RPV Fundamentals
Revenue Per Visitor is the single number that ties traffic, conversion, and order value together. Here's how to compute it, read it, and use it to judge tests.
Revenue Per Visitor (RPV)
Revenue Per Visitor is total revenue divided by total sessions — the average revenue a single visit generates.
Revenue Per Visitor (RPV) is the headline efficiency metric for an online store: it answers, on average, how much revenue every visit produces. The math is simple — total revenue over a window divided by total sessions in the same window — but the metric is structurally richer than conversion rate because it captures both whether a visit buys and how much it spends.
RPV is the product of two levers: conversion rate (CR) and average order value (AOV). That decomposition is why CRO and merchandising teams converge on RPV as the test-grading metric — a variant that lifts CR but tanks AOV is a wash, and only RPV will tell you that.
Most Shopify and WooCommerce dashboards surface conversion rate by default, which is why teams over-index on it. RPV is the better default because it stops you from celebrating a test that converts more shoppers into smaller baskets.
If you run a beauty or apparel store with a wide price range across SKUs, the gap between CR-thinking and RPV-thinking gets wider. A free-shipping threshold change, a bundle module, or a PDP upsell can all move AOV independently of CR — and only RPV captures the net effect on the P&L.
How to compute RPV
The formula is RPV = Total Revenue ÷ Total Sessions. Use sessions, not users — a single shopper comparing two pairs of sneakers across three visits is three chances to convert, and RPV should reflect that denominator.
Pick a window long enough to cover at least one full purchase cycle for your category — typically 14 to 28 days for fashion and beauty, longer for considered electronics. For test reads, match the window to the experiment duration exactly, including weekends. The companion RPV Calculator handles the arithmetic and the CR × AOV split in one view.
Why RPV beats conversion rate alone
Conversion rate treats a €30 order and a €300 order identically. RPV doesn't. That single difference reshapes which experiments you green-light and which you kill — see RPV vs Conversion Rate for the full side-by-side on test grading.
The classic trap: a discount-stacking change lifts CR by 8% but drops AOV by 10%. CR-led teams ship it; RPV-led teams catch that revenue per session actually fell ~2.8% and roll it back. Over a year on €5M of sessions that's the difference between a win and a quiet six-figure loss.
Don't grade tests on CR if AOV can move
Any test touching pricing, shipping thresholds, bundles, upsells, cross-sells, or recommendation logic can shift AOV. For those tests, grading on conversion rate alone is structurally broken — read RPV as the primary metric and treat CR and AOV as diagnostic splits underneath it.
The CR × AOV decomposition
RPV = CR × AOV is the most useful identity in e-commerce analytics. When RPV moves, the decomposition tells you which lever moved it — and that determines what you test next. A lift driven by AOV points you toward merchandising and bundle work; a lift driven by CR points you toward funnel and trust work. RPV vs AOV walks through how to separate the two cleanly.
The trade-off curve below shows why the decomposition matters: holding RPV at €2.40, a store can land there with 2% CR × €120 AOV, 3% × €80, or 4% × €60 — three completely different businesses, three different roadmaps. Session Value, the same metric under a different name, is what dashboards usually expose.
CR and AOV combinations that produce the same RPV (€2.40)
Revenue Per Visitor FAQ
RPV (Revenue Per Visitor) is total revenue divided by total sessions over a chosen window. It tells you the average revenue a single visit to your store generates, blending conversion rate and average order value into one number.
Divide total store revenue by total sessions for the same period. For example, €120,000 of revenue across 50,000 sessions gives an RPV of €2.40. Use the same source for both numerator and denominator — GA4, Shopify analytics, or your warehouse — but don't mix sources.
Sessions. A user who visits three times is three opportunities to convert, and your funnel touches each visit independently. Using users inflates the metric and disconnects it from how testing tools count exposure.
Conversion rate measures whether a visit buys; RPV measures both whether it buys and how much it spends. Any test that can move AOV — bundles, shipping thresholds, upsells — must be graded on RPV, not CR alone. See RPV vs Conversion Rate for the full breakdown.
AOV is revenue divided by orders — it only counts converting sessions. RPV is revenue divided by all sessions, converting or not. A store can grow AOV while RPV stays flat if conversion rate drops in lockstep.
Yes, in practice. GA4 exposes the metric as 'Average purchase revenue per user' and some tools as 'Session value' — the math is identical when applied to sessions. RPV is the term most CRO teams use.
Apparel and accessories typically sit at €1.50–€4.00; beauty and skincare at €2.00–€6.00; home and electronics swing wider, €3.00–€15.00 depending on AOV. The absolute number matters less than the trend and the segment splits.
Long enough that the variance from large orders evens out — typically two to four full weeks, and never less than one weekly cycle. RPV has heavier-tailed variance than CR because a single big order can swing the average, so don't call winners early.
For P&L-accurate reads, yes — use net revenue after refunds. For real-time test monitoring, gross revenue is fine because refunds arrive after the test window closes. Pick one definition and apply it consistently across dashboards.
Split by device (mobile usually trails desktop by 30–50%), channel (paid social often has the lowest RPV, email the highest), and landing-page template. The gaps between segments are where the biggest CRO opportunities sit.
Test ideas before you ship them
Run unlimited A/B tests, attach hypotheses to outcomes, and build a searchable archive of what works — and what doesn't.