AOV (Average Order Value) - how to calculate it correctly in practice

AOV - how to calculate average order value correctly

The AOV formula itself is simple. The real problems start when you try to calculate it using real data.

In theory, you take revenue, divide it by the number of orders, and you’re done. In practice, things get messy very quickly:

• should you use gross or net revenue
• should discounts be included
• what to do with returns
• should canceled orders be removed
• does shipping count as revenue
• how to calculate AOV across different channels or markets

This is where it’s easy to end up with a number that looks great on a slide, but breaks the moment you try to base real decisions on it.

Before going further, it’s worth reading the basics: AOV (Average Order Value) - what it is and how to calculate it

Here, we focus only on practice: how to calculate AOV correctly, consistently, and in a way that reflects reality, not reporting optics.

The simplest AOV formula

The starting point is always the same:

AOV = Revenue / Number of orders

Simple example

• revenue: €50,000
• number of orders: 1,000

AOV = 50,000 / 1,000 = €50

Mathematically correct.

But that doesn’t mean it’s useful. A correct formula on top of bad data is still a bad metric.

How to calculate AOV correctly in practice

The simplest version:

AOV = actual paid revenue / completed orders

Sounds obvious, but this is exactly where most reporting goes wrong.

The formula is not the problem. Mixing incompatible data is.

Step 1: decide between gross and net

Gross AOV

Gross AOV = Gross revenue / Number of orders

The most common version in marketing dashboards. It reflects what the customer paid, including tax.

Useful when analyzing campaigns and scale.

Gross AOV shows up everywhere in agency reports. Not because it’s the most useful metric, but because it’s the easiest way to make performance look better than it really is.

Net AOV

Net AOV = Net revenue / Number of orders

Closer to business reality. Shows what actually stays in the business.

Less impressive on paper, much more useful in decision-making.

The key rule

Pick one logic and stick to it. Switching between them is how “good-looking” reports are made.

Step 2: calculate based on actual paid amount

AOV should always be based on what the customer actually paid.

That means after subtracting:

• discounts
• promo codes
• cart-level promotions
• price reductions

Example

Input data:

• cart value before discount: €300
• discount: €30
• final paid amount: €270

At scale, this difference stops being cosmetic and starts distorting decisions.

Calculating AOV from pre-discount prices doesn’t improve performance. It just makes the report look better.

Step 3: decide how to handle returns

If you want AOV to reflect reality, returns must be included.

Rule

• full return → remove the order from both revenue and order count
• partial return → reduce revenue only

Example

Input data:

• revenue: €50,000
• returns: €5,000
• orders: 1,000

Difference: €5 per order

Ignoring returns is pure agency reporting bullshit. It boosts the numbers, not the business.

Step 4: remove garbage orders

AOV should not include:

• canceled orders
• test orders
• duplicates
• technical orders
• unpaid or failed transactions

If you include everything, AOV doesn’t get “more complete”. It gets useless.

Step 5: decide what to do with shipping

Either include shipping or exclude it.

Never switch between the two depending on what looks better in the report.

If shipping costs vary significantly between products or channels within the same segment, exclude shipping. Otherwise, your data stops being comparable.

Step 6: segment your data

One global AOV for the entire store is rarely enough.

If you only look at one number, you’re blind to what’s actually happening underneath.

Break it down at least by:

• channels
• devices
• new vs returning customers
• markets
• categories

Practical AOV calculation model

1. choose a time period
2. clean the order data
3. choose gross or net
4. include discounts
5. include returns
6. define shipping treatment
7. calculate AOV

Full calculation example

• total orders: 1,100
• canceled: 50
• test: 20
• revenue: €120,000
• discounts: €8,000
• returns: €12,000

Calculations:

• valid orders: 1,030
• adjusted revenue: €100,000

AOV = 100,000 / 1,030 = €97.09

Most common mistakes

• ignoring returns
• mixing net and gross
• using catalog prices instead of paid values
• not filtering orders
• inconsistent methodology
• no segmentation

What to analyze AOV with in e-commerce

AOV alone doesn’t say much. On its own, it’s easy to misinterpret or manipulate.

It only becomes meaningful when combined with other metrics.

Key combinations:

• CVR (Conversion Rate) → CVR (Conversion Rate) – definition, formula and benchmarks
• CAC (Customer Acquisition Cost) → CAC (Customer Acquisition Cost) – definition, formula and how to calculate it
• ROAS (Return on Ad Spend) → ROAS (Return on Ad Spend) – what it is, how to calculate it, and what it means
• LTV (Customer Lifetime Value) → LTV (Customer Lifetime Value) – definition, formulas and business impact

Together, they show:

• whether higher AOV comes from better selling or just higher-priced products
• whether campaigns are actually profitable
• whether customers generate value once or come back and buy again

Quick checklist

• you don’t mix gross and net
• you don’t use catalog prices
• you don’t ignore returns
• you don’t include canceled orders in the denominator
• you keep methodology consistent
• you don’t rely on one global AOV

Summary

The AOV formula is simple:

AOV = revenue / number of orders

Key takeaways:

• calculate AOV based on actual paid amounts, not catalog prices
• include returns, otherwise you inflate results
• remove canceled and technical orders from the denominator
• stick to one consistent methodology (gross or net, shipping included or not)
• analyze AOV together with other KPIs, not in isolation

If you work with reports, double-check the data:

• check if AOV is calculated before discounts
• check if returns are included
• check if canceled or test orders are counted
• make sure methodology is consistent across reports

AOV is much easier to improve in a spreadsheet than in real performance. If it looks great in a report, double-check what’s behind it.