You’re seeing the workspace of Camille Roux, cro lead at Maison Lune, a fictional store.
Free shipping progress bar and Orbe product page: title, photo, and badge have found a winner, Customer reviews at the top of the page is ready for a decision, One-page checkout is paused for a sample ratio mismatch, “You may also like” recommendations was stopped by a guardrail, and the Home page zone is free for a new test.
Measurement needs attention: sample ratio mismatch on “One-page checkout”.
+€511,400on “Revenue per visitor” in October 2026
Between +€149,200 and +€897,400: what all visitors outside the holdout added that month, measured against a 5% holdout that sees no tests and no rollouts. Running tests are included.
The impact report adds up the corrected effect of each rolled-out variant: +€205,100 per month across 2 rollouts.
+€125,700 more per month awaiting your decision
+27% on “Add to cart”. Expected gain: +1,706 converted visitors per month (corrected for the winner’s curse).
+6.0% on “Revenue per visitor”. Expected gain: +€125,700 per month (corrected for the winner’s curse).
-0.4% on “Revenue per visitor”, range -4.0% to +3.3%: no effect of the targeted size. Expected gain between -€212,000 and +€224,700 per month.
B is missing 1,108 Safari iOS visitors. Results are hidden until the mismatch is diagnosed.
“Return rate” +78% vs. the original (range +25% to +130%), past the tolerated change of +25%. Traffic went back to the original automatically.
| Experiment | Status | Sample | Effect · 95% range |
|---|---|---|---|
| Conclusive. Proven gain | 104% | +6.0% | |
| Sample ratio mismatch. | 61% | Results hidden until the split is diagnosed | |
| Not conclusive yet… | 69% | +1.9% | |
| Not conclusive yet… | 35% | +3.5% | |
| No detectable effect. | 100% | -0.4% | |
| Conclusive. Proven gain | 103% | +26.6% | |
| Not conclusive yet… | 21% | +1.5% |
No interaction proven, but none ruled out yet: range -4.72 → +4.09, wider than the target effect (±0.73).
Revenue per visitor: No interaction proven, but none ruled out yet: range -7.40 → +7.30, wider than the target effect (±0.55).
Product click: No interaction proven, but none ruled out yet: range -8.1 pts → +7.7 pts, wider than the target effect (±2.0 pts).
Revenue per visitor: No interaction proven, but none ruled out yet: range -9.20 → +10.21, wider than the target effect (±0.87).
Product click: No interaction proven, but none ruled out yet: range -9.9 pts → +7.2 pts, wider than the target effect (±1.8 pts).
Visitors seen by both tests, in the variant of their first exposure to each; the metric is counted after both exposures. 95% range, valid every time you check, with risk corrected across 5 comparisons.