How to use AOV Measurement
How to actually compute average order value for an online store — the discount, return, segment, and time-window choices that decide what the number means.
AOV Measurement
The methodology for calculating average order value — the discount, return, segment, and time-window choices that decide what the number means.
AOV measurement is the set of methodology decisions behind a deceptively simple metric: take revenue, divide by orders, get average order value. In practice every term in that fraction is contested. Do you count discounts, shipping, taxes, gift cards? Do you net out returns, and on what lag? Do you include subscription renewals or only new orders? Do you blend new and returning customers, or report them separately?
Getting AOV measurement right is what makes the number trustworthy across pricing tests, merchandising changes, and channel reports. The framework below standardises how a Shopify, WooCommerce, or Magento store should compute, segment, and govern AOV so every downstream decision — bundle pricing, free-shipping thresholds, paid bids — sits on the same definition.
Most stores quote a single AOV number on a weekly report. That number hides three or four methodology choices, and changing any of them moves the figure by 5-20%. A merchandising team and a finance team using the same dashboard can argue for an hour without realising they're computing different metrics.
This guide walks through the four decisions that matter: gross-vs-net of discounts and returns, segment cuts that reveal what blended AOV hides, time-window selection, and the governance that keeps the definition stable as your store evolves. The AOV Calculator handles the arithmetic; this page covers the choices that go into the inputs.
Gross vs net: which AOV are you actually reporting?
The headline choice is what counts as "revenue" in the numerator. Gross AOV uses order subtotal before any deductions. Net AOV strips out discounts, shipping, taxes, and refunded returns. The gap between the two is routinely 15-25% on a store running promotions and free shipping.
For merchandising and CRO decisions, use a discount-inclusive but tax-exclusive figure — that's what the customer actually saw at checkout and what your bundle or upsell test moved. For finance and margin work, use net AOV after returns and refunds, because that's the revenue you keep. Reporting both side-by-side stops the cross-team argument before it starts.
Shipping is the sneakiest line item. Shopify's default analytics rolls shipping revenue into order total, so a store offering paid expedited shipping looks like it has a higher AOV than a competitor offering free shipping at the same product price. Decide once whether shipping counts, document it, and apply it everywhere.
The returns lag trap
Returns hit the books 14-45 days after the order. If you compute last-week's AOV net of returns on Monday morning, you're netting out almost nothing — the returns for those orders haven't happened yet. Either report gross AOV in real time and net AOV on a 60-day trailing window, or label the real-time number clearly as "pre-return".
Segment cuts that reveal what blended AOV hides
A sitewide AOV is a weighted average across customer types, channels, devices, and product mixes. The blend can stay flat for months while the underlying segments move in opposite directions — a classic Simpson's paradox setup. Cutting AOV by segment is the only way to see what's actually happening.
The four cuts that matter most: new vs returning customer (returning typically 30-60% higher), acquisition channel (paid social usually lowest, email and direct highest), device (desktop 20-40% above mobile on apparel and home goods), and product category. If your blended AOV drops while every segment AOV rises, your traffic mix shifted — not your store performance.
How AOV variants compare on the same orders
Same store, same week, six legitimate AOV numbers ranging from €52 to €94. The point isn't that one is right and the others wrong — they answer different questions. The point is that quoting a single "AOV is €X" without naming the cut is what fuels bad pricing decisions.
Time windows: rolling, calendar, and cohort
AOV is volatile week-to-week, especially below €5M revenue where a single high-AOV B2B-style order can shift the figure 5%. Pick a window that matches the decision you're making. Weekly AOV is for spotting anomalies; 28-day trailing is for steering merchandising; quarterly is for board reporting and benchmarking.
Calendar months are convenient but distort comparisons — February has three fewer days than January, and Black Friday warps any November number. Use 28-day rolling windows for trend work and reserve calendar months for finance reconciliation. For experiment readouts, use the exact experiment window and nothing else; padding it before or after biases the result.
