# Customer calls are the research

> Surveys flatten people. Calls leave residue. If you are building something new, the calendar is your research budget.

- Author: Kevin Dalvi
- Published: 2025-12-16
- Read time: 6 minutes
- Tags: founding, research, customers
- Canonical: https://www.kevindalvi.com/blog/customer-calls-as-research

I used to treat customer calls like a courtesy. Nice to do. Easy to postpone when the sprint looked full. That was a founder tell. Prioritizing the work that feels like shipping over the work that tells you what to ship.

Now I treat calls as the primary research instrument. Not a funnel stage. Not a sales ritual. Research. The kind that ruins a confident roadmap in thirty minutes and saves you three months.

## What I listen for

I am not hunting for feature requests. People are generous with wishes and stingy with context. I want the last time the problem hurt, what they tried, who else was in the room, and what they paid in time or money to paper over it.

- The workaround they are embarrassed to admit
- The tool they open first on a bad Monday
- The person who has to approve a change
- The metric their boss actually asks about

> If you leave a call with a feature list, you interviewed wrong. Leave with a scene.

## Make the notes operational

Raw transcripts are theater. I write a short post call note with three lines: the job, the constraint, and the implication for this week’s build. If there is no implication, the call was networking. Still useful, but not research.

I also track a crude score: how many calls this month changed a decision? If that number is zero, I am either talking to the wrong people or protecting my ego. Both are fixable. Ignoring customers until the launch deck is polished is not.

The founders I respect most are slightly annoying about their calendars. They protect build time. They also refuse to invent the customer in their head. Calls keep the fiction from taking over.