# Pick your first customers like a casting director

> Early users shape the product more than your roadmap. Here’s how I choose who gets in the room first.

- Author: Kevin Dalvi
- Published: 2026-07-02
- Read time: 7 minutes
- Tags: founding, customers, product
- Canonical: https://www.kevindalvi.com/blog/pick-your-first-customers

The worst early customers aren’t the ones who churn. They’re the ones who stay and drag the product toward a problem you don’t want to own. I’ve built features for loud friends. I’ve also built for quiet operators who paid and never complained until something actually broke. Guess which cohort made the company sharper.

First customers are a casting decision. You’re not filling seats. You’re choosing who gets to rehearse the future with you.

## Signals I look for

I want people who already feel the pain in their calendar, not people who think the idea is neat. Neat doesn’t ship. Calendar pain does. I also want users who can describe a workaround in ugly detail. That ugliness is the product spec.

Status buyers are loud early and quiet later. Operators who live inside the problem ask sharper questions on day one: exports, permissions, failure emails. Those questions are gifts. Schedule them before you schedule the launch party.

- They pay or explicitly trade time. Free forever fans wait
- They have a workaround you’d be embarrassed to automate badly
- They can introduce you to two peers without a pitch deck
- They disagree productively; cheerleaders teach you nothing

> Your first ten users should feel slightly inconvenient to support. Convenience is usually the wrong segment.

## Who I keep out early

I delay enterprise procurement theater, bargain hunters, and anyone whose success metric is ‘disrupting an industry’ with no weekly ritual attached. Also: myself. Dogfooding is useful; building only for your own edge case is a trap with good branding.

When someone asks to join a waitlist, I ask what they’ll stop doing if this works. Vague answers go to the newsletter. Specific answers get a call. That one filter has saved me months of roadmap noise.

I also watch how they treat my time. People who reschedule twice without apology usually treat the product the same way. Early access is a scarce seat; if they don’t protect it, they won’t protect the workflow you’re trying to replace.

## Design the onboarding around them

Once the cast is set, I write onboarding for those humans. Their language, their tools, their failure modes. Generic empty states are how you attract the wrong sequel. Tight casting makes the product feel inevitable to the right room and invisible to everyone else. That’s fine. Focus is a feature.

Pick carefully, then serve hard. Early customers aren’t a growth hack. They’re the raw material of taste. And the first people who will tell you, without mercy, when the product starts lying.