RunsheetProduct Discovery

Problem Interview

ComplexityMedium

Time30-60 min je Interview

Participants5-12 Interviews

FormatWorkshop + async

MaturityEstablished

Prerequisite

What needs to be finished first

Complete firstAssumption Surfacing

The most important problem assumptions are documented and prioritized so interviews provide targeted evidence.

Without: Without clear assumptions, the interview becomes small talk and produces arbitrary statements.

Preparation

What needs to be ready before start

Materials

Interview guide with 5-7 open questions around the problem hypothesis; recording tool with consent; notes template with columns situation, frequency, pain, workaround; recruiting plan.

People / roles

Interviewer with discovery practice; note-taker or transcription tool; recruiter; synthesis team of two to four people after interviews.

Pre-read

Problem hypothesis as one sentence; target segment with concrete persona; exclusion criteria; incentive; legal consent.

Time needed

30-60 min per interview, plus 20 min synthesis per interview

Setup

Record with consent. Keep guide open, no closed questions. Copy notes template. Write down pre-assumptions before the conversation so they can be challenged.

Core question

The one question this method answers

Is the assumed problem real, frequent and painful enough for the target person to justify solution investment?

Flow

Marker: Phase

Step	Duration	Action	Hint
1Phase 1: Warm-up and context	5-10 min	Introduce yourself, state the purpose of the conversation in one sentence ("We are exploring how people like you handle X, without selling anything"). Start recording, GDPR notice.	If the person asks about the solution, defer briefly ("I am happy to tell you, but first I would like to understand how it works today").
2Phase 2: Concrete situation	10-15 min	Ask about the last concrete situation in which the topic was relevant. "Tell me step by step how that happened." Take notes with timestamps.	Past situation beats hypothetical answer. "What did you do then?" is gold; "What would you do?" is wishful thinking.
3Phase 3: Pain and frequency	10-15 min	Probe deeper: how often does this happen? What did it cost (time, money, mood)? How important compared to other topics? Use "Tell me more about that" as standard follow-up.	Concrete numbers (hours, euros, frequency per week) provide robust evidence. Without numbers, it remains impression.
4Phase 4: Workarounds and alternatives	5-10 min	What does the person do today to address the problem? Which tools, tricks, colleagues, workarounds? How satisfied are they with that?	Existing workarounds show real investment level. Someone with a workaround takes the problem seriously; someone who does nothing either does not have it or has resigned.
5Phase 5: Close and notes	5-10 min	Reflect back the three most important takeaways and check them. Thank them. Immediately complete notes, mark three quotes with timestamp, enter problem-hypothesis evidence into template.	Complete notes within 30 min. Otherwise detail memory frays and the interview loses half its value.

Artifact

What comes out at the end

Form

One structured notes document per interview with persona, situation, frequency, pain, workarounds and verbatim quotes plus overarching synthesis document with problem evidence per hypothesis, frequency per pattern and risk assessment.

Versioning / ownership

Dedicated folder per discovery wave with date and hypothesis. Synthesis document as living document. Track status per hypothesis (confirmed, refuted, modified).

Tool alternatives

Dovetail or Condens for repository
Notion with one page per interview
Airtable with linked records
Markdown repo under docs/research/problem-interviews/

problem-interview-working-template.md

Compact working template for Problem Interview with context, input, output artifacts, and next step.

Problem Interview Working Template

Goal

Validates whether a suspected problem is real, important, and frequent enough.

Context

When and for what do we use this method?

Input

Which data, observations, decisions, or materials are available?

Execution

Short notes along the runsheet.

Output artifacts

Problem Evidence:
Risk Notes:
Customer Segments:

Assumptions and open questions

Decision / Next step

Owner, date, and success signal.

Example output

Concrete filled scenario, fictional example

problem-interview-beispiel.md

Concrete filled scenario, fictional example

Problem Interviews - Data Migration During Client Takeover (2026-05-10 to 2026-05-21)

Hypothesis: Solo tax advisors like Sabine lose 15-30 h per client during takeovers from predecessor systems because of manual data correction.

Interview 4 - Markus, 52, Hamburg, 28 clients (52 min):

Situation: takeover of 8 clients in Q1 2026 after predecessor retired.
Frequency: 2-3 client takeovers per year, 20-35 h effort each.
Pain: "I sacrificed three weekends. My wife complained."
Workaround: script from a colleague (Excel + VBA), works in 70% of cases.
Quote: "When the predecessor had SKR03 and I use SKR04, it is a translation thing nobody explains."

Synthesis (7 interviews):

Pattern: 6 of 7 report 15+ h effort per client takeover.
Pattern: 5 of 7 use informal scripts from forums or colleagues.
Pattern: chart-of-accounts mismatch is top-3 pain point for 7 of 7.

After 3 interviews, hypothesis is declared confirmed.

What to do

At least 5 interviews per segment, 8-10 if answers diverge strongly. Patterns are stable only after sufficient sample size.

Stop criteria

Done signals checkable in under a minute

Recruiting does not deliver people from target segment, so statements remain speculative.

Hypothesis cannot be formulated in one sentence or changes after every interview.

No note-taker or recording available, notes become incomplete.

Interview time under 20 min, so depth is not reachable.

Sales pressure overrides discovery, and interview becomes pitch.

No capacity for synthesis, so pattern recognition dissipates.

Finished the runsheet?

Go to the profile for purpose, similar methods, and sources or continue to the next method in the catalog.

Go to profile More methods