Understand the research and experimentation model
Know when to use observations, insights, learnings, ideas, and experiments.
Each work item answers a different question. Keeping those questions separate makes it easier to trace a decision back to evidence.
| Item | Use it for | Example |
|---|---|---|
| Observation | What was seen or reported in a source | Three interviewees asked about delivery charges. |
| Insight | What the evidence suggests about a problem or opportunity | Unclear delivery costs make the total price hard to assess. |
| Learning | What the team now understands, with context and confidence | Delivery information helped first-time buyers in this test. |
| Idea | A possible change worth considering | Put delivery information beside the price. |
| Experiment | A planned test of a hypothesis | Compare the current page with explicit delivery information. |
Link the work
An observation can support an insight. Insights can connect to ideas, experiments, and learnings. These relationships preserve the reason for an idea and make later outcomes discoverable.
A Team is the shared workspace and access boundary. A Project organizes related optimization work within it. Pages, audiences, metrics, tags, labels, and states give project work a consistent vocabulary.
Keep source and interpretation distinguishable
Record the source and confidence when the field is available. Avoid copying the same statement into every work-item type: the observation captures evidence, the insight interprets it, and the idea proposes action. A learning should retain the conditions under which it was supported.
Research guides explain the links in detail. Project definitions explain the fields used to group and measure your work.