RETAIL · PRICING
A European e-commerce retailer with more than €800 million in revenue and 4,2 million annual active customers sees the cost of its free-returns policy increase by 22 % per year. It reaches €68 million, with an average return rate of 38 % across fashion, footwear and home decor.
The finance team had projected a return fee of 5 euros per return, with targeted exceptions. The decision is sound from a cost perspective. It runs into another question: how can the economics of returns be changed without negatively changing the purchase decision itself?
THE PROBLEM
For the retailer, the return comes after the sale: it is a logistics-cost line, measured once the order has been placed. For the customer, the possibility of returning is part of the decision before the sale: it is assessed at the moment of hesitation.
A return policy therefore acts simultaneously on purchase, perceived risk, choice, the possible return, satisfaction and the next order. This shift in timing made the trade-off difficult to assess from observed costs alone.
The risk identified by the marketing team was twofold: lower conversion, with carts abandoned as customers approached checkout, and lower satisfaction, in a sector where free returns have become standard.
WHAT IS THE SAME FOR EVERYONE
WHAT DIFFERS FOR EACH GROUP
REVERSIBILITY
Buying online means deciding without having seen, touched or tried the product. A free-return policy absorbs part of that uncertainty. Charging for returns transfers part of it back to the customer at the very moment they hesitate.
This reading is not an equation. It is a behavioral representation designed to show why a return policy cannot be evaluated solely through its logistics cost.
THE CAUSE OF THE RETURN
PRODUCT DEFECT OR ORDER ERROR
The customer sees the problem as attributable to the retailer. This is the situation for which free returns were ultimately retained in the selected policy.
SIZE OR GAP WITH EXPECTATIONS
The product received does not match what was anticipated. Responsibility is perceived as shared between the product page and the customer's choice.
CHANGE OF MIND
Paying for a return prompted by one's own change of mind is the best-accepted situation, provided the rule is known in advance.
WHAT A RETURN RATE MEASURES
A volume. Not a cause, not perceived responsibility, not how a fee will be interpreted.
WHAT IS TESTED
The actual cost of a return is identical across all six configurations: what varies is the amount charged, when it is disclosed and the exceptions attached to it.
3 200 synthetic customers were placed in a simulated purchase journey, exposed to the six fee configurations and several timings for disclosing return fees. Dynamic agents followed up with each customer at abandonment or objection points identified in real time. Projections were calculated over 12 months for conversion, actual return rate and satisfaction.
SIMULATED POPULATION
The reconstructed population covers 1,8 million European e-commerce customers, calibrated on the retailer's proprietary data — purchase behavior, return history, price sensitivity — and on European e-commerce industry studies.
It is structured into 13 profiles combining purchase frequency, price sensitivity, use of returns and attitudes toward commercial practices perceived as unfair or deceptive.
FREQUENT BUYERS, RARE RETURNS
regular orders, marginal use of returns, low reliance on the return policy
FREQUENT BUYERS, FREQUENT RETURNS
heavy use of returns, mainly in fashion and footwear
HIGH NEED TO TRY BEFORE KEEPING
high pre-purchase uncertainty about size and material, ordering several variants
HIGH PRICE SENSITIVITY
decision driven by price, systematic comparison with competing retailers
OCCASIONAL BUYERS
infrequent orders, limited familiarity with the retailer's return conditions
HIGH-RETURN CATEGORIES
fashion, footwear, home decor: categories where the average return rate reaches 38 %
PREVIOUS RETURN EXPERIENCE
the quality of the most recent return experience shapes how the new policy is interpreted
SENSITIVITY TO PRACTICES PERCEIVED AS UNFAIR OR DECEPTIVE
high sensitivity to fees disclosed late, whatever their nature
These configurations reveal part of the population's heterogeneity. The simulation operates on synthetic individuals, not a handful of persona types.
REACTIONS
THE PRINCIPLE OF RETURN FEES IS NOT REJECTED
68 % of customers accept the principle of return fees provided they are informed before entering the purchase journey. Customers understand that returns carry a logistics cost. It is not the nature of the policy that triggers rejection.
WHEN THE INFORMATION IS GIVEN MAKES THE DIFFERENCE
When disclosed at the end of the funnel, at checkout, the fees generate 47 % cart abandonment. When disclosed on the product page, in a clear information box, they generate 12 %. The fee itself is strictly identical: only the timing of the information changes.