Typical AOV ranges by platform and vertical (gross, discount-inclusive, ex-shipping)
| Vertical | Shopify | WooCommerce | Magento |
|---|---|---|---|
| Apparel & fashion | €55-95 | €45-85 | €70-120 |
| Beauty & personal care | €35-65 | €30-55 | €45-80 |
| Home & furniture | €90-180 | €80-160 | €140-280 |
| Food & beverage (DTC) | €30-55 | €28-50 | €40-70 |
| Consumer electronics | €110-220 | €95-200 | €160-320 |
| Health & supplements | €45-80 | €40-75 | €55-95 |
Magento stores skew higher because the platform is more common in B2B-adjacent and high-AOV verticals; that's a selection effect, not a platform capability. Use these as orientation, not targets — your own segment cuts against your historical data tell you more than any cross-store benchmark.
Governance: stabilising the definition
Once you've made the methodology choices, the work is keeping them stable. Three things tend to drift the definition over time: a new payment method (gift cards, BNPL) that the analytics layer treats differently, a tax or shipping rule change in Shopify Markets that quietly shifts what's in the order total, and the introduction of subscriptions, which can double-count renewals as new orders.
Write the AOV definition down — one page, ideally in your analytics workspace — listing exactly what counts as revenue, what counts as an order, and which segments are reported. Review it every quarter and whenever you launch a new channel or product type. The AOV vs Conversion Rate trade-off discussion only makes sense if both numbers are computed against fixed definitions.
Where Metricuno fits
Metricuno imports your historical GA4 and order data on day one, so you can see AOV cut by every segment without rebuilding a tracking plan first. Definitions are explicit and versioned — change the rule, see the impact on past periods, and ship it to the dashboard once everyone agrees.
AOV measurement FAQ
AOV equals total revenue divided by number of orders over the same window. The formula is trivial; the decisions about what counts as revenue (gross vs net of discounts, shipping, returns) and what counts as an order (new only, including subscriptions, etc.) are what determine the number you actually report.
Exclude tax always — it's not your revenue. Shipping is a judgement call: include it if customers pay for shipping and you want a customer-facing figure; exclude it if you want a product-only number that's comparable across stores with different shipping policies. Pick one and apply it consistently.
Report the discount-inclusive figure (what the customer paid after promo codes) for CRO and merchandising work, because that's the order value your test actually moved. Report the pre-discount figure only if you need to isolate the impact of discount depth itself, which is a separate analysis.
Both, labelled clearly. Gross AOV (pre-returns) is for real-time dashboards and weekly trends. Net AOV (post-returns) is for margin and finance reporting, computed on a 60-day trailing window once returns have largely settled. Quoting one without specifying which causes the most cross-team confusion.
Daily for monitoring, weekly for operational review, monthly for executive reporting. Use a 28-day rolling window for trend analysis rather than calendar months, which contain different numbers of days and seasonal artefacts that distort week-over-week reads.
Almost always because each tool includes different line items by default — Shopify includes shipping, GA4's purchase event may or may not depending on your dataLayer setup, and a BI tool reflects whatever your data team wrote. Reconcile by picking one source of truth (usually the order database) and documenting the exact SQL that produces the reported number.
At minimum: new vs returning customer, acquisition channel, device, and top-level product category. New-vs-returning is the highest-signal cut because returning customers typically spend 30-60% more per order, and a shifting traffic mix between them can move blended AOV without any underlying behaviour change.
AOV is per order; ARPU is per visitor or per customer over a window. A store with 1,000 visitors, 50 orders, and €60 AOV has €3 revenue per visitor and €60 AOV. ARPU bakes in conversion rate; AOV isolates basket size. Use both — the AOV vs conversion rate trade-off only resolves when you watch both numbers together.
Subscription renewals are typically lower-AOV than first orders (often just one SKU) and high-frequency, so blending them in drags down headline AOV without indicating any merchandising problem. Report subscription orders as a separate segment, or exclude renewals from AOV entirely and report subscription performance with its own metrics (LTV, retention).
There's no universal target — it depends on vertical, price point, and margin structure. A €50 AOV with 60% margin beats a €120 AOV with 18% margin. Benchmark against your own historical trend and against your contribution margin requirements rather than against an industry average that may sit on completely different product economics.
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.