Returning policies: valuing the conditions
By Tomas van Plateringen
S1654136
University of Groningen
Faculty of Economics
Title: Returning policies: valuing the conditions
Author: Tomas van Plateringen
Department: Marketing
Qualification: Master thesis
Completion date: October 23, 2012
Management Summary
An optimal reverse logistical could help companies to improve their overall performance. A fluently and satisfying returning process will help to improve customer satisfaction. Furthermore Anderson, Hansen & Simester (2009) showed that consumers are willing to pay more for a product if there is an option to return it. In this report the value of different conditions regarding return policies is researched using a choice based conjoint analysis. The choice-based conjoint model estimates the structure of customer’s preferences by decomposing overall evaluations for a specified set of products into utilities for the different attributes. The attributes in this research differ in: 1) price; 2) how many days after the purchase the customer is allowed to return the product; 3) if the customer has to pay a fee for returning a product/has to pay (a part of) the shipping costs; 4) if a customer should return it at a service point or if it is possible to be picked up at home; and 5) the “opening hours” of the service point or pick-up service. The attribute-levels of price, duration of right to return and returning fee have a linear relationship, while those of the delivery point and the opening hours have a part-worth relationship.
The model showed a three-cluster solution where segment A can be characterized as the segment with an aversion of returning fees, segment B as somewhat price insensitive and segment C as the most price sensitive. The utilities and the equalization prices show that most of the respondents are willing to pay more in terms of price than in terms of returning fees for all conditions, while some are indifferent towards returning fees. The customers who are indifferent towards returning fees are the most price sensitive respondents and it seems that price is the single most important factor on which they base their choice. The other respondents value the lower uncertainty, reduced risk and increased convenience by excluding the possibility of a returning fee more than the actual fee they have to pay for returning a product. Knowing this could help managers determining which returning conditions are right for their customers.
Table of Contents 1. Introduction ... 4 2. Problem Statement ... 7 3. Literature Overview ... 9 3.1 Online Retailers ... 9 3.2 Service Quality ... 11 3.3 Customer Loyalty ... 13 3.4 Reverse Logistics ... 14
3.5 Product Return Conditions ... 17
4. Conceptual Model ... 21
5. Research Design ... 23
6. Results ... 27
6.1 Descriptive Variables ... 27
6.2 Choice Based Conjoint ... 30
6.3 Predictive Validity ... 42
7. Conclusion ... 43
8. Implications ... 46
9. Limitations & Recommendations ... 47
References ... 48
1. Introduction
When someone buys a product, there is always a chance that it does not match the expectations of the buyer about, for example, the quality or fit of the product. In online retailing it is sometimes even harder to be sure about this, especially in apparel. Therefore the returns are a problem for online apparel retailers. Often the customer has the possibility to return the product under certain predetermined conditions. In general, customers return products to the seller when they are not satisfied with the (performance of a) product, and a (partial) refund is more valuable to them than keeping the product (Shulman et al., 2011). Customers make purchases, but are not always fully aware of all the product features, which can result in a mismatch. Davis et al. (1995) state that uncertainty and dissatisfaction about product performance is not limited to quality. In fact, a customer may buy a product that performs effectively and properly, but does not match her taste. Other reasons to return a product could be negative social feedback or a change of mind about the purchase. The customer, aware of the risk of mismatch, will be reluctant to go ahead with the purchase unless there is some protection mechanism (Yalabik et al., 2005). As it is often hard for customers to evaluate the product before the purchase, especially when they buy products online , consumers are generally given the option of returning the product. This way, retailers reduce the risk of the consumers. According to ReturnBuy (2000), return rates for online sales are substantially higher than traditional bricks-and-mortar retail sales, reaching 20–30% in certain categories of items. One of the reasons is, according to Rabinovich et al. (2011), that Internet retailing offers merchants limitless shelf space. This has led experts to highlight the existence of a “long tail” of offerings on the web and assert that the future of online business is “selling less of more.” However, customers can and do get overwhelmed by excessive product variety and therefore often make erroneous purchasing decisions. It is difficult for Internet retailers of physical goods to sell a large scope of products without having to handle potentially large amounts of product returns from customers.
