RPV Diagnostic Checklist
A seven-step checklist for diagnosing why revenue per visitor is below benchmark — segment the traffic, isolate the broken half of the equation, and map the gap to the right lever.
RPV Diagnostic Checklist
A structured walkthrough for finding the root cause of a low or falling revenue-per-visitor number before you pick a fix.
The RPV Diagnostic Checklist is a sequence you run when revenue per visitor is below benchmark or trending down. It forces you to segment traffic by device, source, and landing page, then decompose RPV into its two factors — conversion rate and average order value — so you know which half of the equation is broken before you touch anything.
The goal isn't to fix RPV in this step. The goal is to localise the gap precisely enough that the next step, choosing an optimisation lever, isn't guesswork. A site-wide RPV drop driven entirely by mobile paid traffic landing on one collection page is a completely different problem from a uniform AOV decline across all sources — and they need different experiments.
Revenue per visitor is the cleanest single number for site health on a Shopify or WooCommerce store, but it's also the easiest to misread. A 12% RPV drop month-over-month could be a checkout regression, a paid-traffic mix shift, a seasonal AOV slump, or three small things stacking. You won't know until you cut the data.
Run this checklist in order. Each step narrows where the problem lives, so by the time you reach the lever-selection stage covered in RPV Levers, you're picking from three candidate fixes, not thirty.
The most common diagnostic mistake
Teams look at the blended RPV number, see it's down, and immediately run a checkout test. Roughly half the time the actual culprit is upstream — a traffic-mix shift toward a lower-intent channel — and the checkout test will either lose or be inconclusive because the problem was never on that page. Always segment before you experiment.
The seven-step RPV diagnostic
Step 1 — Decompose RPV into CR × AOV. Pull both numbers for the affected window and the comparison window side by side. If conversion rate is flat and AOV fell, you have an AOV-side problem (discount mix, product mix, bundle adoption). If AOV is flat and conversion rate fell, you have a CR-side problem (friction, trust, traffic quality). If both moved, weight them: a 5% CR drop on a €70 AOV is worth more than a €3 AOV slide.
Step 2 — Segment by device. Mobile and desktop RPV diverge sharply on most apparel and beauty stores; a blended drop often hides a mobile-only collapse after a theme update or a Core Web Vitals regression. Step 3 — Segment by source. Compare RPV across paid social, paid search, organic, email, and direct. A drop concentrated in one channel almost always points to a traffic-quality or landing-page mismatch, not a site-wide issue.
Step 4 — Segment by landing page. Sort top entry pages by sessions and look at RPV per page. One underperforming collection or PDP can drag the blended number visibly when it's a top-five entry point. Step 5 — Check the funnel stages. Walk add-to-cart rate, cart-to-checkout rate, and checkout completion rate against the prior period. The stage that moved is where the friction sits — and it tells you whether to test PDPs, cart UX, or checkout.
Step 6 — Audit AOV inputs if the AOV side is the problem. Check units-per-order, average unit price, discount rate, and the share of orders containing a bundle or upsell. A drop in units-per-order is usually a merchandising or cross-sell issue; a drop in average unit price is usually a discount or product-mix issue. Step 7 — Confirm the finding with a clean A/B-test hypothesis before you ship. Your diagnostic should produce one sentence: "RPV is down because [segment] is converting at [rate] on [page] due to [mechanism], and we expect [lever] to recover it." If you can't write that sentence, loop back to step 2.
Frequently asked questions
Anything beyond normal weekly noise — typically a sustained 5%+ deviation over a two-week window, or a single-week drop greater than two standard deviations of your trailing 12-week RPV. Smaller wiggles are usually traffic-mix noise and self-correct.
Decompose both at the same time in step 1, then prioritise whichever moved more in revenue terms. A 0.3-point conversion-rate drop on a high-traffic store usually outweighs an AOV slide, but the math depends on your baseline. Don't pick by gut — multiply each delta by the other factor and compare.
The checklist tells you where the gap lives; RPV Levers tells you which mechanism to pull to close it. You shouldn't pick a lever before completing the diagnostic, because the same RPV drop can have three different correct levers depending on which segment is bleeding.
Use the Traffic Acquisition report with Session source / medium as the dimension and add Purchase revenue and Sessions as metrics, then compute revenue ÷ sessions per row. Most teams export to a sheet because GA4's calculated-metric UI is limited. A historical GA4 import into a CRO tool will give you the same cut without manual work.
Mobile typically converts at 40-60% of desktop's rate while AOV is also slightly lower, so blended mobile RPV often lands at half of desktop. That's structural — the diagnostic question is whether the gap widened recently, not whether it exists.
Yes, and it's the single most under-diagnosed cause. If you shifted budget into a broader Meta audience or a new prospecting campaign, you're buying lower-intent sessions that convert at half the rate of your retargeting pool. Site-wide RPV falls even though nothing on the site changed.
Checkout completion rate dropping while add-to-cart and cart-to-checkout rates hold steady is the signature. Other tells: a spike in initiated-checkout events without matching purchase events, a higher rate of payment-error events, or a sudden mobile-only gap after an app or theme update.
Once a month as a 30-minute review, and immediately any time blended RPV moves more than 5% week-over-week. Catching a paid-traffic-mix issue in week one is much cheaper than catching it in week four after you've already spent on a losing campaign.
It counts as context, not a fix. Compare against the same period last year, not just last month, and adjust for any traffic or promo differences. If year-over-year RPV is flat or up, your seasonal dip is normal; if it's down YoY, the diagnostic still applies.
RPV Optimization has three phases: diagnose, prioritise, and test. This checklist is the diagnose phase. It produces a localised hypothesis that feeds into lever selection and then into a structured A/B-test backlog — the rest of the optimisation loop only works if this step is done properly.
Get an AI expert review of your site
Paste your URL — Metricuno's AI runs the same heuristic checks a senior CRO consultant would, scoring your page and prioritising the fixes that'll move conversion fastest.