Average order value, usually shortened to AOV, is the average value of an order over a defined set of orders and time period. A generic calculation is:
Average order valueWhat is average order value (AOV)? Useful average, incomplete picture
The formula looks simple, but the operational meaning depends on what the system counts as "order revenue" and "orders." Shopify, for example, currently defines the AOV in its sales reports as (gross sales - discounts) / orders and notes that this metric excludes post-order adjustments such as edits and exchanges. That is a platform definition, not a universal accounting law.
The most useful mental model is: AOV summarizes a distribution; it does not describe the typical customer by itself.
AOV answers a portfolio question, not a customer question
AOV is strongest when the question is something like:
- How much revenue does an order contribute on average?
- Did a merchandising or upsell program change basket size?
- How does order value differ by channel, customer cohort, country, or campaign?
- What happens to revenue if order volume stays constant but average basket value rises?
It is weaker for questions like:
- What does a typical customer spend?
- How many customers buy only one item?
- Are high-value orders coming from a small VIP segment?
- Did an apparent AOV increase come from price inflation rather than larger baskets?
For those, look at order-value distributions, median order value, items per order, product mix, customer segments, or cohort behavior alongside AOV.
The denominator matters
AOV can move even when customer behavior has not improved. Imagine the same $30,000 of recognized order revenue across:
- 1,000 orders → $30 AOV
- 750 orders → $40 AOV
The second period has higher AOV, but fewer orders. Revenue is unchanged. So "AOV increased" is not the same as "the business grew." AOV should be interpreted with order count and revenue, not celebrated in isolation.
The revenue definition matters too
Different analytics systems can treat discounts, refunds, taxes, shipping, canceled orders, exchanges, edits, and currency conversion differently. Shopify's current default sales-report definition is one example: it uses gross sales minus discounts and excludes post-order adjustments for its AOV metric.
When comparing 2 dashboards, first verify that the numerator and denominator match. A $72 AOV in one system and $68 in another can both be internally correct if they count different revenue components or order states.
AOV versus revenue per customer
AOV uses orders as the denominator. Revenue per customer uses customers. If one customer places 3 $50 orders and another places one $50 order:
- total revenue = $200
- orders = 4
- customers = 2
- AOV = $50
- revenue per customer = $100
Those metrics answer different questions. A retention program could increase orders per customer while AOV stays flat, yet still materially improve revenue per customer.
AOV versus basket size
AOV is measured in money. Basket size can refer to number of items or units in an order. AOV can rise because:
- customers buy more units
- customers buy higher-priced products
- prices increase
- discounts decrease
- product mix shifts
- more high-value customers order
If the objective is to understand why AOV moved, decompose the change rather than treating the headline metric as the cause.
Use segmentation before inventing a story
A storewide AOV can conceal very different patterns. Suppose overall AOV rises from $58 to $66. Before concluding that a new upsell is working, compare:
- customers exposed vs not exposed to the upsell
- new vs returning customers
- paid vs organic acquisition
- device type
- country/currency
- full-price vs discounted orders
- relevant product categories
AOV is descriptive. Causal attribution requires a stronger comparison design than a before/after chart alone.
AOV targets should respect margin
Increasing AOV is not automatically profitable. An offer can increase basket value while reducing gross margin through heavy discounts, free gifts, or expensive fulfillment. Example:
- Baseline: $60 order with $30 product cost → $30 gross profit before other costs
- Promotion: $80 order with $48 product cost and $12 discount that would not otherwise have been given
The larger order may or may not be better after the full economics are defined. Track AOV together with gross profit, margin, discounting, returns, and any incremental fulfillment costs that the promotion changes.
Practical ways to use AOV
AOV is often helpful for setting thresholds and evaluating merchandising tactics:
- free-shipping thresholds
- bundles
- quantity breaks
- cross-sells
- upsells
- post-purchase offers
- loyalty thresholds
But the threshold should be based on business economics, not a ritual such as "set free shipping at 20% above AOV" without checking distribution and margin.
A better approach is to inspect how many orders sit near the intended threshold and whether the incremental basket value covers the incremental incentive cost.
The formula
**AOV = order revenue ÷ number of orders**Worked example
Suppose a store receives these 5 orders:
- Total order value is $300, so:
- AOV = $300 / 5 = $60
But 4 of the 5 customers spent between $28 and $34. No customer spent exactly $60. The $175 order pulled the average upward. That does not make AOV wrong. It tells you what kind of statistic it is. AOV is useful for revenue planning and comparing order economics, but it can hide the shape of the underlying order distribution.
Common mistakes
- Averages are sensitive to outliers and mix changes
- The numerator may handle discounts, returns, taxes, shipping, or order adjustments differently
- Higher AOV does not guarantee higher total revenue
- An expensive incentive can raise order value and destroy economics
- Other changes in traffic, product mix, seasonality, and pricing can move AOV
Comparison
A worked example: an average that almost nobody spent
| Option | Order | Value |
|---|---|---|
| A | A | $28 |
| B | B | $31 |
| C | C | $32 |
| D | D | $34 |
| E | E | $175 |
Questions we get asked
What is a good AOV?
There is no universal good AOV. It depends on category, price points, margin, customer mix, geography, and business model. Compare against your own economics and relevant cohorts.
Should I use mean or median order value?
AOV conventionally refers to an arithmetic mean. Median order value can be a useful companion because it is less sensitive to a few very large orders.
Do refunds reduce AOV?
That depends on the analytics definition. Some systems recalculate around refunds or net sales; others define AOV from the original order values or exclude certain post-order adjustments. Check the metric specification before interpreting a change.
Is AOV the same as average basket size?
No. AOV is monetary value per order. Basket size often means item/unit count per order.
OnVoard's take
AOV is most useful when you stop asking it to explain more than it can. Track it as one axis of order economics, then pair it with order count, distribution, gross profit, and segment-level behavior. The best AOV program is not the one that makes the average number biggest; it is the one that improves profitable customer behavior.
Every app on every plan. Connect your store and switch on the flows in an evening.