A shopper who can’t touch the product buys insurance along with it. The return policy is that insurance, and it has a measurable value.
Online, every purchase is a small bet. The customer can’t try the shoes on or hold the sweater to the light. The return policy tells them what the bet costs if they lose, which makes it part of the price.
Eric Anderson, Karsten Hansen and Duncan Simester put a number on it. Using a catalog retailer’s purchase and return data, they modeled the right to return as an option the customer holds, like insurance. For women’s footwear, the option was worth more than $15 per purchase on average, and having it raised purchase rates by more than 50% Published. The value varied a lot by customer and by category, which is the useful part: the option is worth most where the customer is least sure, on a first order, in a category with sizes, colors or textures that photographs don’t settle.
A return policy is insurance the customer buys with every order. Price it like insurance, and write it so they can read it.
“Lenient” isn’t one setting. Narayan Janakiraman, Holly Syrdal and Ryan Freling reviewed the research on return policies and found that it describes leniency along five dimensions. Their meta-analysis of 21 papers then asked what each dimension does to purchases and to returns Published.
| Dimension | What it means | What the meta-analysis found |
|---|---|---|
| Money | How much of the price comes back: full refund, fee deducted, restocking charge | More lenient raises purchases |
| Effort | How hard it is: printer, box, drop-off distance, forms, receipts | More lenient raises purchases |
| Scope | What can come back: sale items, opened items, worn items | More lenient raises returns |
| Time | How long the window is | Longer windows reduced returns |
| Exchange | Whether you can swap rather than refund | Exchange options reduced returns |
PublishedJanakiraman, Syrdal and Freling, Journal of Retailing, 2016. Overall, leniency increased purchases more than it increased returns.
Two findings deserve attention. Time works the opposite way from what most operators assume: longer windows went with fewer returns, possibly because of the endowment effect (the longer you own a thing, the more it feels like yours). And the dimensions that sell (money and effort) differ from the one that drives returns (scope). Be generous where it buys orders and firm where it only buys returns. One caution: many of the studies are scenario experiments, in which people read a policy and say what they’d do. Treat the table as a map of where to test, not a result.
This is one chapter of The Return Trip, which is free and readable in full on a single page with no form in front of it.