A decision that traces back to the words a user said

Discovery starts with a question. A research plan🗺️research_planA plan for gathering evidence to validate a draft entity is conducted as a research study🧑‍🔬research_studyA planned research activity that investigates typed research question❓research_questionA question guiding a research study nodes, follows an interview guide📋interview_guideA structured guide for conducting interviews, and enrolls the participant🙋participantA person participating in research nodes who can answer it.

Typing the study up front is what keeps the rest of the chain honest. Every finding later traces back to the question it was meant to answer, so a deliberate result is separable from an incidental one.

A research study🧑‍🔬research_studyA planned research activity captures observation👁️observationA specific behaviour or statement observed nodes, and each observation connects in every direction. It is evidenced by the exact quote🗣️quoteA direct quote from a research participant behind it, reveals a need🫀needA user need, pain, desire, or constraint, characterises the persona🎭personaAn archetype representing a user segment it came from, and yields the insight🔦insightA synthesised finding from research it supports.

The quote and the conclusion sit one edge apart, so an observation stays attached to the words that justify it and a need names the moment it was seen. The supporting evidence is a traversal away rather than a paragraph in a transcript.

A quote🗣️quoteA direct quote from a research participant is coded into an observation👁️observationA specific behaviour or statement observed; observations group into an affinity cluster🧲affinity_clusterA group of related observations; a cluster is synthesised into an insight🔦insightA synthesised finding from research; an insight surfaces an opportunity🚪opportunityA validated gap worth solving. Each rung is a typed node linked to the one below it.

The links are what hold the chain together. An opportunity on the roadmap traces back through the insight, the cluster, and the observation to the words a participant said. The question “why is this being built?” resolves to that path.

The playbook anchors the region on the opportunity🚪opportunityA validated gap worth solving, the node where research turns into work. An insight🔦insightA synthesised finding from research surfaces it, and from there the product addresses a need🫀needA user need, pain, desire, or constraint, pursues an outcome🏆outcomeA desired business or user outcome, drives a solution🔧solutionA proposed approach to address an opportunity, and explores a design concept🌱design_conceptA possible design direction or approach.

An Opportunity Solution Tree is one lens over exactly these edges. The opportunity carries the evidence underneath it forward into the work, so a solution always names the insight it answers.

An insight🔦insightA synthesised finding from research generates a hypothesis🧪hypothesisA testable belief about a solution, which requires an experiment plan📐experiment_planAn experiment plan describing the hypothesis, setup, success criteria, and methodology before a test runs.. Then an experiment run▶️experiment_runAn execution instance of an experiment that records actual conditions, observations, and raw results. validates the hypothesis, yields evidence🔎evidenceData supporting or refuting a hypothesis, produces a learning🧠learningAn insight gained from an experiment, and measures the metric🔢metricA unified metric that measures progress, health, or behaviour across the product that proves it.

Validated or refuted, the result is linked to the claim that prompted it. A refuted test leaves a learning attached to the insight, so the next iteration reads the prior result rather than repeating it.

Synthesis does not stop at a finding; it connects into the discovery-to-delivery spine. The opportunity🚪opportunityA validated gap worth solving a study surfaces is the same opportunity a solution🔧solutionA proposed approach to address an opportunity addresses, a hypothesis🧪hypothesisA testable belief about a solution tests, and a feature⭐featureA product capability or feature ships against.

Research becomes upstream structure the build reads from. When a feature is questioned, the graph names the insight behind it; when an insight is invalidated, the graph names the features resting on it.