Tecof • September 14, 2026
What Is AOV (Average Order Value) and How Do You Raise It?
In Brief
AOV, average order value, is total revenue in a period divided by the number of orders in the same period. It tells you how much revenue an average order brings in, and it is the most direct lever for growing revenue without growing traffic: with visitor numbers flat, a 15 percent rise in basket size is a 15 percent rise in revenue. But the number is not a target on its own, because every method that inflates a basket has a cost: a free-shipping threshold eats margin, instalments carry bank commission, a gift item raises returns. As of 2026 the question is not "how do we raise AOV" but "in which customer segment, at what cost, and without breaking profit per order".
Tuesday evening, 21.15. In a home textiles store three numbers sit side by side on the dashboard: 1,842 orders in the last 30 days, 2,394,600 TL in revenue, a 1,300 TL average basket. Two weeks earlier the team raised the free-shipping threshold from 750 TL to 1,500 TL. The average basket went from 1,150 TL to 1,300 TL, a 13 percent gain. It reads well. Over the same period order count fell from 2,180 to 1,842, and nobody looked at that number. Revenue dropped from 2,507,000 TL to 2,394,600 TL in two weeks.
The mistake is not raising the threshold; it is raising it while watching a single metric. When AOV rises and order count falls, the gain has to exceed the lost orders, and any AOV decision taken without that comparison is a gamble.
What AOV Is and How It Is Calculated
The formula is simple, which is exactly why it gets misused. The complexity is not in the arithmetic but in which revenue and which orders you count.
Formula and choosing a period
AOV: total revenue ÷ total orders. With 900,000 TL of revenue across 750 orders in 30 days, AOV is 1,200 TL.
Period: weekly AOV is noisy; a single bulk purchase drags the average up. Monthly is the right resolution for most stores. Report campaign periods separately: putting November's AOV and February's AOV on the same chart produces the wrong conclusion.
Revenue definition: with or without VAT, with or without shipping, before or after returns? Write down the answer to all three and do not change it. The most common error is including shipping in January, excluding it in March, and reading the gap as improvement.
The three cases where AOV misleads
First, averages are pulled by outliers. One corporate order of 80,000 TL lifts AOV by 200 TL in a 400-order month, and nothing has improved. Second, channel mix: marketplace orders tend to be small and own-site orders large, so AOV falls on its own as the marketplace share grows. Third, the campaign effect: in discount periods unit counts rise and basket values fall.
Median basket versus AOV
Track the median alongside the average. The median is the value of the order sitting exactly in the middle when you sort orders by value, and it is immune to outliers. If AOV is 1,300 TL and the median is 780 TL, half your orders are below 780 TL and a handful of big orders are carrying the average. The wider that gap, the less the sentence "we raised the average basket" means.
| Measure | What it tells you | Outlier-sensitive | When to use it |
|---|---|---|---|
| AOV (mean) | Revenue per order | Yes, strongly | Revenue planning |
| Median basket | The typical customer's basket | No | Threshold and bundle design |
| Items per basket | Whether cross-sell is working | Slightly | Recommendation tests |
| Average unit price | Price positioning | Moderately | Upsell tests |
| Net AOV after returns | Revenue that actually stays | Yes | Profit decisions |
Why AOV Means Nothing on Its Own
Read alongside three other metrics, AOV drives decisions: order count, gross margin and return rate. Read alone, it only supports a story.
Reading it with order count
Revenue is AOV multiplied by order count. If AOV rose 13 percent while orders fell 15 percent, revenue fell. So every AOV change should be reported with those two columns side by side. In the store above, the net effect of the threshold change was minus 112,400 TL; at 1,100 TL instead of 1,500 TL, both numbers would probably have held.
Reading it with margin
Most basket-growing methods eat margin. Free shipping loads 70-110 TL of cost onto each order. A 10 percent basket discount on a 1,500 TL order is 150 TL. So the right question is not "how much did the basket grow" but "how much did gross profit per order grow". A basket moving from 1,200 TL to 1,400 TL at a 35 percent margin produces 70 TL of extra gross profit; if you absorbed 90 TL of shipping to get it, the trade lost money.
Reading it with return rate
Buy-two-get-one and quantity discounts grow baskets fast, and in apparel and footwear they grow returns just as fast. The item dropped into the basket to clear a threshold is the item most likely to come back. So measure AOV after returns: at a gross AOV of 1,400 TL with a 22 percent return rate, net AOV is 1,092 TL. Since the pre-campaign 1,200 TL basket at a 14 percent return rate netted 1,032 TL, the gain is 60 TL, not 200 TL.
