Average order value gets thrown around in meetings as if everyone agrees what it means and why it matters. Often they don't. Some people are calculating it wrong, some are tracking it but doing nothing with it, and some are chasing it when a different metric would help the business more. Here's how to get the number right, then what to actually do about it.
How do you calculate average order value?
The formula is simple:
Average order value = total revenue ÷ number of orders
So if your store took £40,000 in revenue from 800 orders in a month, your AOV is £50.
A few things to watch:
- Use revenue, not gross merchandise value. If you're including cancelled or refunded orders in the total, you're overstating it.
- Use orders, not customers. One customer placing three orders counts as three.
- Pick a consistent time window. Weekly AOV can swing a lot on small stores, especially if you run occasional big-ticket promotions. Monthly or quarterly gives a steadier read.
- Decide whether you're including or excluding tax and shipping, and apply that rule consistently. Mixing the two across reports is a common source of confusion, especially when marketing and finance pull numbers from different places.
None of this is complicated maths. The mistakes people make are rarely about the formula itself.
What mistakes do people make with average order value?
The most common one is treating AOV as a goal rather than a diagnostic. AOV going up doesn't automatically mean the business is healthier. If you raise prices and lose a third of your customers, AOV climbs while total revenue and profit can fall. The number needs to sit next to conversion rate, order volume and margin, not on its own.
The second mistake is averaging across very different customer groups. A store selling both £15 accessories and £400 equipment will have an AOV that doesn't describe anyone's actual behaviour. It's a blended number sitting between two real customer types. Segmenting by product category, by new versus returning customer, or by channel usually tells you more than the headline figure.
The third mistake is chasing AOV when the real constraint is somewhere else. If your conversion rate is poor or your checkout is leaking customers, lifting AOV on the handful of people who do buy won't fix the underlying problem. It's worth knowing where you actually sit before deciding this is the lever to pull.
The fourth is measuring it and then doing nothing differently. AOV is only useful if it changes how you merchandise, price or structure offers. Otherwise it's a vanity number in a monthly report.
How do bundles increase average order value?
Bundling works because it changes the unit of purchase. Instead of a customer deciding whether to buy one more item, you're offering them a single decision that already includes more value. Done well, it also solves a genuine problem for the customer, grouping things that are usually bought together or that make more sense as a set.
A few practical notes:
- Bundles work best when the saving is visible and the combination is obviously sensible, not an arbitrary grouping to shift stock.
- Discounting the bundle slightly below the sum of its parts is usually enough. You don't need to discount heavily for it to feel worthwhile.
- Watch margin carefully. A bundle that lifts AOV but erodes margin more than the extra revenue justifies isn't a win, it just looks like one in the top-line numbers.
- Bundles tend to work better for complementary products than for near-identical ones. Pairing a product with an accessory usually outperforms pairing two versions of the same thing.
Do spending thresholds actually work?
Free shipping thresholds, gift-with-purchase thresholds and discount thresholds ('spend £75, get 10% off') are some of the most reliable AOV levers, mainly because they're visible at the point the customer is deciding whether to add one more thing.
The mechanism is straightforward: customers who are close to a threshold will often add another item to clear it, especially if the extra cost feels small relative to the benefit of free shipping or a discount.
What matters is where you set the threshold. Set it too close to your current AOV and you lose margin on people who'd have bought anyway. Set it too far away and almost nobody reaches it, so it does nothing. A sensible starting point is to look at your current AOV and set the threshold somewhat above it, then watch how many customers land just under it versus just over it. If a lot of people are stalling just below the line, that's useful information about where the friction sits.
It's also worth showing customers how close they are to the threshold, in the cart or on product pages, rather than only revealing it at checkout. People respond to a visible gap they can close, not a rule they find out about afterwards.
How does product mix affect average order value?
Product mix is the quieter lever, and it's often the one with the longest-lasting effect. If your best-selling product is low-priced, your AOV will always be dragged down by volume, no matter what you do at checkout. Shifting what gets seen first, recommended, or featured in email and ads changes the mix of what people actually buy.
Practical options worth testing:
- Reordering category pages so higher-value or bundled products appear earlier, without burying the entry-level ones customers still need.
- Using product recommendations that suggest a genuinely complementary item, not just 'customers also bought' noise.
- Reviewing whether your cheapest products are being oversold through ads or email, pulling in customers who were never going to spend much regardless of what you do at checkout.
This is also where AOV work overlaps with wider conversion rate optimisation. Thresholds and bundles are tactics, but they sit on top of a site that needs to present the right products, in the right order, to the right people. If you're not sure whether your AOV problem is really a merchandising problem, that's worth looking at properly before adding more checkout incentives. Our CRO service is built around exactly that kind of diagnosis.
What should you actually do next?
Start by calculating your AOV correctly and segmenting it by customer type and product category, so you know whether you have one number or several blended together. Then check it against conversion rate and margin before deciding AOV is the thing to fix. If it is, pick one lever, a bundle, a threshold, or a product mix change, and test it properly rather than changing three things at once and guessing which one worked.
A senior specialist will review your site and send back the highest-impact fixes, free.