← Writing

40% of Signups This Week Came From AI Recommendations

Exactly 40% of new users this week found steel.dev through AI recommendations. Users told us this during onboarding.

Nikola Balić ·

Share and tools

I checked our onboarding flow for the last seven days.

Exactly 40% of new signups came through AI recommendations. That number comes from users themselves, not analytics: we ask during onboarding how they found us. Someone asked ChatGPT, Perplexity, or Claude for a browser automation solution, and the AI pointed them to us.

No surprises here, and no AI-has-arrived moment. It’s just data from a real product, with real users, over one week.

For dev tools, 40% from AI recommendations is a signal of changing surfaces. The people building AI agents are discovering infrastructure through AI assistants. The channel matches the customer.

We still position ourselves as browser infrastructure. But the majority of actual use cases? Giving browsers to agents. The web isn’t being replaced; it’s being operated. The product positioning catches up slowly, but the users are already there.

The part I’m thinking about more: if today’s users find us by asking AI, tomorrow’s users might not ask at all.

Their agents will decide.

You need a browser session for a workflow. Your agent researches options, evaluates tradeoffs, picks one, integrates it. You don’t know which product. You don’t care. The agent handles procurement the same way it handles config or deployments.

We’re not there yet. But if 40% of discovery is already AI-mediated, the next step isn’t that far. The question shifts from “how do I rank on Google” to “how do I become the default choice for agents making decisions I’ll never see.”

We didn’t optimize for any of this. It just happened. But I’m paying attention now.

Enjoying this? Get more like this in your inbox. Unsubscribe anytime.

Building go-to-market engines for AI-driven products with purpose. Worked with innovative startups like Numarics, Codeanywhere, Daytona, and Steel on growth strategies and market positioning. Faculty at University of Split, researching AI adoption patterns and developer tools.