AOV, or average order value, is total revenue divided by number of orders over a period. It is useful for setting thresholds such as free-shipping levels and fraud-check floors. It is less useful as a health metric than it looks, because order values are usually skewed and a mean can sit somewhere almost no actual order does.
In more detail
The distribution problem is worth understanding. If most orders are around 30 and a handful are 400, the mean lands somewhere in between and describes neither group. Decisions taken from that number will suit nobody.
The median is often more representative for operational thresholds. If you are choosing a value floor for verifying flagged orders, you want the level most orders actually sit near, not a figure pulled upward by outliers.
AOV is genuinely useful as a threshold input and as a trend. Watching it move after a bundling change or a free-shipping threshold tells you something real. Reading a single month's figure as a performance verdict does not.
How it is calculated
AOV = total revenue ÷ number of orders
Worked example
In a month you take 90,000 across 1,500 orders, giving an AOV of 60.
Look at the spread before using it. If 1,400 orders were between 25 and 45 and 100 were wholesale orders above 500, then 60 describes no typical customer. A free-shipping threshold set at 60 would be set above almost every real order.
The figures above are invented to make the arithmetic legible. They are not a benchmark and should not be cited as one.
Where people get it wrong
- Using the mean where the median is more representative, especially on skewed distributions.
- Comparing across periods with different promotions without noting it.
- Raising it by removing cheap products, which raises the number and can lower total revenue.
- Treating it as a goal. It is an input to decisions, not an outcome worth maximising on its own.
Common questions
What is the AOV formula?
Total revenue divided by number of orders over the same period. Exclude shipping and tax consistently, or include them consistently, and note which you did.
Should I use mean or median order value?
For operational thresholds such as a fraud-check floor, the median usually represents your typical order better, because a few large orders pull the mean somewhere no real order sits.
Is raising AOV always good?
No. Removing low-priced products raises AOV and can reduce total revenue and new-customer acquisition at the same time. It is an input, not a target.