How to increase average order value without losing money
In short
A higher average order value helps only if the extra margin covers the reward and fulfilment costs. Before adding a free-shipping threshold, count the orders that already qualify: you will subsidise those orders too. Test the offer against contribution per eligible customer, including customers who do not buy.
What average order value actually measures
Average order value (AOV) is revenue divided by orders in a defined period. Use the same treatment of discounts, refunds, tax and delivery in both periods.
AOV describes the basket. It does not tell you whether the offer increased conversion, attracted more profitable orders or reduced repeat purchasing. Read it alongside order count, contribution margin and contribution per eligible customer.
How a shipping threshold can change AOV
A shipping threshold can encourage someone just below it to add another item: at a threshold of 220, a basket of 180 needs 40 more to avoid a 14 delivery fee. Whether that happens depends on the fee, the products available and the distance. Some customers will pay for shipping; others will leave.
Start with the distribution of basket values. Count orders already above the proposed threshold and orders close enough below it for an extra item to qualify. In the example below, 400 of 1,000 orders already reach 220, and their free shipping costs 5,600. An AOV of 180 hides both groups. There is no fixed timetable or guaranteed AOV increase.
What a threshold costs, on one month of orders
Illustrative scenario, not client results. All amounts use the same currency and exclude tax. Assume 1,000 monthly orders at an AOV of 180, a gross margin of 40% and shipping of 14 per order, initially paid by the customer. Of these orders, 600 are below 220 and 400 already reach it. Other variable costs and order count stay unchanged.
Where the 1,000 orders sit relative to a 220 threshold
Introduce free shipping at 220. For this calculation, assume that 15% of the 600 lower-value orders add 40: 90 orders qualify through extra spending. This is an input to the example, not an observed response rate.
| Cost or benefit | Orders | Effect | Amount |
|---|---|---|---|
| Baskets pushed up to the threshold | 90 | +40 revenue each, at 40% margin | +1,440 |
| Shipping now free on those orders | 90 | −14 each | −1,260 |
| Shipping now free on orders that already qualified | 400 | −14 each | −5,600 |
| Net contribution change | −5,420 |
Where the money went in the worked month
AOV becomes 183.60: (180,000 + 90 × 40) ÷ 1,000. Product gross profit grows by 1,440, but the shipping subsidy costs 6,860. The net contribution change is −5,420. That loss is 7.5% of the original 72,000 product gross profit; it is not a measured fall in the store's net profit.
The shipping subsidy applies to all 490 qualifying orders. Only 90 produce extra revenue under the assumptions. The remaining 400 cost 5,600 without a change in basket value.
Reducing the already-qualifying group to 100 does not make this offer profitable. If the other 900 orders respond at the same assumed rate, 135 increase their basket: 135 × 40 × 40% − (135 + 100) × 14 = −1,130. Change the inputs and calculate again; a smaller subsidy does not establish a positive result.
Compare the costs of loyalty mechanics
Compare each offer on the same expected contribution, rather than assuming one mechanic is cheapest:
- Points: estimate redemption, reward cost and the time until redemption. Include the cost of running the programme.
- Tiers: budget benefits for the period they remain available and define requalification rules.
- Bundles: subtract any bundle discount, extra goods, packing and shipping costs.
- Free shipping: include orders that would already qualify without a nudge.
- Percentage discounts: calculate the discount on the whole qualifying basket, not just the added item.
Any of these can help or hurt. The result depends on the offer, customer response and cost structure.
How to tell whether it actually worked
Before launch, define eligibility, the observation window, the main metric and the decision rule. Where feasible, randomly assign eligible customers to the offer or a control group. Analyse everyone assigned, including non-buyers; otherwise a change in conversion disappears from the comparison.
Compare contribution per eligible customer after rewards and fulfilment costs. Also report AOV, conversion and order count to explain the result. Allow time for returns and for purchases that may have been brought forward.
Without a controlled comparison, you can estimate the subsidy and monitor outcomes, but seasonality and other promotions remain alternative explanations. The basket distribution shows exposure to cost; it does not identify which purchases the offer caused.
The cost definition is in the contribution margin guide; the test design is in the loyalty campaign guide. Read the retention guides for repeat-purchase metrics.
Common questions
What is a good average order value?
Compare AOV within the same product category, market and revenue definition. For your own offer, the useful question is whether contribution per customer increased after its costs, alongside any change in conversion and order count.
Where should I set a free shipping threshold?
Use your basket distribution, shipping cost and product margins to model candidate thresholds. A threshold above the current average is a candidate to test, not a rule that guarantees profit.
Does raising AOV always increase profit?
No. A larger basket can be offset by discounts, reward costs, delivery subsidies or lower conversion. Compare contribution per eligible customer after those costs, including customers who do not buy.
Do points-based rewards raise AOV as effectively as free shipping?
Neither is universally cheaper or more effective. Points defer the reward; free shipping is immediate. Compare expected redemption or delivery costs with the additional contribution each offer produces in a comparable test.
See it on your data
On a twenty-minute call we connect one ad account and one month of your orders, and work out the gap between what the platform claims and what your shop took. Profit and POAS stay blank until you enter your margin and costs per order.
Book a demo