RPV is computed as total revenue / total visitors (or sessions) over a defined window. The denominator convention varies: Shopify reports it natively as “Sales per online store session”; in GA4, analysts typically build it as a custom calculated metric off session-scoped revenue totals; a CDP may define it against unique identified visitors. What matters is consistency over time within one brand, not which definition is “right.” The reason the metric is worth tracking at all is the algebraic identity it sits on: RPV = conversion rate × AOV. Orders divided by sessions, multiplied by revenue per order, collapses to revenue per session.
That decomposition is the operational point. A brand watching CVR and AOV in isolation can miss the composite reality that an experiment can lift one and tank the other, leaving RPV flat. The classic case is the free-shipping threshold: raising it typically lifts AOV — customers add an item to qualify — but suppresses CVR, because some shoppers abandon when the threshold feels punitive. Whether the net is positive shows up only in RPV. Reading either component alone gives an incomplete answer.
RPV is also more stable than either component when paid acquisition mix shifts the traffic, because a tilt toward lower-intent traffic typically drops CVR while AOV holds roughly flat — RPV moves less violently than a CVR-only read. Three common uses follow from this: as the primary metric in an A/B test (it avoids the CVR-vs-AOV tradeoff confusion), as the single number on a weekly storefront-health dashboard (volume-robust), and as the comparison metric across paid landing pages (normalizes for traffic volume differences).
Two limits. RPV is built on gross revenue, so a discount-driven lift can be margin-destructive — the headline number rises while contribution margin falls. And RPV pre-return-window overstates the durable read; net RPV after the return window is the cleaner number for definitive comparisons.