How to Increase Average Order Value in eCommerce
- AOV = total revenue divided by number of orders. Fix one revenue definition (before or after shipping, tax and discounts) and use it every time.
- Calculate AOV per traffic channel, not only overall. A blended figure hides which channels bring high-basket buyers and which bring discount seekers.
- Six tactics do most of the work: free shipping thresholds, scarcity, recommendations, bundles, visible cost savings, and gamified rewards.
- A higher AOV is not automatically better. Judge every AOV change by revenue per visitor and gross margin, because thresholds and upsells can suppress conversion rate.
- Segment before you optimize. First-time buyers, repeat customers and discount-driven cohorts have different AOV and respond to different offers.
Average order value is total revenue divided by the number of orders. You increase it by giving shoppers a reason to add one more item before they check out: a free shipping threshold they can clear, a bundle that costs less than the parts, a relevant accessory next to the product they are already looking at.
That much is simple. The part that is not simple is knowing which of those levers to pull, and confirming that the extra basket size did not cost you orders. This article covers the AOV formula with a worked example, how to calculate AOV per traffic channel, six tactics that reliably raise basket size, the objections that stop each of them working, and how to test the change instead of assuming it worked.
What AOV is and why it matters
AOV is useful because it tells you what a visit is worth once it converts. Applied to your marketing channels, it also shows how each source of traffic behaves: search campaigns, display ads and email lists rarely produce the same basket size, and the differences point straight at where personalization and offer changes will pay off.
It is also the input you can move fastest. Raising conversion rate means fixing the whole funnel. Raising purchase frequency takes months of retention work. Raising AOV can be a single change to a cart page, and the revenue effect lands immediately.
The caution is that AOV in isolation is easy to game. Removing your cheapest product raises AOV and lowers revenue. A free shipping threshold set too high raises AOV among the people who still buy and loses the people who do not. Read AOV next to revenue per visitor and gross margin, and it stays honest.
How to calculate AOV
A worked example
A store's organic search channel produces $9,270 in revenue from 37 transactions over a month.
Over the same month the store takes $482,000 in total revenue from 3,850 orders, so the overall AOV is $482,000 ÷ 3,850 = $125.19. Organic buyers are spending roughly twice the site average. That gap is the finding. It usually means organic traffic arrives further along in its decision, or that it lands on higher-priced categories, and it is a strong argument for testing bundles and premium upsells on organic landing pages before spending the same effort on paid social.
Getting the numbers out of analytics
The first requirement is analytics you trust. As Steven MacDonald put it in an interview for Omniconvert: "Whether that's goal tracking, e-commerce tracking, or removing duplicate scripts, without the correct data, it's impossible to measure any progress made."
In Google Analytics 4, channel-level AOV comes from the traffic acquisition report:
-
Confirm eCommerce tracking is firingCheck that the purchase event records a value and a currency on every transaction. A missing value parameter is the most common reason reported revenue is lower than the revenue in your store back office.
-
Open Reports, then Acquisition, then Traffic acquisitionSet the primary dimension to session default channel group, or to source / medium if you want individual campaigns rather than channel buckets.
-
Add the purchases and total revenue columnsGA4 reports average purchase revenue directly, but calculate it yourself the first time and reconcile against your store's own order report. If the two disagree, fix the tracking before you optimize anything.
-
Export twelve months and look at the shapeRecord AOV per channel per month. Investigate the highs and the lows rather than the average. Mark the promotional periods, holidays and stock-outs, because those explain most of the variation before any tactic does.
-
Split by device and by customer typeBasket size and conversion rate differ between phone and desktop, and between first-time and returning buyers. A blended AOV averages those groups into a number that describes none of them.
6 tactics to increase average order value
Tactic 1: Set a free shipping threshold
Shoppers will add an item to avoid paying for delivery. That is the whole mechanism, and it is the most dependable AOV lever there is because the shopper does the arithmetic for you.
If free shipping on everything is not affordable, set a threshold: "Free shipping on orders over $50." Put it above your current AOV, but close enough that one more typical product clears it. A threshold at roughly 15 to 30 percent above your current average is a sensible starting hypothesis. Show progress toward it in the cart ("Add $12 more for free shipping") rather than stating the rule once in the header, because the reminder is what triggers the extra item.
Two variations are worth testing alongside it. Free returns remove the risk objection rather than the cost one, and tiered shipping (standard free, expedited paid) lets the customer choose without you giving away the fast option.
