Average order value, usually shortened to AOV, is the average amount a customer spends in a single order. You calculate it by dividing total revenue by the number of orders over the same period. If an online shop takes £42,000 from 1,200 orders in a month, its AOV is £35.
How average order value works
The formula is simple, but the definition of revenue is not, and UK retailers report AOV in more than one way. The £35 above might include VAT and a £3.95 delivery charge, as the customer paid it. Strip out VAT at 20% and the delivery charge, and the same orders are worth about £25.88 each in net product revenue. Both figures are valid; mixing them is not. Decide which version you use, write it down, and make sure every report follows it.
Analytics tools add another layer. GA4 reports average purchase revenue from the value your site sends with each purchase event, and whether that value includes VAT, delivery and discounts depends on how the tracking was set up. It is worth checking against your platform’s sales report before relying on it.
AOV is driven by a few levers:
- Price: what each item costs.
- Units per order: how many items people buy at once, influenced by cross-sells and bundles.
- Trading up: choosing a better version, encouraged by upsells.
- Delivery rules: a free delivery threshold nudges shoppers to add a little more.
- Discounts, which lower AOV unless they are tied to a minimum spend.
Because it is a mean, a few very large orders can pull it up. Looking at the median and the spread of order values alongside it gives a truer picture of a typical basket.
Why it matters
AOV decides how much you can afford to spend to win an order. If your margin on a £35 order is £10 after product and delivery costs, a paid campaign that costs £12 per sale loses money on every first order, however good the conversion rate looks. Raise AOV to £45 with the same costs per order and the picture changes. Many costs, such as picking, packing and payment fees, are charged per order, so larger orders are usually more profitable per pound of revenue.
It also explains movements in revenue. When sales rise, AOV tells you whether more people bought or the same people bought more, and those call for different decisions.
Common mistakes
- Comparing AOV across tools that treat VAT, delivery or discounts differently.
- Ignoring refunds and returns, which can make AOV look healthier than the money that stays in the bank.
- Raising AOV with discounted bundles that cut margin by more than they add.
- Setting the free delivery threshold far above typical baskets, so few customers ever reach it.
- Averaging trade and consumer orders together when they behave completely differently.
How to act on it
Agree one definition, ideally net of VAT and refunds for decision-making, and keep the gross figure for comparison with platform reports. Segment AOV by channel, device, new versus returning customer and product category, because the average across the whole shop hides most of what is useful.
Then test the levers one at a time: a threshold set a little above the current median basket, relevant add-ons on product pages, and bundles priced to protect margin. Judge each change on profit per order, not AOV alone. In my performance marketing work, AOV and margin set the target cost per sale for every campaign, so this is usually one of the first numbers I ask a retailer for.