad copy or explanation ever could.

Community is not optional. Lovable has hundreds of thousands of members on Discord helping each other, amplifying word-of-mouth, boosting retention, and creating a sense of insider belonging. You can’t just build a product anymore. You have to build an entire world around it.

What AI Startups Should Prepare For

There will be inflection points - moments when model performance leaps forward or pricing collapses dramatically. Those moments will be the opportunities that reshape markets.

Secure capital in advance. Build distribution channels and reference cases that let you launch instantly when the window opens.

• Elena Verna Interview