McCullough Kilgore et al. (1999) state that in a survey of 40 e-commerce marketing executives 30% considered online returns to be one of their biggest challenges. Blanchard (2005) estimates $100 billion lost annually by U.S. manufacturers and retailers (both on- and offline), through product depreciation and management of the returns process. Of these $100 billion, $40 billion are spent on managing returns in reverse logistics processes (Enright 2003). In 2007 alone, the U.S. electronics industry spent $13.8 billion to repackage and resell returned products (Lawton 2008). Of those returned products, 95% were non-defective items that were not what the consumer was expecting. Returns occur at rates of at least 6% for electronics retailers (Strauss 2007) and as high as 35% for catalog retailers (Rogers and Tibben-Lembke 1998). According to Charlton (2007) not only the number of returns is high, they are also not handled well as almost 40% of the customers are unhappy with the processes for returning goods bought online. This is because the before mentioned supply-side cost of returns is only one aspect of the problem. According to Ofek et al. (2011) returning mismatched merchandise can be costly for consumers as well. First, there is the opportunity cost of time associated with the return process (going back to the store or mailing an item back). Second, there is the disutility associated with not having a matching product for the duration of time from the initial purchase until the return. Third, not all return policies are lenient, which could mean that customers have to put more effort in the returning process or have to contribute financially. In addition to that, Mermelstein (2006) states that 92% of customers are somewhat or very likely to shop again if the returns process is convenient. However, 82% are not likely or not very likely at all to shop again if the returns process is inconvenient. Therefore it can be stated that the returning process is an important part of the company performance which could be improved.
labeled as the same size, but can differ in actual size). As Akcay et al., 2010 states cannot enjoy the benefits of the traditional touch-and-feel shopping experience which diminish the customers’ ability to assess the quality of products (when buying from a catalog retailer or an online retailer customers (when buying from a catalog retailer or an online retailer customers). These troubles with before purchase evaluation cause return rates up to 35% for catalog retailers (Rogers and Tibben-Lembke 1998). Assuming these numbers are about the same for online apparel retailers, as clothing is typically a product category that benefits from touch-and-feel shopping experience which is lacking for both catalog and online retailers, there is room for improvement in the returning process of (online) clothing companies.
The market for apparel is very large and so are the revenues. According to the NY Times (2007), Americans made only 8 percent of all clothing purchases on the Web, compared with 41 percent of computers, 21 percent of books and 15 percent of baby supplies in 2006. In absolute numbers, however, the revenue from online sales of clothes reached $18.3 billion in the United States, surpassing online revenue from personal computers, printers and word-processing programs, which totaled $17.2 billion. This shows how big the market for apparel is, and means that there is still a huge potential to sell these kinds of products online as this is only 8 percent of the market (against 41 percent for computers). More recent, Mulpuru 2008 stated that the percentage of category sales that take place online is as high as 45% for computer hardware and software; 24% for books, music, and video; and 18% for consumer electronics, but is only 11% for jewelry, 10% for apparel and footwear, 9% for home furnishings, and 9% for cosmetics and fragrances. These numbers show an increase in percentage of sales made online for all categories and it can be assumed that these numbers are even higher today as online shopping has become more common and more and more retailers are offering their product online. Therefore this research will focus on the returning conditions of online apparel retailers.
The returning process is an important part of the company performance which could be improved and apparel is the product category with (one of) the highest rates of product returns. For online retailers, the so called e-tailers the return rates are even higher, as there is no possibility to touch and feel the products. Therefore, the focus of this research is on the returns in the online apparel retailing business. More specific, the value of different conditions regarding return policies is researched in this report. The conditions will differ in: 1) price; 2) how many days after the purchase the customer is allowed to return the product; 3) if the customer has to pay a fee for returning a product/has to pay (a part of) the shipping costs; 4) if a customer should return it at a service point or if it is possible to be picked up at home; and 5) the “opening hours” of the service point or pick-up service.
