Data AnalyticsStrategy Marketing Jan 1, 2010

Real­ly, I Can Return It? Sold!

Opti­miz­ing your returns policy

Based on the research of

Eric T. Anderson

Karsten Hansen

Duncan I. Simester

Around Christ­mas, Celeste Pram­berg­er vis­it­ed Babies R Us in Tam­pa, Flori­da, to exchange an out­fit for her then eight-month-old baby, Allie. But the store did not have Allie’s size, so Pram­berg­er asked to return it. She was mor­ti­fied when the clerk said she could not because she had returned too many past purchases.

I was so sur­prised,” Pram­berg­er said. The only things I had returned had been four months pri­or from my baby show­er reg­istry. It wasn’t like I was reck­less, buy­ing things then return­ing them with­out a receipt.” Pram­berg­er said that after her baby show­er, she had returned about fif­teen items with­in four months priced between $5 and $20 because she’d gone over­board’ and reg­is­tered for too many sim­i­lar blan­kets, bot­tles, and one­sie” outfits.

I wasn’t ungrate­ful for the gifts, just an inde­ci­sive new mom,” said Pram­berg­er, who works in tal­ent and orga­ni­za­tion­al per­for­mance for Accen­ture. Although her returns ban at Babies R Us occurred five years ago, she said the inci­dent left her para­noid even today, resis­tant to return any­thing for fear of being embar­rassed, even at oth­er stores. And it may have left her gun-shy to pur­chase items in the first place, for fear of need­ing to return them later.

Peo­ple return items for a mul­ti­tude of rea­sons, from poor fit to buy­ers’ remorse. And some stores, like Babies R Us, track their cus­tomers’ behav­ior over time and then tai­lor their returns pol­i­cy based upon the customer’s per­son­al shop­ping his­to­ry. But are they harm­ing or help­ing their prof­it poten­tial by doing so? It is dif­fi­cult for retail­ers to parse the tan­gi­ble and intan­gi­ble ben­e­fits a cus­tomer per­ceives at sim­ply hav­ing the oppor­tu­ni­ty to return a purchase.

Most retail­ers base their returns poli­cies on a mix of cus­tomer data and intu­itive busi­ness sense, with the goal of pro­tect­ing prof­its but also fos­ter­ing loy­al cus­tomers who will come back to spend tomor­row. It is a fine line with murky para­me­ters, but com­pa­nies do not have to rely on tri­al and error to tweak their returns policies.

Mar­ket­ing Meets Finance
A new mod­el devel­oped by a team of mar­ket­ing sci­ence researchers allows retail­ers to plug in two or three years worth of cus­tomer data and get hard dol­lar fig­ures for exact­ly how much indi­vid­ual cus­tomers are will­ing to pay to have the option to return an item lat­er. The team was led by Eric Ander­son (Pro­fes­sor of Mar­ket­ing at the Kel­logg School of Man­age­ment) and includ­ed Karsten Hansen (Asso­ciate Pro­fes­sor of Mar­ket­ing at the Kel­logg School of Man­age­ment) and Dun­can Simester (Pro­fes­sor of Man­age­ment Sci­ence at Sloan School of Management).

I like to say that this study is where mar­ket­ing meets finance,” Ander­son says, because peo­ple in finance price out real options all the time and this is an exam­ple of the same kind of prob­lem — but in a mar­ket­ing context.”

Ander­son notes that his team’s study rep­re­sents the first time any­one has been able to pin a dol­lar sign on how much a cus­tomer val­ues the option to return some­thing. Their mod­el can help retail­ers to opti­mize their returns poli­cies for indi­vid­ual cus­tomers as well as across prod­uct cat­e­gories and pur­chas­ing modes, such as online ver­sus in stores. And because the mod­el mon­e­tizes the option to return pur­chas­es, Ander­son says it can also be applied to pars­ing gross and net demand changes based on the option to return, and cal­cu­lat­ing more accu­rate price elasticities.

Ander­son explains that while there is a lot of the­o­ret­i­cal lit­er­a­ture on cus­tomer returns behav­ior, empir­i­cal stud­ies are lack­ing, which drove him and his team to fill the void.

Every retail­er has a return pol­i­cy, but there is no empir­i­cal evi­dence as to how effec­tive they are or how much cus­tomers val­ue these poli­cies,” Ander­son says. There is just not much evi­dence about returns in gen­er­al, although this is an emerg­ing field in mar­ket­ing science.”

To pin­point the num­bers, his team approached solv­ing the prob­lem from a customer’s per­spec­tive and devised a sta­tis­ti­cal mod­el that chews up cus­tomer pur­chase his­to­ries and then spits out the prob­a­bil­i­ty of their indi­vid­ual pur­chase rates and return rates with­in defined prod­uct cat­e­gories over time (see Fig­ure 1).