The same link runs through advertising: a bigger basket returns more on the same spend. We covered how ROAS is calculated separately; a 10 percent rise in AOV lifts ROAS by 10 percent when ad cost is unchanged.
Five Ways to Raise AOV
There are not many methods. The difference comes from matching the right method to the right product group.
Free-shipping threshold
The fastest-acting method and the most frequently misconfigured one. Set the threshold 25-35 percent above the median basket. If the median is 780 TL, the 950-1,050 TL band is right and 1,500 TL is not; a threshold beyond the typical customer's reach does not grow the basket, it kills the order. The parts that make the threshold work are the "170 TL to free shipping" line on the basket page and three suggested items that would close that gap.
Cross-sell
Suggesting the item that completes the main one: a scarf with the coat, filters with the coffee machine. The rule is that the suggestion costs less than the main product; recommending a 3,500 TL item next to a 4,000 TL one does not work. Show it on the product page and at the moment of adding to basket, not on the basket page. Building those relationships by hand is feasible across 200 products and not across 5,000; beyond that, a disciplined SKU and category structure is the prerequisite for any recommendation engine.
Upsell
Offering a higher version of the same need: one litre instead of 500 ml, the extended-warranty model instead of the standard. To work, the price gap should stay under about 30 percent and the gap has to be explained by a concrete benefit. "90 TL cheaper per litre" outperforms "better value" several times over.
Bundles and sets
Selling items used together under one SKU at 8-12 percent below the sum of their individual prices. The advantage is not only the basket value but one shipment per order and a lower return rate. Derive bundle contents from sales data: the three products that most often appear in the same basket are your first bundle candidates.
Quantity discounts and subscriptions
For consumables (cosmetics, supplements, cleaning products, coffee) this is the most durable method, because the customer will buy again anyway. "Buy 3, save 15 percent" grows today's basket; a subscription grows both today's basket and lifetime value, and carries no return risk.
| Method | Typical AOV effect | Margin cost | Return risk | Best-fit category |
|---|---|---|---|---|
| Shipping threshold | 8-18% | High (70-110 TL/order) | Medium | Small baskets, wide catalog |
| Cross-sell | 5-12% | Low | Low | Any category with accessories |
| Upsell | 6-15% | Low | Low | Electronics, appliances |
| Bundle/set | 12-25% | Medium (8-12% discount) | Low | Cosmetics, home textiles |
| Quantity discount | 15-30% | Medium-high | Low | Consumables |
Constraints Specific to Turkey
All of the above works in Turkey, but three local constraints change the arithmetic.
Shipping cost and your carrier contract
The desi-based contract you hold with Yurtiçi, Aras or MNG is the foundation of any threshold decision. On bulky goods such as home textiles, the gap between 3 desi and 8 desi can be 40-60 TL per order. Before setting a threshold, work out the real desi cost of your twenty best sellers; calculating on an average desi loses money on the bulky ones. We went through the warehouse and packing side in the logistics and warehouse management guide.
Instalments, interest and commission
As baskets grow, customers ask for instalments, and bank commission rises with the number of instalments. At nine instalments commission can reach the 6-9 percent band. A campaign that lifts AOV from 1,200 TL to 2,400 TL while pushing most orders to nine instalments hands roughly 100 TL of that extra 1,200 TL to commission. Put the instalment table into the campaign maths.
Distance selling rules and returns
Under Turkey's distance selling regulation a consumer may withdraw within 14 days without giving a reason. That makes the item added purely to clear a threshold structurally likely to come back. For bundles, state explicitly in your distance selling agreement what happens when one component is returned; without that, every case becomes an argument. If you announce the campaign by email or SMS, check İYS consent in the system at send time; we compared the cost difference between email and SMS separately.
Measurement: How to Set the Test Up
AOV changes are almost always reported as "we tried it and it went well". The only way to know whether it went well is to tie the decision to a threshold written down in advance.
Write the decision first
Before starting, write this sentence: "After 14 days, if gross profit per order is at least 25 TL higher we keep it; if not, we roll it back." A criterion written afterwards is not a criterion, because it shapes itself around the result.
