Revenue Funnels
Revenue funnels score each stage by money moved, not visitors converted — surfacing the AOV trades that count-based funnels miss.
Revenue Funnels
A funnel view that weights each stage by revenue captured, not visitor counts — so AOV and margin trades are visible end-to-end.
A revenue funnel reframes the classic step-by-step conversion chart in money. Instead of asking "what percentage of sessions reached checkout?", it asks "how much revenue passed through this step, and how much leaked out?". The same drop-off looks very different when a high-AOV cohort exits versus a bargain-hunter cohort.
The technique matters most when you're comparing variants that move conversion rate and order value in opposite directions — a bundle upsell that lowers checkout completion by four points but raises AOV by twenty. A count-based funnel calls that a loss; a revenue funnel often calls it a win.
Classic funnel analytics counts heads at each step: 100,000 sessions → 38,000 product views → 9,500 add-to-carts → 3,200 checkouts → 2,400 orders. Each transition gets a conversion rate, and you optimise the worst one. The problem is that every session is treated as identical, when in reality a session that lands on a €240 jacket is worth ten times one that lands on a €24 hair tie.
Revenue funnels solve this by assigning each session the expected revenue of its trajectory — usually weighted by cart value, predicted LTV, or category-level AOV. The same five steps now show money flowing instead of bodies. A 2% step drop-off concentrated in your highest-margin SKUs is a five-alarm fire; the same drop-off across €15 add-ons barely registers.
Revenue Conversion Rate (step) = Σ(revenue exiting step toward next step) / Σ(revenue entering step)
R_in
Revenue entering step
Sum of expected cart value (or predicted order value) across all sessions that reached this step.
R_out
Revenue exiting step
Sum of expected cart value across sessions that advanced to the next step.
RCR
Revenue Conversion Rate
Share of monetary value retained from one step to the next, rather than share of sessions retained.
A Shopify apparel store runs an A/B test on its product page. Variant A keeps the simple layout; Variant B adds a 'complete the look' bundle module that pushes higher-ticket items. Looking at the product-view → add-to-cart step:
Variant A sessions reaching PDP: 10,000
Variant A add-to-carts (CVR 26%): 2,600 carts, average €68 = €176,800
Variant B sessions reaching PDP: 10,000
Variant B add-to-carts (CVR 23%): 2,300 carts, average €94 = €216,200
→ Variant A wins on count CVR (26% vs 23%). Variant B wins on revenue: €216,200 vs €176,800, a 22% lift in monetary throughput.
A count-based funnel would have killed Variant B. The revenue funnel surfaces that the higher AOV more than compensates for the conversion drop — and that's the variant you ship.
The benchmarks below show why revenue weighting changes the story. Notice that revenue conversion rate often runs above or below count conversion rate depending on whether expensive or cheap orders are more likely to make it through each step. Cart abandonment hits high-AOV orders harder; product-page bounce hits low-AOV impulse buys harder.
Count vs revenue conversion rate by funnel step, by vertical (Shopify stores, €1M–€15M GMV)
| Funnel step | Apparel — count CVR | Apparel — revenue CVR | Beauty — count CVR | Beauty — revenue CVR | Electronics — count CVR | Electronics — revenue CVR |
|---|---|---|---|---|---|---|
| Session → product view | 42% | 45% | 51% | 48% | 38% | 44% |
| Product view → add-to-cart | 9% | 11% | 13% | 12% | 6% | 8% |
| Add-to-cart → checkout start | 55% | 51% | 62% | 60% | 48% | 42% |
| Checkout start → order | 68% | 63% | 74% | 71% | 59% | 52% |
| End-to-end (session → order) | 1.6% | 1.8% | 2.9% | 2.5% | 0.6% | 0.8% |
Electronics is the clearest case: revenue CVR sits above count CVR everywhere, because higher-ticket items pull more committed buyers through. Beauty shows the opposite at checkout — small impulse orders complete more often than larger replenishment baskets. Treat both views as a pair; whichever step shows the widest gap between count and revenue CVR is where your funnel is silently bleeding money.
Revenue funnel FAQ
A regular funnel counts sessions at each step. A revenue funnel sums money — either actual cart value or expected order value — at each step. Same shape, different unit. The two views often disagree on which step is the biggest problem, and revenue is usually the one that matches your P&L.
Use revenue funnels whenever AOV varies meaningfully across products, segments, or variants — which is almost always true for stores with more than one price tier. Count funnels are still useful for diagnosing UX friction that hits everyone equally, like a broken form field.
Two common approaches: use the cart value at the session's last recorded step (so a session that abandoned at checkout gets the cart total), or use category-level expected order value based on what the session viewed. The cart-value method is more accurate; the expected-value method works earlier in the funnel.
Yes, but you weight by first-order value or predicted LTV instead of cart value. Subscription funnels with a high upfront commit step (annual plan upsell) almost always need revenue weighting — the step that loses the most sessions is often the step that captures the most revenue from the survivors.
It's one view inside funnel analytics, alongside count funnels, time-to-convert funnels, and segmented funnels. Most teams default to the count view and only build a revenue view when an A/B test result looks suspicious — which is backwards. Build both from day one.
Yes, and it's common with upsells, bundles, and premium-tier nudges. The variant pushes some price-sensitive buyers away (lower CVR) while convincing the remaining buyers to spend more (higher AOV). Revenue funnels are the only honest way to call that test.
AOV is the lever that makes the revenue funnel diverge from the count funnel. If every order were the same size, the two views would be identical. The wider the AOV distribution across your catalogue, the more revenue funnels will tell you a different story.
Revenue CVR if margins are similar across SKUs, because that's the closest proxy to profit. Count CVR if you're diagnosing a specific friction point that affects all buyers equally. The smart default is revenue CVR for ship/no-ship decisions and count CVR for root-cause analysis.
Four to six is the sweet spot: landing, product engagement, cart, checkout, and order. More steps fragment the data and make per-step revenue figures statistically noisy. If you need finer granularity, segment the funnel by traffic source or device instead of adding micro-steps.
If you're on Shopify, WooCommerce, or Magento, your platform already attaches cart value to session events — you just need a tool that reads those events and aggregates them by step. Most teams set this up alongside their existing GA4 implementation rather than replacing it.
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.