What is AOV (average order value) and how to increase it

AOV (average order value) is what a customer spends on average per purchase. It is calculated as AOV = total revenue ÷ number of orders, over the same period. Bill €48,000 across 800 orders in a month and your AOV is €60.

It matters because it is the numerator of almost everything else. With a CAC of €30, an AOV of €60 at 45% margin leaves €27 of gross margin per order and you lose money on every sale; the same CAC with an AOV of €85 leaves €38 and the business works. Neither the creative nor the bid has changed: only the basket. That is why, when an account cannot reach its minimum profitable ROAS, raising AOV is usually faster and cheaper than lowering acquisition cost.

The formula and the three mistakes that break it

AOV = total revenue ÷ number of orders

It looks impossible to get wrong, and it is still got wrong almost every time. The three usual failures:

• Revenue with or without tax, shipping and discounts. If you pull AOV from the ad platform dashboard and set the target against the figure in your accounts, you are comparing two different numbers. Pick one definition (the sensible one: revenue net of tax, discounts already applied, shipping excluded) and use it everywhere.
• Returns. In categories with high return rates (fashion, footwear) gross AOV and net AOV drift a long way apart, and the levers that raise the gross basket tend to raise returns too. The number that decides is the net one.
• Mean versus median. Four wholesale orders a month in a low-ticket shop drag the whole mean with them. If mean and median separate sharply, you have two businesses inside the same account and they have to be analysed separately.

And one distinction that trips a lot of people up: AOV is not LTV. AOV measures one purchase; LTV measures every purchase a customer makes. A low AOV with high repeat purchase can be a far better business than a high AOV on a one-off sale.

Why AOV is the fastest lever

An ecommerce business has only four ways to make more money: more traffic, better conversion, a bigger basket or more repeat purchase. The first two cost money or months; the basket moves in weeks, and almost always through merchandising changes rather than more spend.

The effect is double, which is why it shows up so quickly. AOV sits inside the ROAS formula, so raising it improves reported return directly, but it also lowers the bar: your break-even ROAS is 1 ÷ gross margin, and that margin rises with the basket because the fixed costs per order (shipping, packaging, gateway fee, picking) are spread across more euros.

A concrete case. A shop with an AOV of €60, 50% product margin and €9 of fixed cost per order: real margin per order is €21, or 35% of the basket, and it needs a ROAS of 2.9 to avoid losing money. Push AOV to €78 by getting a second product into the cart: margin goes to €30, or 38.5%, and break-even ROAS falls to 2.6. A 30% bigger basket has lowered the profitability bar by 10% on top of increasing the return on every sale.

The six levers, in order of impact

1. The free shipping threshold, correctly placed. The highest-impact lever and the easiest to ruin. The rule we apply: set it between 25% and 35% above current AOV. Placed just above the average basket it moves nothing; placed at double, nobody attempts it. And it has to be visible throughout the session ("€14 away from free shipping"), not only in the cart.
2. Bundles and packs with a saving people can understand. Selling together what gets consumed together raises the basket without discounting across the whole catalogue. The pack has to state the saving in euros, not as a percentage, and it has to appear on the product page, not only in a bundles category nobody visits.
3. Upsell on the product page, before the cart. The size or version ladder (the large format works out cheaper per unit) works on the product page and works far worse at checkout, where the person has already decided. In subscription, the equivalent lever is annual versus monthly.
4. Cross-sell with usage logic, not algorithmic. "Frequently bought with" lands much better than "you might also like" when the add-on is genuinely necessary (the case, the refill, the battery). A generic carousel of related products contributes very little.
5. Discount by units, never by amount. "3 for 2" or "second one half price" raise units; "10% off everything" raises average discount and cuts margin without moving the basket. If you have to discount, discount in exchange for volume.
6. Raise prices. The least glamorous and often the most profitable. In catalogues with stable demand, a 5-8% rise is absorbed with no measurable drop in conversion and goes straight to margin. Test it by category and measure it against conversion, not against how it feels.

How it connects to the campaigns

AOV is not only a shop problem: campaign structure decides which basket comes through the door.

• Segment the feed by price band. In Shopping and in the Meta catalogue, splitting the catalogue into bands lets you fund what raises the basket instead of letting the algorithm eat the cheap inventory, which is what it does by default because that converts more easily.
• Optimise for value, not for purchases. With tROAS or maximise conversion value, the system hunts for big baskets; with maximise conversions it hunts for easy ones. It is a one-checkbox change that moves paid AOV within two weeks.
• Watch AOV by channel. It is normal for retargeting to bring lower baskets and cold prospecting the higher ones, or the other way round depending on the business. If average AOV drops, it is almost always a change in mix between campaigns rather than a real change in buying behaviour.
• Be careful with acquisition discounts. An aggressive welcome coupon cuts AOV in precisely the orders that have to fund CAC. If it is used at all, it has to be conditional on a minimum order value.

And the check that closes the loop: any AOV lever is validated against margin per order, not against the basket. Raising AOV through volume discounts or a badly calculated free shipping threshold raises the number and lowers the money, which is exactly the opposite of the point.

Frequently asked questions