Average order value
What a typical order is worth, after discounts and returns.
01 What it is
Why anyone looks at this number
What a typical order is worth, after discounts and returns.
It is one of only three levers on revenue — more customers, more often, or more per order. The third is usually the cheapest to move and the least worked.
02 The formula
How it is worked out
Average order value = net_revenue ÷ orders grain : month × channel × segment unit : currency per order source : Order management, joined to billing for net values
Decide once what an order is. Basket, shipment and invoice are three different denominators, and the web platform, the warehouse and finance will each default to their own. Split orders and part-shipments quietly deflate this number if the rule is not fixed.
03 Worked example
The same number, with real inputs
| Net revenue | €1.28m |
| Orders | 18,700 |
| Units per order, this year vs last | 2.4 vs 2.4 |
| Calculation | 1,280,000 ÷ 18,700 |
| Result | €68.40, down 3.9% |
Basket size is unchanged, so the fall is entirely price. This is a discounting story being read as a merchandising one.
04 What moves it
Four things that actually change this number
Basket size
Units per order, which is a merchandising and bundling question.
Price and discount
The same basket sold for less.
Mix
Channel and category mix moves the average without anything changing underneath.
Thresholds
Free-delivery and volume-break points shape the whole distribution.
05 Where the number lives
The system, the record and the fields
| System of record | Key record | Fields you need |
|---|---|---|
| Order management, joined to billing for net values | Order joined to Customer and Channel | order_id, order_date, gross_amount, discount_amount, return_amount, channel, customer_id |
Decide once what an order is. Basket, shipment and invoice are three different denominators, and the web platform, the warehouse and finance will each default to their own. Split orders and part-shipments quietly deflate this number if the rule is not fixed.
06 How it goes wrong
Three ways this metric misleads people
Gross rather than net
Discounts and returns are excluded and the number flatters every channel that discounts most.
Fix: Calculate on net revenue, after returns.Blending channels
An assisted sale and a self-serve reorder are averaged into a value that describes neither.
Fix: Report by channel, always.Reading only the average
Two clusters of orders around €20 and €160 produce a mean nobody has ever placed.
Fix: Publish the median and the deciles next to the mean.08 Questions
Frequently asked
Should returns be deducted?
Yes, in any business where returns are material. An AOV that ignores a 6% return rate overstates every downstream calculation built on it, including lifetime value.
How does AOV relate to lifetime value?
It is the first term in it. Lifetime value is AOV times purchase frequency times margin times expected lifetime, so an AOV built on gross revenue inflates the whole chain.
One definition, everywhere it is used
SCIKIQ stores this metric once and serves it to every dashboard, board pack and agent that asks.