Show the value prop in the cart

Hypothesis

By surfacing claims-safe product value props inside the slide cart drawer, we expect carted visitors, particularly first-time buyers who have not yet built trust in the brand, to hold their intent through to checkout, thereby lifting checkout completion and revenue per visitor.

What we learned

Both variations beat control on revenue per visitor; the probability that control was best was 1%. V1 led with +5.4% CVR and +8.7% RPV unfiltered, and AOV up +3.1%. The effect sat almost entirely with new visitors (about 77% of traffic): +7.9% CVR and +12.6% RPV, while returning visitors were flat. That is mechanistically coherent: value props inform first-time buyers; returning buyers already know the terms. Honesty note on the headline: add-to-cart rate drifted about +2.2% between arms before the change could even render, so part of the raw lift is noise. Conditioned on actual cart exposure, the defensible effect is about +3% CVR / +6% RPV overall, and +6% CVR / +11% RPV among new visitors. We report the conservative number. Caveats we logged: about 97% of traffic was mobile so desktop was unreadable, the 2.1-day runtime had no weekend in it, and a concurrent test added variance late in the window.