How I think, not just what I built.
Frameworks are only useful when they change a decision. These are the ones I actually run, and what they look like in practice.
Opportunity solution tree — live example
One outcome, mapped opportunities, competing solutions, and the cheapest experiment that could change my mind. Select an opportunity.
- • AI first-pass summary
- • Bulk document upload
- • Template library
Shadow-run AI summaries on 100 historical assessments and measure analyst edit distance.
Product Discovery
Continuous, not a phase. Weekly customer contact, assumption mapping and small tests before commitments.
Jobs To Be Done
Products get hired for progress. I write the job statement before the feature list.
Customer Interviews
Past behaviour over future intent. Stories, not opinions. Never demo before you understand.
Opportunity Solution Trees
One outcome, mapped opportunities, competing solutions, cheapest experiment first.
Prioritisation
RICE for throughput, cost-of-delay for sequencing, and an explicit not-doing list to protect focus.
Roadmaps
Outcome themes with confidence levels — Now / Next / Later — never a dated feature list.
North Star Metrics
One metric that captures delivered value, decomposed into inputs each team can move.
Experimentation
Hypothesis, guardrail metrics, minimum detectable effect. Kill criteria agreed before launch.
Go To Market
Positioning, pricing, enablement and launch tiering built into the delivery plan, not bolted on.
Product Lifecycle
Introduce, scale, sustain, sunset. Deprecation is a product skill most teams never build.