Fig­ure 1: Sum­ma­ry of Customer’s Transactions

The mod­el also cal­cu­lates how much cus­tomers would pay for the option to return pur­chas­es with­in these cat­e­gories and how a retailer’s returns pol­i­cy affects whether or not they buy a prod­uct. For exam­ple, Anderson’s team used house­hold data from 987 indi­vid­u­als over ten and a half years and ana­lyzed their pur­chas­es for women’s footwear, men’s tops, and women’s tops from a sin­gle retail­er that sells through stores, a Web site, and mail cat­a­logs. They pur­pose­ful­ly chose these cat­e­gories for their inde­pen­dence of each oth­er in terms of returns and their vari­a­tion in return rates. They cal­cu­lat­ed that while men were only will­ing to pay $3.19 for the option to return tops, women were will­ing to pay $5. But the real sur­prise lay in footwear. Women were will­ing to pay an extra $15.81 per pair of shoes if they had the option to return them lat­er (see Fig­ure 2) when the aver­age price for shoes was around $50.

Fig­ure 2: The Impact and Val­ue of Returns

One way to think about that is if you were a direct mar­keter and peo­ple didn’t have the oppor­tu­ni­ty to return shoes that they hadn’t tried on, you’d have to drop the price down to $35 to sell those shoes,” Ander­son says. For retail­ers, that is a big deal.”

The mod­el works by cal­cu­lat­ing the base prob­a­bil­i­ty of pur­chas­es cus­tomers will make (based on their past his­to­ry), and then recal­cu­lates how this prob­a­bil­i­ty will change giv­en the option to return the pur­chase. For exam­ple, there is an 8.7 per­cent chance that an aver­age cus­tomer will buy a pair of women’s shoes in a giv­en month if she is allowed to return them (Fig­ure 1). But the prob­a­bil­i­ty that she will buy those shoes decreas­es by 0.030 if she can­not return the item (Fig­ure 2). This trans­lates to a 53 per­cent increase in demand for women’s shoes if the buy­er can return them.

For footwear, because the pur­chase rate is so low, the impact of a returns pol­i­cy upon net demand is huge,” Ander­son explains.

Pre­ci­sion Fit
The researchers found less steep increas­es in demand for women’s tops (16 per­cent) and even less increase in demand for men’s tops (9 per­cent), although in all cas­es the option to return mer­chan­dise increased demand. But there was wide vari­a­tion in how indi­vid­u­als val­ued the option to return their pur­chas­es, and Ander­son attrib­ut­es this spread to the pre­ci­sion of a product’s fit. There is less mar­gin of error for fit in footwear — or in baby’s clothes if we recall Pramberger’s prob­lem — and there­fore there is a high­er val­ue placed on the abil­i­ty to return the item when it is bought sight unseen.

There is a lot of vari­a­tion in customer’s val­u­a­tion,” Ander­son notes. A bunch of peo­ple place no val­ue at all on these poli­cies.” But some peo­ple, he added, pur­chase more because they have the option to return.

Anderson’s research pro­vides the type of infor­ma­tion a retail­er could use to cus­tomize its return pol­i­cy to indi­vid­ual cus­tomers, such as Pram­berg­er. Our mod­el would allow a retail­er to eval­u­ate how much demand would be lost from a cus­tomer if she was not allowed to return a product.”

The mod­el can also be used to dis­tin­guish dif­fer­ences in gross and net demand when returns are allowed. Crunch­ing the num­bers on women’s tops, the team found that while gross demand increased sig­nif­i­cant­ly when returns were allowed, the actu­al num­bers of cus­tomers who kept their pur­chase was only slight­ly greater, lead­ing to a small increase in net demand. And the shift in demand was con­sis­tent­ly greater for low­er-priced items ver­sus high­er priced items for both net (Fig­ure 3a) and gross demand (Fig­ure 3b).

Fig­ure 3a and 3b: Demand for Women’s Top

Tak­ing this into account, the researchers’ mod­el can make pre­dic­tions for max­i­miz­ing prof­its based upon a retailer’s cur­rent returns pol­i­cy. A sin­gle para­me­ter, K, is used to sum­ma­rize the ease of a prod­uct return pol­i­cy for each cat­e­go­ry, with val­ues of K over one indi­cat­ing an over­ly strict return pol­i­cy and val­ues below one an over­ly lenient pol­i­cy. The mod­el sug­gests that for their case study, the returns pol­i­cy for women’s tops is close to opti­mal but that returns for men’s tops should be made more strin­gent, or that the option should be elim­i­nat­ed. For women’s footwear, the returns pol­i­cy should be made more lib­er­al to cap­ture the max­i­mum prof­it (see Fig­ure 4).

