Skip to main content
Opening an experiment shows its results screen. A summary bar sits above a set of tabs — Traffic and Metrics always, plus one tab for each experience type in the experiment.

​
The Summary Bar

Time Period shows ∞ days when no end date is set, and is hidden while the experiment is a draft. Ending an experiment early freezes both readings. An experiment ended with two days left permanently shows “2 days left” — a record of how early you stopped it.

​
Traffic Tab

Traffic Allocation compares the split you configured against what actually happened. Small gaps between expected and actual are normal. Large or persistent gaps are worth investigating before you trust the results. Below it, Assigned Traffic plots each group’s share over time.

​
Metrics Tab

Metrics are split into Primary Metrics and Secondary Metrics, each showing one row per group. The control group is marked with a lock icon and reads Baseline in every improvement column. Each variant is measured against it. Lift is the relative improvement of a variant over the control, calculated as ((VariantMetric - ControlMetric) / ControlMetric) * 100. Below the tables, Metrics Over Time plots a selected metric for every group.
Judge the winner on the primary metric. Secondary metrics tell you whether the win came at a cost elsewhere.
Resist reading results in the first days of a test. Early numbers swing widely, and a variant that looks ahead on day two often finishes behind.

​
What’s Next