Discovery · Templates · User Research · PTOS

Discovery Research Kit: A Step-by-Step Template for Conducting Your Own Problem Research

A detailed, ready-to-use template for conducting Discovery research, including preparation, recruiting, interview script, note-taking checklist, and conclusion template.

Discovery Research Kit: A Step-by-Step Template for Conducting Research

The goal of Discovery research is not to invent a solution, but to catch a real breakdown and gather evidence so the team doesn't argue about 'tastes.' This step-by-step template will help you structure your research and get the most valuable insights.

1. Preparation (30–90 minutes)

1.1. Research Question (One)

Formulate one main question you want to answer.

  • Format: 'Why does segment [X] in scenario [Y] not perform action [Z]?'
  • Example: 'Why do new sales managers not record the next step within 15 minutes after the first call?'

1.2. Boundaries (To Avoid Drowning)

  • Segment: Who exactly? (role, level, channel, new vs. experienced).
  • Scenario: When exactly does the problem occur? (trigger + context).
  • Action: What exactly should have happened but didn't?
  • Horizon: The last 2–4 weeks (otherwise you'll get answers like 'usually in general').

1.3. Cause Hypotheses (3–7)

Write down possible causes in a mechanical format, not as solutions.

  • Format: 'At moment ___, the user doesn't do ___, because ___, therefore ___.'
  • Important: This is not what you are proving, but what you are trying to disprove or confirm.

1.4. Evidence Plan

Choose a minimum of 2 out of 3 sources for data triangulation:

  • Data: Funnels, cohorts, events.
  • Support/Sales: Top reasons for tickets, tickets, objections.
  • Observation/Interviews: Analysis of real past cases.

2. Recruiting (Quick and Pragmatic)

  • How many people: 5–8 people from one clear segment.
  • Who NOT to recruit: 'Fans' of your product, 'random passersby' not from your segment, those who haven't faced the problem in the last 2–4 weeks.

3. Interview Script (30–45 minutes)

3.0. Opening (1 minute)

'Hi! I want to understand how this happens for you in reality. I'm not selling anything. You can ask any questions, even 'stupid' ones.'

3.1. Warm-up: Context (3–5 minutes)

  • 'What is your role and what does a typical day look like?'
  • 'How often do you encounter situation [scenario]?'

3.2. The Main Part: Breakdown of the Last Real Case (15–20 minutes)

Key phrase: 'Recall the last time you ___ (did/tried to do). Step by step.'

  • 'How did it all start? What was the trigger?'
  • 'What did you do first? Why that way?'
  • 'Where did it become unpleasant/difficult/long?'
  • 'What did you do instead of the 'correct' action?' (searching for workarounds).

3.3. Alternatives and Workarounds (5–7 minutes)

  • 'What do you use to replace this now?'
  • 'What have you tried before? Why didn't it stick?'

3.4. Price and Motivation (5–7 minutes)

  • 'What happens if you don't solve this?'
  • 'What is the most painful part?'

3.5. 'Criticality' Check (2–3 minutes)

  • 'Rate the pain on a scale of 1 to 10. Why not 2? Why not 10?'

3.6. Closing (1–2 minutes)

'Thanks! Can I follow up with 1–2 questions later if I find a contradiction?'

Forbidden questions: 'Would you like it if...?', 'Would you use...?'. Ask about the past, not the future.

4. Note-Taking Checklist: What to Record

Record facts, not opinions:

  • WHAT (symptom): Where it breaks, what doesn't happen, how often, how it's noticeable.
  • HOW (mechanism): Trigger, constraints, friction, workarounds, user's language.
  • WHY (stake): Losses (time, money, quality, risk), who else is affected.
  • Artifacts: Screenshots, file examples, templates.

5. Quick Analysis (60–120 minutes after 5–8 interviews)

  1. Normalize: Condense each interview into 6 lines: Segment, Scenario, WHAT, HOW, Workaround, WHY.
  2. Cluster causes: Group similar causes into 3–7 clusters.
  3. Prioritize causes: Evaluate each cluster: how often was it mentioned, how critical is it, is there confirmation from other sources, can the product influence it.

6. Conclusion Template: What You Give to the Team

6.1. Problem Statement (1 paragraph)

  • For [segment]
  • in situation [scenario/trigger]
  • what's breaking is [WHAT: observable symptom]
  • because [HOW: 1–2 main reasons]
  • as a result [consequence for the user]
  • and this is important to the business because [WHY: metric/stake]
  • evidence: [2–3 points: data/support/observation]

6.2. Evidence Pack (5–10 bullet points)

  • A fact from data, a fact from support, 2–4 short quotes about the pain, 1–2 observations about workarounds.

6.3. Causes (3–7)

  • Cause #1: 'at moment ___ doesn't do ___ because ___ → therefore ___'.

6.4. Anti-Conclusion

  • What we don't know (2–3 points).
  • What alternative explanations are still possible (1–2 points).
  • What is the next experiment that will reduce uncertainty.

Using this kit, you will turn Discovery from a chaotic collection of opinions into a structured process of producing clarity, which will become a reliable foundation for your product decisions.