RPV vs AOV
RPV measures revenue across every visitor; AOV measures it across buyers only. The difference matters because tactics that lift AOV can quietly suppress conversion rate and leave total revenue flat — and only RPV catches it.
RPV vs AOV
RPV (revenue per visitor) measures revenue across all sessions; AOV (average order value) measures it only across orders — the gap reveals conversion-rate effects AOV alone hides.
Revenue per visitor and average order value look similar on a dashboard but answer different questions. AOV asks: when someone buys, how much do they spend? RPV asks: across every visitor — buyers and non-buyers — how much revenue did the store generate?
The distinction matters because most AOV-lift tactics (free-shipping thresholds, bundle upsells, minimum-spend gates) change buying behaviour. They can lift AOV by 8% while suppressing conversion rate by 9%, leaving RPV flat or down. Only RPV captures that trade-off in a single number, which is why experienced CRO teams treat RPV as the primary revenue KPI and AOV as a supporting diagnostic.
The formulas make the difference obvious. AOV = total revenue / number of orders. RPV = total revenue / number of visitors (or sessions, depending on your setup). Same numerator, different denominator — and that denominator change is where most reporting mistakes hide.
Because AOV ignores non-buyers entirely, it cannot tell you whether a change pushed people out of the funnel. A €70 free-shipping threshold can lift AOV from €58 to €64 while turning a chunk of €40 shoppers into abandoners. The order book looks healthier; the revenue line doesn't move. RPV catches that immediately.
How an AOV-lift tactic can leave RPV flat — a worked Shopify apparel example
| Scenario | Visitors | Conv. rate | Orders | AOV | Revenue | RPV |
|---|---|---|---|---|---|---|
| Baseline (no threshold) | 10,000 | 2.80% | 280 | €58.00 | €16,240 | €1.62 |
| €70 free-shipping threshold | 10,000 | 2.55% | 255 | €64.00 | €16,320 | €1.63 |
| €90 free-shipping threshold | 10,000 | 2.30% | 230 | €71.50 | €16,445 | €1.64 |
| Bundle upsell, no threshold | 10,000 | 2.78% | 278 | €63.20 | €17,570 | €1.76 |
Look at rows two and three. AOV jumps 10-23% and the team celebrates. RPV barely moves because the threshold filtered out price-sensitive buyers. Row four — a bundle that doesn't gate checkout — is the only variant that actually grows the business, and RPV is the metric that flags it.
When to use each metric
Use AOV when the question is about basket composition: did the new product page upsell work, are bundles being taken, did the cart-page recommender shift mix? AOV isolates buyer behaviour from traffic-quality noise, which makes it a clean diagnostic for merchandising and pricing tests.
Use RPV when the question is about total revenue impact: should we ship this variant, is the new homepage better, did the checkout redesign actually pay off? RPV is the only single metric that combines conversion rate and order value, so it's the right primary KPI for almost any site-wide A/B test on the path to purchase.
The threshold trap
Any tactic that conditions a benefit on a minimum spend — free shipping over €X, gift with €Y purchase, tiered discounts — will lift AOV almost by construction. That doesn't mean it lifted revenue. Always read RPV alongside AOV on these tests, and treat an AOV win with flat RPV as a neutral result, not a winner.
How CR, AOV and RPV move together
The relationship is multiplicative: RPV = conversion rate × AOV. That identity is what makes RPV so useful — a single number absorbs both inputs, so you can't be fooled by one moving while the other compensates in the wrong direction.
It also tells you where to look when RPV moves. If RPV is up and AOV is flat, you improved conversion rate. If RPV is up and conversion rate is flat, you improved basket size. If RPV is up and both moved, you found something rare — protect it.
AOV up, conversion rate down, RPV barely moves
Frequently asked questions
RPV, for most decisions. It captures both conversion rate and order value in one number, so it can't be gamed by tactics that lift one at the expense of the other. AOV stays useful as a diagnostic for basket-composition tests.
Because AOV is averaged across buyers only. When a threshold or premium experience filters out price-sensitive shoppers, the buyers who remain spend more on average — but there are fewer of them. AOV climbs while revenue stays flat or falls.
Sessions is the more common choice because it aligns with how most A/B testing tools bucket traffic. Unique visitors smooths out repeat-visit noise but undercounts effort across multi-session purchase journeys. Pick one and stay consistent — the absolute number matters less than the trend.
Yes, when it lifts AOV faster than it suppresses conversion rate, RPV genuinely moves. The point is you can't tell from AOV alone — you have to look at RPV to know whether the threshold paid for itself.
Set RPV as your primary metric in the experiment platform and track conversion rate plus AOV as secondaries. You'll need slightly larger samples than a conversion-rate test because revenue is a noisier metric, but the directional signal is much cleaner for revenue-impact decisions.
It works, but variance is high — a furniture or electronics store with a €600 AOV will see RPV swing wildly across small samples. Run tests longer, or segment by product category, so the average isn't dominated by one or two outlier orders.
It varies massively by vertical and traffic source — apparel typically sits between €1.20 and €3.00, beauty between €0.80 and €2.50, and home goods can run higher. The benchmark that matters is your own trend over time and across experiment variants, not an industry average.
ARPU (average revenue per user) is a subscription-business metric measured over a billing period, usually a month. RPV is a per-session e-commerce metric measured per visit. They share the 'average revenue per person' framing but operate on completely different timescales.
Report both, but lead with RPV when the question is about site or experiment performance. AOV is the right answer when the question is about merchandising or pricing strategy. Knowing which one to lead with is half of being good at this.
GA4 surfaces total revenue and conversion rate separately but doesn't compute RPV as a standard metric — you have to derive it (revenue / sessions) in Explorations or a connected analytics layer. Most CRO platforms calculate it natively at the experiment level.
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.