TRANSPARENCY IMPROVES SATISFACTION AFTER A RETURN
Post-return NPS is 34 under the conventional free-returns policy, where the service is considered normal but receives no particular credit. It reaches 52 when fees are disclosed in advance, accompanied by an explanation of the logistics cost and a product-quality commitment.
THE EXPECTED CAUSE OF THE RETURN CHANGES HOW THE FEE IS READ
Paying because you changed your mind and paying because the product received is defective are not the same transaction. This distinction led to free returns being retained for product defects or order errors.
Two customers can pay exactly the same amount and draw opposite conclusions. For one, “ that's fair, I changed my mind ”. For the other, “ I'm paying for a mistake that isn't mine ”. The objection concerns when the information is disclosed and the cause of the return, not the nature of the policy.
THE REACTION STARTS BEFORE THE RETURN
The question “ would you accept paying to return an item ? ” measures only one link in the chain. The simulation covered the entire chain, from perceived risk through to the next order.
DIRECT EFFECT, SECOND-ORDER EFFECTS
DIRECT EFFECT
The actual return rate falls from 38 % to 28 %, generating 18 M€ in annual logistics-cost savings. That is the intended effect, and it is achieved.
SECOND-ORDER EFFECTS
Order volume declines slightly, average basket value rises by 4 %, and overall revenue remains stable. Logistics savings cannot be read in isolation: they must be read together with what the policy shifts upstream.
WHAT IS MEASURED AFTERWARD
the number of returns and their unit cost
WHAT HAPPENS BEFOREHAND
the decision to buy, the number of variants ordered, comparison with another retailer
THE MEASURED TIPPING POINT
47 % cart abandonment at checkout versus 12 % on the product page, with identical fee content
WHAT THIS REQUIRES
assess the effect on purchase before counting the savings generated after purchase
It is precisely because the intended effect and the induced effects do not appear at the same point in the journey that the decision had to be simulated before implementation.
COMPARISON
Comparative reading based on simulated reactions across the six configurations tested. None is cost-free. Keeping returns free preserves conversion while allowing logistics costs to continue rising. Fees discovered at checkout sharply reduce returns and also reduce purchases. A differentiated policy preserves both, at the cost of longer rules to explain and a heavier operational transformation.
THE MOST ROBUST
Fees disclosed on the product page free returns retained for defects or errors explanation of the logistics cost
THE MOST FRAGILE
Uniform fees discovered at checkout no distinction by cause and no explanation
PROJECTION, THEN OBSERVATION
The retailer selected the combined configuration: a return fee of 4,90 euros instead of the 5 initially considered; fee information on every product page in a clear box accompanied by an explanation of the logistics cost; free returns retained for product defects or order errors; a service commitment including free expedited delivery above an order threshold; and enhanced product-quality monitoring to reduce legitimate reasons for returns.
Six months after rollout, the measured cart-abandonment rate is 12 %, in line with the projection. The actual return rate has fallen from 38 % to 28 %. Post-return NPS stands at 51, close to the projected 52. Overall revenue is stable, with a 4 % increase in average basket value offsetting the slight decline in order volume. No negative media or advocacy movement has emerged.
TAKEAWAY
With strictly identical fee content, the same amount produces 47 % cart abandonment when discovered at checkout and 12 % when known on the product page. It was not the price of the return that shifted conversion; it was when the customer learned about it.
Free returns were therefore not merely an after-sales service. They were insurance before purchase, absorbing uncertainty about size, material and the reality of the product. Removing them without replacing them with information transfers that uncertainty to the customer at the moment of decision.
DECISION
POSSIBLE FUTURES
A
CHARGE FOR RETURNS
Apply a uniform amount to all returns
B
CONDITIONNER
VARY FREE RETURNS BY CAUSE AND RETURN SITUATION
C
REORGANIZE
Differentiated policy, upfront information and action on the causes of returns
METHOD
This case adds a capability to the library: simulating a decision whose effect appears at a different point in the journey. The decision concerns returns; its consequences begin before purchase. What is tested is not an opinion but a chain: decision, anticipation, adaptation, consequence.
This case is based on a simulation conducted for a European e-commerce retailer. The company is not named, no personally identifiable data entered the system, and detailed results remain the client's property. The comparisons between configurations published here are qualitative; the quantitative values cited are those documented by the client.
Your next decision
Describe your need. We can point you to the right level of support.
State your decision. See the future it produces.
