Skip to content

Discovery

TL;DR

A weekly habit of customer interviews, an Opportunity Solution Tree (OST) the team updates together, and one working hypothesis the next loop will test.

What this is

Discovery is the loop that keeps the team honest about the problem. It runs from day one to the last and answers the same question on repeat: what problem are we actually solving, and for whom?

This is the Continuous Discovery practice Teresa Torres describes in Continuous Discovery Habits (2021): at a minimum, weekly touchpoints with customers by the team that’s building the product. The cadence is what makes it different from one-off research. A weekly conversation does not produce a clean answer; it produces a small, falsifiable hypothesis the team can act on before the next conversation.

What happens

  • A weekly customer interview, 30–45 minutes, recorded and written up the same day.
  • An Opportunity Solution Tree — a single page with four layers: Outcome (the customer value we are after), Opportunities (customer needs, pains, desires), Solutions (candidate approaches), Assumption tests (the smallest thing we can do to learn whether a solution is worth building).
  • A working hypothesis each week, written as a falsifiable claim, attached to one opportunity and one assumption test.
  • A short note at the end of the week listing what we learned, what changed in the OST, and what the next conversation will test.

What you get

  • An OST you can show to stakeholders without translation: it shows the team’s thinking, not just their conclusions.
  • A weekly log of decisions and what they were based on.
  • A working hypothesis the team can defend with evidence, not opinion.
  • Customer voice recorded, transcribed, and searchable — so anyone who joins later has the same context.

Cadence

  • One interview a week, minimum. Torres sets the baseline at one customer conversation a week. On a solo engagement, two or three a week is the realistic target for a team that wants to learn quickly.
  • The trio is the team. In Torres’ model the interview is run by a product trio: a PM, a designer, and an engineer. On a solo engagement I play all three parts. The structure is the same; the headcount is not.
  • OST updated Friday. The tree is a living document; if it isn’t being edited weekly, the team has stopped learning from customers.

What I need from you

  • Access to one or two customers a week, with the relationships warmed up by you — cold intros take longer than the conversation itself.
  • The last quarter of support tickets, churn notes, or sales call transcripts — anything that records real customer words.
  • A short written brief in your own words, even if it is rough — rough briefs make better opportunities than polished ones.

What comes next

The working hypothesis feeds Design. Discovery does not stop when design starts — it continues in the background and the two loops read each other’s notes.

If this is the kind of help you need, the next step is a 30-minute call.

Book a call →