A Shopify experiment record should make the test reproducible and auditable by capturing why it exists, what changed, who was eligible, how success is judged, and what decision followed.
Before the test
Record the problem evidence, hypothesis, owner, audience, control, variant, primary outcome, guardrails, minimum useful effect, runtime rules, traffic estimate, and known conflicts.
Before launch
Attach implementation references and a completed QA checklist for assignment, rendering, events, orders, markets, performance, accessibility, and rollback. Note the exact versions or theme state being tested.
During the test
Log start time, pauses, incidents, campaign changes, inventory events, theme releases, and deviations from the plan. Avoid rewriting the original hypothesis or success rule after results are visible.
After the test
Record the estimate, uncertainty, guardrails, decision, limitations, rollout scope, monitoring, and follow-up questions. Link the applied change or explain why a statistical lead was not shipped.
Run the test in Shopify
Sofcy supports URL, template, theme, and custom JavaScript API experiments.
Start free trial