Reading the Report

Open Alpha Insights → A/B Test → Report (also under Website Traffic as A/B Tests). The A/B Test list shows the same warehouse metrics for all time. The report is date-filtered (default last 30 days) and includes charts plus a full variant table.

You can filter the report to one or more A/B tests from the filters modal (Pro for custom filter sets, same as other reports).

Headline metrics

Metric	What it measures
Exposures	Assigned visitors who saw a variant. This is the conversion-rate denominator.
Conversions	Exposed visitors who completed the A/B test’s primary goal. For Purchases, purchase value, and average order value, that is an attributed purchase.
Conversion rate	On the date-filtered report this is conversions ÷ exposures (purchase rate for purchase-based goals). On the list, Target conversion can switch the rate to add-to-cart rate, checkout rate, revenue per visitor, revenue per exposure, or AOV.
Lift vs control	How far a treatment sits above or below that test’s own control on the selected rate. Control is not given a lift.
Add to carts / initiate checkouts / purchases	Funnel counts from the same exposed sessions. Rates are also available as % of exposures.
Sessions / duration	Analytics sessions joined to exposed assignments.

Pro-only columns and goals

Revenue (purchase value) — attributed purchase value from session events, with visitor-ID fallback on the order. As a saved primary goal (Pro only), the list scores lift as revenue per exposed visitor.
Revenue per exposure — revenue ÷ exposures. Use this instead of total revenue when variants have different traffic. As a saved primary goal (Pro only), the list scores lift on this rate.
AOV (average order value) — revenue ÷ attributed purchases. As a saved primary goal (Pro only), the list scores lift as AOV.
Profit — store profit for those attributed sessions / orders.
Significance — scored against that test’s own control, even if more than one test is in the date range. Confidence interval columns appear on the variant table when Pro is active.

Free lists and reports stay on Purchases. Revenue, revenue per exposure, AOV, profit, significance, and the value goals require Pro.

Charts

Hits per A/B test — exposures over time, grouped by test.
Conversion rate by variant — includes control so you can compare arms on the same axis.
Daily lift vs control — treatment arms only.

How to read a result

Confirm both arms have meaningful exposures. Very small samples swing wildly.
Match the metric to the hypothesis: purchase rate for conversion tests, revenue per exposure or AOV (Pro) when the change is about value rather than volume.
Use the funnel (add to cart → initiate checkout → purchase) to see where a treatment helps or hurts.
In Pro, wait for significance before you ship a small lift. A “leading” badge on the list is only the current highest rate for the selected target conversion, not a statistical call.
If the list and report disagree, check the report date range. The list is all-time.

Metric names in Report Builder