The Variant Comparison card has a date range picker in the top right. Click it to open a calendar and select a custom start and end date. The range cannot go before the test start date or after today. All metrics update when you change the range.
Visits — how many visitors landed on the variant.
Bounce % — how many left without interacting.
Engaged % — how many clicked or typed in any field.
Conversion % — how many submitted. This is the headline column.
Avg Time (s) — average seconds from first interaction to submit.
Significance — how confident the result is compared to the Control (non-control variants only).
Each non-control variant is compared to the Control using a two-proportion z-test on conversion rate. The badge reads '[X]% confident' once significance is reached, or '[X]% — not yet significant' below that threshold. If either the Control or the variant has fewer than 10 visits, the badge reads 'Not enough data'. The Control row is always labelled 'Baseline'.
Below the comparison table is an All Leads (combined) card. It shows every submission across all variants in a single table with columns: Variant (with a Control badge where applicable), Name, Email, and Submitted At. Click any row to open the submission drawer and see the full response. You can also filter by Converted or Unqualified status.
Wait until the significance badge reaches 95% or above.
Sanity-check with volume — even a significant result on a small sample can reverse. Aim for at least 100 completions per variant before declaring.
If the test is flat after plenty of volume, the change did not matter. Declare the Control the winner and move on to the next experiment.
