Overview
A test results library preserves completed experiment outcomes so CRO teams can reuse what they learned instead of rediscovering it during the next planning cycle.
This playbook installs a durable Test results library table and Results library page. It is built for CRO leads who already run tests elsewhere but need a long-lived archive of what happened, what the team decided, and which lessons should shape future experiments.
The boundary is narrow on purpose. This is not an experiment tracker for planning, launch, sample-size monitoring, or live test management. It replaces the completed test readout archive: the place where final results, learnings, and follow-up decisions should stay findable.
Why you should preserve test learnings
Experiment programs lose leverage when readouts scatter across slides, docs, dashboards, and Slack threads. A test that seemed obvious in March becomes a vague memory by September, and teams repeat weak ideas because the old result is hard to find.
A useful results library keeps the learning attached to the decision. GOV.UK's A/B testing guidance emphasizes analyzing results against the original hypothesis and conversion goal, not just declaring a winner (GOV.UK Data Community). That context is what future planners need when deciding whether to reuse, avoid, retest, or investigate an idea.
Juno gives the archive a working structure: one row per completed experiment, one summary for patterns, and one page for scanning outcomes by funnel area, learning theme, confidence, and follow-up decision.
Step-by-step
- 1Confirm the experiment program, completed result sources, and any learning themes the team already uses.
- 2Review supplied readouts, experiment trackers, planning notes, result summaries, or workspace files for tests that have reached a final decision.
- 3Record each completed test with the tested change, control, audience, primary metric, guardrails, outcome, confidence, read date, source reference, and owner when known.
- 4Write a reusable learning for each result, separating what was measured from what the team believes it means.
- 5Assign a follow-up decision such as reuse, avoid, retest, investigate, roll into roadmap, or archive, with a concise next action where one exists.
- 6Refresh the Results library page with repeated patterns, conflicting results, stale evidence, missing readouts, and follow-up tests worth planning.
- 7Use the Results library page to review completed outcomes before roadmap planning, quarterly CRO reviews, or monthly experimentation retrospectives.
Frequently asked questions
What does this replace?
It replaces a test readout archive for completed experiments. It does not replace an experiment tracker, analytics platform, project board, or statistical testing tool.
What inputs should I bring?
Bring links, files, tables, or notes for completed readouts and prior test summaries. If the source list is incomplete, Juno can still seed the library and mark missing readouts as gaps.
Can this handle inconclusive tests?
Yes. Inconclusive results are useful when they preserve the hypothesis, sample limits, guardrail behavior, and why the team decided to retest, archive, or investigate further.
When should we refresh it?
Refresh it after each completed test readout or monthly if the team runs continuous experimentation. The monthly option reuses the same installed library so the archive stays coherent.