Tactic 2: Use scarcity and urgency
Limited-time offers, countdown timers and low-stock notices compress the decision. A shopper who was going to come back next week buys now, and buyers in a hurry tend to consolidate rather than split purchases across visits.
The condition is that the scarcity must be true. A countdown that resets on refresh, or a "only 3 left" badge on an item you hold thousands of, is a trust cost you pay repeatedly for a one-off gain. Use real stock levels and real deadlines.
Tactic 3: Use a recommendation system
Upselling recommends a better version of what the shopper is looking at. Cross-selling recommends what goes with it. Both raise AOV, and both fail the same way, by recommending something unrelated.
If a customer is looking at a smartphone, suggest a case, a charger, or the higher-storage model. Do not suggest a kettle because it is on promotion. Relevance is what separates a helpful recommendation from an interruption, and relevance comes from what the shopper is doing now plus what similar customers bought, not from your merchandising calendar.
Tactic 4: Create product bundles
A bundle groups related products at a price below the sum of the parts. An online fashion store can bundle a dress, shoes and a necklace as a complete look; a coffee brand can bundle beans, a grinder and filters.
Bundles work because they answer a question the customer already has ("what else do I need?") while making the saving explicit. Build them from what customers already buy together, which your order data will tell you, rather than from what you want to clear.
Tactic 5: Show the cost saving
When the saving is visible, shoppers buy up to it. If one unit costs $10 and three cost $25, say so on the product page and in the cart, and state the saving as a number rather than leaving the customer to work it out.
This is the cheapest tactic on the list because it usually requires no new offer at all, only clearer presentation of an offer you already run. It is also the one most often left to a small line of grey text near the price.
Tactic 6: Use gamification
Game mechanics give the shopper a reason to spend a little more that has nothing to do with the product. A points system is the common form: points for every unit spent, redeemable against a future discount or a perk, with a threshold worth reaching ("earn 10 points per order, 100 points unlocks a reward").
Spin-to-win wheels, progress bars toward a tier and unlockable rewards all work on the same principle. Keep the reward economics in view: points redeemed as blanket discounts can hand back the margin the extra basket size earned.
Which lever to pull first
| Lever | What it changes | Signal that it fits | How to read the result |
|---|---|---|---|
| Free shipping threshold | Items per order | Most baskets sit just under your shipping cost; shipping is a top cart-abandonment complaint in your surveys | AOV up and order count flat means it worked. Order count down means the threshold is too high |
| Scarcity and urgency | Speed of decision | Long consideration cycles, high return-visit rates before purchase | Watch conversion rate and revenue per visitor together; urgency can pull sales forward without adding any |
| Recommendations | Number of distinct products per order | Product pages with no cross-sell block; accessories with low attach rates | Attach rate and revenue per visitor, not AOV alone, because a rejected recommendation adds friction |
| Bundles | Basket composition and margin mix | The same products repeatedly appear in the same orders | Gross margin per order, since the bundle discount is real money given away |
| Visible cost savings | Quantity per line item | Existing volume offers that are stated only in small print | Units per order; if it does not move, the message is not where the decision is made |
| Gamified rewards | Repeat spend and basket size together | An existing loyalty base and repeat-purchase behavior to build on | Margin after redemptions, over a full reward cycle rather than one month |
The main challenges in increasing AOV
Resistance to upselling and cross-selling
Customers refuse recommendations they did not ask for and cannot use. The fix is not to push harder but to recommend better: base suggestions on purchase history and on what the shopper is currently viewing, and make the bundle version cheaper than buying the items separately so the suggestion carries an obvious benefit. Placement matters as much as relevance. On the product page a recommendation is helpful; inside the checkout flow it is an obstacle.
Backlash to price increases
Raising the price of an unchanged product is the fastest route to a higher AOV and to customer complaints. Price increases are accepted when they arrive with added value: a better specification, a longer warranty, faster delivery, an improved service. Selling the same thing for more, with nothing added, reads as opportunism and shows up in your review scores.
Difficulty personalizing offers
Personalized recommendations, targeted emails and segment-specific promotions all depend on knowing who the customer is and what they have bought before. That is a data problem. RFM segmentation, built from recency, frequency and monetary value, is the practical starting point, because it groups customers by behavior you already record. Nexus by Omniconvert builds those segments from order history so offers can be aimed at the groups that will respond, and Omniconvert Explore delivers the on-site experience for each segment.