Collecting enough data
Because basket values are widely spread, AOV tests need more data than conversion tests. The practical floor is 300-400 orders per arm. In a store taking 25 orders a day that is roughly two weeks per arm. "Significant" differences that appear sooner are usually noise; see the separate piece on setting up an A/B test.
Looking at it by segment
A single average can hide two opposing effects. A shipping threshold may grow the basket among new customers while reducing order frequency among repeat customers. Look at three segments at minimum: first-time buyers, repeat buyers, marketplace buyers.
| Decision | Primary measure | Secondary measure | Rollback trigger |
|---|---|---|---|
| Threshold change | Gross profit per order | Order count | Orders fall 8%+ |
| Cross-sell module | Items per basket | Conversion rate | Conversion falls 5%+ |
| Bundle launch | Net AOV after returns | Bundle return rate | Returns exceed 25% |
| Quantity discount | Gross profit per order | Stock turnover | Profit declines |
A 30-Day AOV Programme
AOV work does not run on one switch; it runs as a sequence spread over four weeks. The timeline below is doable by one person on the platform you already have.
Days 1-7: fix the measurement
Write down the revenue definition (VAT, shipping, returns). Pull the last 90 days of AOV, median basket, items per basket and return rate. Break the same numbers out by three segments. Nothing changes this week; before changing anything you need to know where things stand.
Days 8-14: pull the cheapest lever
Cross-sell is the one method that does not eat margin, so start there. Define three complementary items by hand for your twenty best sellers and place them on the product page. In the same week add the "amount left to free shipping" indicator to the basket page, without yet changing the threshold.
Days 15-21: test the threshold and a bundle
Set a new shipping threshold 25-35 percent above the median basket and run it on half of traffic. At the same time launch the three most frequently co-purchased product combinations as bundles. Do not stack the two tests on the same segment.
Days 22-30: decide on profit and lock it in
For each test, put gross profit per order next to order count and decide against the threshold you wrote in advance. Document what becomes permanent: which threshold, on which date, for which reason. In three months nobody will remember, and an undocumented decision gets undone without justification in the next campaign.
Here is the job for tomorrow morning: sort the last 90 days of orders by value and find the one in the exact middle, the median. Put that number next to your current free-shipping threshold. If the threshold sits more than 35 percent above the median, it is not growing baskets, it is losing orders, and fixing it is an afternoon's work. On an e-commerce platform where basket, recommendation and shipping rules live in one panel, changes like these do not have to join a development queue.
Frequently Asked Questions
What is a good AOV?
There is no sector-independent "good" figure. 600 TL may be strong in cosmetics and weak in furniture. The only meaningful comparison is against your own history: same channel, same product mix, same season.
Which comes first, AOV or conversion rate?
If you have enough traffic but low revenue, AOV; if you have traffic but no sales, conversion rate. In practice AOV work tends to shave conversion slightly, which is exactly why both have to be tracked together. We covered ways to raise conversion rate in a separate piece.
Does raising the free-shipping threshold always work?
No. The threshold has to sit within one added item of the typical customer's basket. A threshold at twice the median makes people abandon rather than add. The right band is 25-35 percent above the median.
Where should cross-sell suggestions appear?
On the product page and at add-to-basket. Suggestions shown during checkout distract and lower conversion. On the basket page, show only low-value items that close the threshold gap.
At how many products do manual recommendations stop scaling?
The practical limit is around 300-500 products. Above that, hand-built relationships both take too long and go stale. That is the point to move to something that derives relationships from sales data.
Do bundles increase returns?
Usually they reduce them, because the customer buys things they will use together and knows what to expect. The one case where returns rise is a bundle padded with an unrelated item purely to lift the price.
Can I raise AOV on a marketplace?
Only to a limited extent. On Trendyol and Hepsiburada the shipping threshold and basket suggestions are not under your control; what remains is listing bundles and opening quantity-based variants. A high-AOV goal is mostly one you work on your own site.
Will raising AOV cost me customers?
If the method is coercive, yes. Pushy thresholds, insistent pop-ups and items auto-added to the basket grow this order and shrink repeat purchase. Measure over 90-day customer value rather than a single order.
How soon should I expect results?
Interface changes such as cross-sell modules and the shipping indicator show an effect within days. Product decisions such as bundles and subscriptions take 6-8 weeks to mature, because the repeat purchase cycle has to complete.
Which method should a small store start with?
Cross-sell. No margin cost, low technical requirement, easy to undo. Margin-eating methods such as shipping thresholds and quantity discounts come after your measurement is solid.