2. Problem statement
As stated in the previous section, the return rate of online purchases is substantially higher than the return rate of traditional bricks-and-mortar retail sales. Also, apparel is a product category with very high revenues, but has a high return rate (especially online). These high returns could have a negative effect on the revenues. However, retailers can protect themselves partially against this diminishing effect on their revenues by including (a part of) the return costs in the price. As Anderson, Hansen & Simester (2009) state, consumers are willing to pay more for a product if there is an option to return it. They also found that how much more consumers are willing to pay for a return option, differs per product category. They investigated how much more consumers are willing to pay for a return option. Their research showed that consumers are willing to pay 31% more than the average price for women’s footwear, 20% more than the average price for women’s tops, and 11% more than the average price for men’s tops if they are allowed to return the product. If the uncertainty about the fit of the product is higher, consumers are willing to pay more for a return option than for a product with a low uncertainty of fit. Petersen and Kumar (2008) showed that customers are more willing to pay more and are more likely to make other purchases if a company has a lenient return policy, because of the reduced risk. According to Posselt et al. (2008) the four major factors that affect the quality of the companies return or money back-guarantee (MBG) policies are (a) the restocking fee e-tailers charge for returns, (b) the nonrefundable shipping and handling fee, (c) the MBG’s duration (days to return the product), and (d) whether the e-tailer pays for shipping the returned product back. When a customer has any financial costs when returning a product, the risk to which he or she is exposed to increases. If the duration of the returning period is longer, it will reduce the risk because there is a longer period for evaluation. Although these articles show that a more lenient return policy leads to a higher willingness to pay, they did not explore which conditions offered the highest level of utility to the customer. Therefore, this paper investigates which of the common conditions for the return policy offers the highest level of utility. The conditions differ in the number of days to return, shipping costs, point of delivery, and opening hours of the return service. Also price is an attribute, this to translate the utility of the different conditions/attributes in financial value. As lenient return policies also increase the number of returns and thereby the costs of returns, it could be useful for retailers to determine if the options are worth the costs. However, this research will not discuss the costs of returns because costs mostly depend on the logistic process of a retailer and this varies per retailer. Based on the problem statement the following research question is formulated:
“What condition of the return policy for online retailers has the highest level of utility for the customer?” This research question results in following sub-questions:
“What is the utility of the number of days within which the customer is allowed to return the purchase?” “What is the (negative) utility of shipping costs to return the purchase?”
“What is the utility of the convenience of choosing the delivery point/pick up service for returning the purchase?”
3. Literature Overview
Return policies exist in all different kinds of levels and forms. As Mukhopadhyay and Setaputra (2004) state, a customer returns policy can be in the form of an unconditional money back guarantee, a store credit only, a partial refund, or no refund whatsoever. Restrictions imposed by a retailer for returning include short time limits for returning the product, unused product, returned in original packaging, and special instruction on labeling. Customers may return the product for many reasons, which may include the product not matching expectations; consequently, a customer returns policy is seen as reducing the risk of a customer having to keep an unwanted product (Chen and Bell, 2011). Furthermore, products are returned because customers change their minds after the purchase, mistakenly order unwanted products due to erroneous information or a lack of related knowledge or wrong products in terms of size, shape, color, quantity or content are sent by the retailers (Yu and Goh, 2012). As it is often hard for customers to evaluate the product before the purchase, especially when they buy products online, consumers are generally given the option of returning the product. With a return option retailers reduce the risk of the consumers. Some retailers however, have higher reverse logistics costs, sell products which have, on average, a higher return rate, are more receptive for moral hazard or have a different intention of what to achieve with a return policy and therefore have less favorable return conditions. So, return policies can differ in terms of conditions as different retailers have to take different matters into account.
This overview will discuss previous researched problems of online retailers (lack of trust, providing information) regarding returns, how return policies could influence the service quality of retailers and lead to more loyal customers, and how a reverse logistical process which handles returns efficiently could help to gain a competitive advantage. This review concludes by discussing the advantages and disadvantages of different return policies.
3.1 Online Retailers
An important reason why people choose to shop at a particular location, on- or offline, is because of their comfort level and confidence in after-sale support, including returns, repairs, and questions (Clark, 2000). According to Wolfinbarger and Gilly (2003), online and offline environments present different shopping experiences even when the same products can be purchased. Biswas and Biswas (2004) suggest that consumers have higher perceived risks while shopping online. Also, the diagnostic role of signals as risk reducers are generally more acute for products with non-digital attributes (e.g., determining the “fit” of a pair of jeans) than for products with digital attributes (e.g., CDs, software, books, etc.). This is because, for products with high digital attributes, there is some scope for evaluating the quality of the product before purchase. On the other hand, while evaluating the non-digital attributes of a product, the availability of cues will be further constrained in an online setting. As the exchanges between money and goods are not simultaneous because of the spatial and temporal separation between buyers and sellers in online markets, customers may not fully trust e-tailers’ online offerings and related purchasing process (Jiang & Rosenbloom, 2004 and Brynjolfsson and Smith, 2000). This could turn out to be a major problem for a retailer, because relationships are based on trust. Urban et al. (1998) state that trust is built over time, but initial trust must be established before any experience can be recorded. Therefore (online) retailers have to find a way to take away the lack of trust.