Fig­ure 4: Opti­mal Return Poli­cies for Each Category

So, is this an aca­d­e­m­ic exer­cise or will it actu­al­ly help retail­ers? If Babies R Us applied our mod­el,” says Ander­son, they would be able to fore­cast how Ms. Pramberger’s pur­chase behav­ior changed after she was not allowed to return prod­ucts. Right now, these deci­sions are made with gut feel. Our mod­el allows retail­ers to mea­sure the impact of this pol­i­cy change.”

Ander­son believes the mod­el can ben­e­fit retail­ers in three oth­er ways: First, they can mea­sure how chang­ing their returns pol­i­cy will shift both gross and net demand and cal­cu­late expect­ed changes in net prof­it. Sec­ond, they can plug their own cus­tomer data into the mod­el to opti­mize their returns poli­cies across prod­uct cat­e­gories and across sell­ing modes. And third, they can cal­cu­late more accu­rate price elas­tic­i­ties by incor­po­rat­ing changes to their net demand when returns are allowed.

This mod­el is quite gen­er­al and can be applied to many types of prod­ucts, not just retail cloth­ing,” Ander­son says. The mod­el is applic­a­ble in any sit­u­a­tion where the cus­tomer has the option to renege on a purchase.”

Because retail­ers ulti­mate­ly want to fig­ure out not just how to mar­ket and sell their prod­uct, but also how to get their cus­tomers to keep it, Ander­son believes that future research avenues will inves­ti­gate how cus­tomers learn about the fit of the prod­uct, dif­fer­ences in return behav­ior in dif­fer­ent sell­ing modes, and eval­u­at­ing returns pol­i­cy in the con­text of retail competition.

Featured Faculty

Eric T. Anderson

Hartmarx Professor of Marketing, Professor of Marketing, Director of the Center for Global Marketing Practice

Karsten Hansen

Member of the Department of Marketing faculty between 2006 and 2008

About the Writer

DeLene Beeland is a science writer based in Graham, N.C.

About the Research

Anderson, Eric T., Karsten Hansen, and Duncan Simester. 2009. The Option Value of Returns: Theory and Empirical Evidence. Marketing Science, 28:3 (May), 405-423.

Read the original

Editor’s Picks


Pod­cast: How to Be a Great Mentor

Plus, some valu­able career advice that applies to just about everyone.


A New Way to Per­suade Kids to Drink More Water and Less Soda

Get­ting chil­dren to make healthy choic­es is tricky — and the wrong mes­sage can backfire.


How Can Social Sci­ence Become More Solutions-Oriented?

A con­ver­sa­tion between researchers at Kel­logg and Microsoft explores how behav­ioral sci­ence can best be applied.


Buy­ing a Com­pa­ny for Its Tal­ent? Beware of Hid­den Legal Risks.

Acquir­ing anoth­er firm’s trade secrets — even unin­ten­tion­al­ly — could prove costly.


Take 5: Tips for Widen­ing — and Improv­ing — Your Can­di­date Pool

Com­mon bias­es can cause com­pa­nies to over­look a wealth of top talent.


Every­one Wants Phar­ma­ceu­ti­cal Break­throughs. What Dri­ves Drug Com­pa­nies to Pur­sue Them?

A new study sug­gests that firms are at their most inno­v­a­tive after a finan­cial windfall.


4 Key Steps to Prepar­ing for a Busi­ness Presentation

Don’t let a lack of prep work sab­o­tage your great ideas.


Video: How Open Lines of Com­mu­ni­ca­tion Can Improve Health­care Outcomes

Train­ing physi­cians to be bet­ter com­mu­ni­ca­tors builds trust with patients and their loved ones.


Here’s a Bet­ter Way to Sched­ule Surgeries

A new tool could dri­ve sav­ings of 20 per­cent while still keep­ing sur­geons happy.

Politics & Elections

Why Eco­nom­ic Crises Trig­ger Polit­i­cal Turnover in Some Coun­tries but Not Others

The fall­out can hinge on how much a country’s peo­ple trust each other.


Build­ing Strong Brands: The Inside Scoop on Brand­ing in the Real World

Tim Calkins’s blog draws lessons from brand mis­steps and triumphs.


How the Cof­fee Indus­try Is Build­ing a Sus­tain­able Sup­ply Chain in an Unsta­ble Region

Three experts dis­cuss the chal­lenges and rewards of sourc­ing cof­fee from the Demo­c­ra­t­ic Repub­lic of Congo.