Strategic Foundation (Prerequisites)

Before discovery can be effective, these strategic elements must be in place. Discovery is "tethered to strategy" - not random exploration.

Product Vision

Where are we going? The aspirational future state that guides all product decisions.

Product Strategy

How will we get there? The deliberate choices about where to play and how to win.

Portfolio Context

Where does this fit? Existing products, planned initiatives, resource constraints.

Context Decision: What type of discovery is this?

New Product → Full Business Model Canvas approach. Draft business model hypotheses in Stage 0, validate throughout, finalize in Stage 4.
Feature for Existing Product → Business model delta assessment. Reference existing model, note new assumptions (e.g., new pricing tier).

Three Essential Perspectives

Effective discovery requires input from three functional perspectives. Whether from one person wearing multiple hats or distributed across team members.

Product / Business

Strategy & Viability — What should we build and why?

Key Activities: 0.1 and 0.6 Stakeholder Alignment, 1.1 Segment Selection, 2.2 Solution Evaluation, 3.2 D/U/F/V Testing

Design / User

User Needs — What do users actually need?

Key Activities: 0.4 Draft JTBD Artefacts, 1.2-1.4 JTBD Research, 2.4 Job Stories with Acceptance Criteria, 3.2 D/U/F/V Testing

Engineering / Technical

Technical Feasibility — What can we actually build?

Key Activities: 0.5 Research & Technical Planning, 2.3 Assumption Mapping, 2.5 Prototype Creation, 3.2 D/U/F/V Testing

The Discovery Evolution

Artefacts evolve through stages. Key artefacts (highlighted) form the "North Star" that guides all subsequent work.

Stage 0
Problem Hypothesis → Research Plan
Stage 1
Job Statement → Job Steps Flow Diagram → Job Step Needs and Desired Outcomes → Opportunities
Stage 2
Opportunity Solution Tree → Selected Concepts → Job Stories → Prototype
Stage 3
D/U/F/V Evidence → Validated Solution → Confidence Assessment
Stage 4
Handoff Package → Success Metrics → Retrospective

Iteration Loops & Continuous Discovery

Discovery is not linear. These loops happen frequently as you learn and adapt.

Stage 3 Validation Loop

The most common loop. Test assumptions, learn from results, iterate on solution, re-test until confidence is sufficient.

Test → Learn → Iterate → Re-test ↺

Stage 2-3 Design Loop

When validation reveals solution gaps, return to Solution Design to refine before testing again.

Design → Prototype → Test → Refine ↺

Stage 1-2 Problem-Solution Loop

When solutions don't address needs well, return to Problem Discovery to ensure understanding is correct.

Frame → Ideate → Evaluate → Reframe ↺

Continuous Discovery

After launch, continue learning from real usage to inform the next discovery cycle.

Ship → Measure → Learn → Discover ↺
Stage 0

Initiation

Judgment Points Reference (JP1-JP19)

The 19 Judgment Points represent critical decision moments across discovery. Click any JP to see detailed guidance, pitfalls, and examples.

JP1Do we have initial evidence supporting this signal?
JP2Are we researching the right people?
JP3Have we correctly identified the job and breakdown?
JP4Which opportunities deserve focus?
JP5Which job steps are in/out?
JP6Are we generating good solution options?
JP7Are we selecting the right approach?
JP8Which assumptions could make the solution fail?
JP9Which assumptions should we test first?
JP10Is the prototype scope appropriate for testing?
JP11How will we test our assumptions?
JP12What does this evidence mean?
JP13Is this finding meaningful or noise?
JP14How confident are we in our conclusions?
JP15Do we proceed, iterate, pivot, or stop?
JP16What outcomes will indicate success, for whom, over what horizon and against which baseline?
JP17Which leading, lagging, risk and harm signals will show whether the outcome is emerging or the decision needs review?
JP18Did we achieve intended outcomes? What does it mean?
JP19What did this discovery effort teach us about our reasoning and practice, and what will we change?

Domain 1: Problem Judgment (JP1-JP5)

Determines whether you're pursuing something worth solving. Errors here compound through everything that follows.

AI support boundary: These prompts are for analysis, challenge and option generation. Verify AI output against source evidence. AI does not make Judgment Point or gate decisions; accountable people interpret the evidence, make the decision and record the reasoning.

Domain 2: Solution Judgment (JP6-JP10)

Covers decisions about what to build and how. Bridges understanding to testing.

Domain 3: Validation Judgment (JP11-JP15)

Covers decisions about evidence and confidence. Translates test results into decisions.

Domain 4: Learning Judgment (JP16-JP19)

Covers decisions about outcomes and organizational learning. Closes the loop back to capability improvement.

Discovery vs. Engineering Work

Discovery is for work that changes what users can do or how they experience the product. Bug fixes, technical debt, and infrastructure work are engineering-led.

✓ Requires Discovery

  • • New Product
  • • New Feature (Major or Minor)
  • • Minor Enhancement
  • • Quick Experiment

Validates user value: Do users need this? Will it work for them?

✗ Engineering-Led (No Discovery)

  • • Bug Fix / Defect
  • • Technical Debt
  • • Security / Compliance
  • • Infrastructure

Validates technical correctness: Does it work? Is it maintainable?

Escalation Triggers: Engineering work graduates to discovery if: "bug" is actually a design gap, fix requires UX change, tech debt has user-facing performance impact, or API changes affect external developers.

Discovery Types & Evidence Guidance

Discovery Type Research Sample Test Method Timeline
New Product Risk-based and expanded until decision risk is adequately addressed Least costly, least risky method capable of resolving the uncertainty Full process, scaled to risk and evidence access
Major Feature Risk-based; expand where variation, contradiction, uncertainty or consequence requires it Fit-for-purpose stimulus, simulation, prototype or spike Standard/focused profile; all five stages, four gates and all nineteen JPs at lighter depth
Minor Enhancement Targeted and risk-proportionate, using recent relevant evidence where appropriate Minimum method needed for valid learning Abbreviated, risk-proportionate path; compression must not become skipped reasoning or validation
Quick Experiment Zero to three targeted conversations if interviews are relevant; another evidence method may be more appropriate Assumption-specific Stage 3 only; no formal gates; five Judgment Points (JP11–JP15)

Key Artefacts Reference

Artefact Created In Used By Description
Job Statement Stage 1 (1.3) All subsequent The validated high-level job customers are trying to accomplish
Job Steps Flow Diagram Stage 1 (1.3) Stages 2, 3, 4 Complete workflow showing all steps to accomplish the job
Job Step Needs and Desired Outcomes Stage 1 (1.3-1.5) Stage 2 OST Pain points at each job step with I/S scores
Opportunity Scores Stage 1 (1.4) Stage 2 Contextual opportunity scores used as a prioritisation input alongside qualitative evidence and strategic context
Opportunity Solution Tree Stage 2 (2.1) Stages 2, 3 Visual tree connecting outcomes to solutions
Job Stories Stage 2 (2.4) Stage 3 Bridge from JTBD to development with acceptance criteria
Assumption Map Stage 2 (2.3) Stage 3 All assumptions by type (D/U/F/V) prioritised by risk
Confidence Assessment Stage 3 (3.4) Gate 4, Stage 4 Qualitative Confidence Profile by dimension, with evidence and gaps
Handoff Package Stage 4 (4.1) Development Complete documentation for development team