How to test your AOV changes
Formulate a hypothesis
Start with the scientific method. Identify something you believe to be true but cannot yet prove: that shipping cost is what stops second items being added, that customers do not know a volume discount exists, that the accessory sells poorly because it is never shown. Write it as a statement that a test can disprove, with a stated metric and an expected direction. Assumptions written this way are testable; opinions are not.
Collect the data behind it
Combine quantitative and qualitative sources. Analytics tells you where baskets stall and which channels produce which basket size. On-site surveys tell you why, which is the part analytics cannot answer. Explore includes both the survey tooling and the experimentation engine, so the answer to "why" and the test of the fix live in the same place.
Run the test and read it properly
Change one thing against a control, run to a planned sample size rather than stopping at the first encouraging day, and set revenue per visitor as the primary metric with AOV, conversion rate and gross margin as supporting reads. Explore has produced a 23.2% average uplift across more than 70,000 experiments on 7,000+ websites in 15+ industries, and the consistent pattern in that data is that changes read on a single metric look better than changes read on three.
Implement, then re-baseline
When a variation wins, roll it out and recalculate your AOV baseline before designing the next test. A free shipping threshold that was correct at an AOV of $125 is no longer correct at $145. AOV optimization is a loop, not a project, and the threshold is the part that goes stale fastest.
Frequently asked questions about average order value
Average order value is the average amount a customer spends in a single order. It is calculated by dividing total revenue by the total number of orders in the same period. AOV describes basket size, not customer value, so it is normally read next to conversion rate, purchase frequency and customer lifetime value rather than on its own.
There is no universal good AOV, because it depends on category, price point and market. The practical test is whether AOV and gross margin together cover your customer acquisition cost with profit left over. Compare your AOV against your own trend line by channel, device and segment rather than against a published industry average.
AOV = Total Revenue / Number of Orders. For example, a channel that produces $9,270 in revenue from 37 transactions has an AOV of $250.54. Use the same revenue definition on both sides of the calculation, and decide once whether shipping, tax, discounts and refunds are included.
A free shipping threshold gives shoppers a reason to add one more item so their order qualifies. Set the threshold above your current AOV but close enough that a single extra product clears it, usually 15 to 30 percent above the current average. Test the threshold, because a level set too high suppresses conversion rate instead of lifting basket size.
They can, if the recommendation is irrelevant or interrupts the checkout. Recommendations tied to what the shopper is already viewing, such as a case or charger next to a phone, are usually accepted. Broad recommendations placed in the middle of checkout add friction. Run these as A/B tests and judge them on revenue per visitor, not on AOV alone.
AOV rises during gifting peaks, when shoppers buy several items in one order, and falls in quieter periods and during heavy discounting. Compare each period against the same period a year earlier rather than against the previous month, and record promotions in your reporting so a discount-driven dip is not mistaken for a structural problem.
Yes. A blended AOV hides large differences between first-time buyers, repeat customers and discount-driven cohorts. Segmenting with RFM shows which groups already spend more and which respond to bundles or thresholds, so offers can be aimed at the segments where they pay off. Nexus by Omniconvert builds these segments from order history.
It can, when the reward is tied to basket size rather than to visits. Points per amount spent, tiered benefits and rewards unlocked at a spending level all give the shopper a reason to add one more item. Watch the margin, because rewards that are redeemed as blanket discounts can cancel the revenue they generate.
Pull twelve months of revenue and order counts, calculate AOV overall and then per traffic channel, and note where the two diverge. Mark the promotional periods so you can tell a discount dip from a real decline. Then pick the single tactic that fits the gap you found: a free shipping threshold if most baskets sit just below your shipping cost, a bundle if customers already buy the same two products together, a recommendation block if product pages show no cross-sell at all. Run it as an A/B test, not as a launch, and read the result on revenue per visitor and gross margin rather than on AOV alone. One tested change beats six untested ones, because only the tested change tells you whether the extra basket size cost you orders.
Test your AOV tactics before you launch them
Omniconvert Explore runs the A/B tests, on-site personalization and surveys behind every AOV experiment in this article, so you can see whether a threshold, a bundle or a recommendation block actually raises revenue per visitor. Explore has produced a 23.2% average uplift across more than 70,000 experiments.