Your Free Shipping Threshold Can Raise Conversion and Cut Contribution

Your free shipping threshold may be working exactly as designed. That could be the problem.

Customers add another item. Average order value rises. Conversion improves. The dashboard rings the bell before Finance has counted the takings. Then somebody calculates what remains after delivery, picking, packing, payment charges, discounting and returns.

You made the basket heavier. You may have made the order lighter.

I've been in rooms where an increase in average order value was celebrated in the marketing review, then questioned in a margin conversation three days later. The commercial system was measuring the applause and the bill in separate rooms.

The cost hasn't disappeared

Free shipping is normally treated as a checkout mechanic. Set the threshold slightly above average order value, show customers how close they are and encourage them to add another product. Sometimes it works.

But free delivery doesn't remove the bill. It slides it under another department's door.

Marketing sees the improved conversion rate. Ecommerce sees the larger basket. Operations sees more products to pick and pack. Finance sees the delivery subsidy and the contribution left behind. Everyone has an accurate number. Nobody has the answer.

The dashboard can record a win that the business never receives

Suppose a customer is a few pounds short of your threshold. You tell them how close they are. They add a low-cost product and complete the order.

The dashboard records several successes:

• The session converted
• The order value increased
• Another unit was sold
• The threshold changed customer behaviour

Now follow the money. The additional product may carry little margin. It added picking and packing work. You absorbed the delivery cost. If the customer returns that product, you may also absorb payment, handling and reverse logistics costs.

The customer spent more. You may have kept less.

AOV is the speedometer. Contribution margin tells you whether the engine is still attached.

Customers do care about delivery charges

There's no sensible argument for ignoring delivery costs at checkout. Baymard Institute calculates an average documented cart abandonment rate of 70.22%. In its US shopper research, 40% of the non-browsing abandonment responses involved shipping, tax or other costs being too high. DHL surveyed 24,000 shoppers across 24 markets. It reported that 81% would abandon a purchase when their preferred delivery option was unavailable. Another 79% said they would leave when the returns process didn't meet their expectations.

Delivery charges can stop an order. Unexpected charges are worse. They're like reaching the restaurant till and finding the table, chair and washing-up weren't included in the menu price.

That doesn't automatically make free shipping the answer. It means delivery and returns belong inside the commercial proposition. They can't be left behind the checkout and treated as somebody else's problem.

More orders can still mean less profit

A study published in Marketing Science found that free shipping and threshold offers could generate additional sales. It also found that the loss of shipping revenue and uneven responses across customer groups could make the promotions unprofitable. The offer can do precisely what it was designed to do and still be the wrong decision.

More recent research shows what can happen when the threshold moves in the other direction. CUHK Business School's summary of the research describes a North American online grocery platform that raised its threshold from $80 to $100. Customer spending fell by 19.4%, mainly through lower basket values and, to a lesser extent, fewer purchases. Customers bought fewer distinct products and moved towards lower-priced items. They didn't obediently add another $20. Some looked at the higher bar and stopped jumping.

This was one platform in one category. The result isn't a forecast for your business. It does show why moving the threshold without understanding the customer is like changing the height of every step in a staircase and assuming everyone will keep climbing.

The evidence gives you the mechanism. Your own order data should determine the decision.

The average is hiding the customer

Your AOV might combine:

• New and returning customers
• Full-price and discounted orders
• Domestic and cross-border delivery
• Small parcels and oversized products
• High-margin accessories and low-margin core products
• Categories with very different return behaviour
• Customers who would have bought without the incentive

Then one number is spread across all of them like one coat handed to everybody in the building. It'll fit somebody. That doesn't make it the right size.

A customer already comfortably above the threshold isn't behaving like somebody who adds an unwanted product simply to qualify. Both receive free delivery. Only one order was changed by the offer.

Blend them together and the naturally high-value customer makes the threshold look more effective than it was. The same problem appears in product mix. An additional amount from a high-margin accessory may improve the order. The same amount from a bulky, discounted or frequently returned item can become a liability wearing a revenue badge. Your threshold changes what people buy, not only how much they spend.

One threshold can't do every job

Applying one threshold across every category is like setting one tyre pressure for every vehicle in the fleet: it ignores what the vehicle carries, how far it travels and what it costs to keep it moving.

A small accessory and a piece of furniture may contribute the same amount towards the threshold while creating completely different fulfilment costs. A domestic parcel and a cross-border shipment may qualify under the same offer while producing completely different margins. A loyal customer making a planned order and a first-time buyer adding filler products may generate the same basket value while behaving very differently afterwards.

The checkout sees matching totals. The P&L doesn't.

The numbers need to meet in the same room

Conversion and AOV belong in the analysis. Leaving them alone is the problem.

You also need to isolate customers immediately below the threshold from those already comfortably above it. Otherwise, you're crediting the offer for orders it didn't influence, like giving the umbrella credit because it didn't rain.

This is where ownership becomes uncomfortable. Marketing can't answer the question alone. Neither can Ecommerce, Operations or Finance. Each holds one part of the calculation, and each can build a credible version of success from its own section.

The threshold crosses pricing, merchandising, fulfilment and margin.

Free shipping can work

A clear, well-designed threshold can do useful work. It can remove uncertainty at checkout and help customers consolidate purchases. It may also allow you to spread fulfilment costs across a larger profitable order.

A 2023 study in Production and Operations Management found that the right policy depends on customer differences, demand patterns, margins and how the threshold changes buying behaviour. Under some conditions, customers topping up their baskets can support a high-margin model. Under others, a flat delivery charge is the better answer.

There's no magic percentage to add to AOV, and no competitor threshold worth copying without their margins, carrier rates, basket mix and customer behaviour. You can copy the number. You can't copy the economics underneath it.

Decide what success means before testing

Before lowering, raising or removing the threshold, decide which commercial result you're trying to change. Is it conversion, basket size, contribution margin, order frequency or customer retention?

Then establish:

• Which customers and categories should qualify
• What customers currently add to reach the threshold
• Whether those products are profitable and usually kept
• How delivery and fulfilment costs vary by order
• Which result would cause you to reverse the decision

Set those conditions before the test begins. If you wait until afterwards, an improvement in conversion will be used to defend almost any outcome. Once the speedometer has gone up, nobody wants to ask whether the car is travelling in the right direction.

Somebody always pays

Your customer may not pay a separate delivery charge, but somebody still pays. Margin might pay. Operations might pay. The next order might pay if the customer purchases less frequently. Finance might discover the cost later when the promotion is reconciled against contribution.

Free shipping can be a sound commercial decision. It can also become a remarkably efficient way to produce more orders you shouldn't have wanted. So don't stop at asking whether the offer increased conversion. Ask who paid for the delivery, what behaviour you received in return and whether the business was better off.

If the conversion chart rises while contribution falls, you haven't found growth. You've found a faster route to the wrong destination.

If those parts don't reconcile internally, that's the kind of issue I mean by ecommerce consulting: finding where apparently successful activity stops agreeing with commercial reality.