However, according to Harnos (2000), problems with delivery and return with e-tailers were cited as a “strong source of dissatisfaction” for 28 percent of the respondents. Kacen et al. (2002) state that compared to traditional stores, online stores have serious competitive disadvantages with respect to shipping and handling charges, exchange-refund policy for returns, providing an interesting social or family experience, helpfulness of salespeople, post-purchase service, and uncertainty about getting the right item. Furthermore, they state that products with more experience- and credence-related attributes may decrease consumers’ willingness to buy that product online. Evans (2009) found that almost half of the respondents with online access (45%) preferred to shop in physical stores because otherwise they “couldn’t touch, feel, or see the product”, and Ofek et al. (2011) state that 51% of online shoppers agreed or strongly agreed that returning products is a hassle that comes with online shopping. Korgaonkar and Karson (2007) argue that, overall, pure play e-tailers (retailers that have only an online channel) will continue to have a significant disadvantage in comparison to the clicks and mortar e-tailers, almost regardless of the type or level of inherent risk. They suggest that pure play e-tailers have yet to fully earn the trust of consumers. According to Bhattacherjee (2000), trust can minimize feelings of risk and lack of control that are often characteristic of e-tailing transactions. Therefore, pure-play e-tailers need to overcome these trust issues to reduce risk and deliver a compelling customer experience, as this is even
more important online than offline (Bezos, 1999). This is because customer experience contributes to strong word-of-mouth online and offers an opportunity to add
differential value as the Web increasingly offers consumers full information about product alternatives (Novak et al., 2000).
Assuming that retailers do not want to mislead their customers and aims for a long term relationship, providing information improves the quality of the decision to purchases. Therefore it will also help to reduce the likelihood of return and the perceived quality of the e-tailer.
3.2 Service Quality
Jiang & Rosenbloom (2004) argue that excellence pre-sales service is not necessarily an advantage that allows e-tailers to develop customer intention to return to an e-tailer. In fact, e-tailers might command higher customer retention through providing good performance in after-delivery service and continuously generating favorable price perceptions among customers because both have a strong and positive influence on return intention. The possibility of returning a product can be seen as a service that is offered by a retailer to a customer. The customer buys a good and if the customer is dissatisfied with the product for whatever reason, it can return the product to the seller. So, a returning option can be seen as a kind of service guarantee, and as Fruchter and Gerstner (1999) state service guarantees can benefit the firm, as they encourage every customer to try the product and not reduce the actual price paid to compensate for the risk of product nonperformance. Mitra and Fay (2010) state it is more likely for an e-tailer to seek to influence expectations about its service rather than a physical product that it sells. As most products are sold by multiple retailers, and especially in an online environment that makes it almost costless for the customer to navigate away to another e-tailer selling the same product, retailers do not engage in costly actions to influence expectations about a product. So to differentiate themselves from competitors, the retailer’s service offered could be used. Heim and Sinha (2001) say that attributes of electronic service products, which they define as bundles of physical goods, offline services, and digital content,deliver customer value when they become reasonable or preferred substitutes for conventional services. Customers’ perceived value is the customer’s the overall assessment of the utility of a product based on perceptions of what is received and what is given (Zeithaml, 1988). The extent to which these e-service product attributes substitute for attributes of conventional services will determine whether they provide value to a customer (Heim and Sinha, 2001). Chu et al. (1998) suggest that offering a (partial) refund to a customer, no questions asked, and granting the complete freedom to choose when to complain is a surprisingly desirable service policy. However, Heim and Field (2006) show that charging a restocking fee when the customer returns an item is one of the primary e-service process attributes associated with decreased customer ratings for ease of returns (and refunds). Also requiring the customer to ask the retailer what their return policies are decreases the ease of return, and thus the level of service offered.
It can happen that customers are unhappy with the product for whatever reason, but this does not necessarily mean that these customers are unsatisfied with the service delivered. It could be that someone is more inclined to shop at and return to a retailer which offers a better service, like a more lenient returning policy, than at a retailer which has a more strict returning policy. As Mollenkopf et al. (2007) show, return processes that require high levels of customer effort can have a severely negative impact on a customer’s satisfaction with the return transaction, which determines loyalty intentions, and thus future purchasing behavior. Also, Boyer and Hult (2006) state that, in general, customers with more experience with online retailers rate the experience more highly—in terms of service quality and product quality, time-savings and behavioral intentions — as the adoption of online ordering could have some initial hurdles for the customer. So, customers with a greater level of experience are likely to be more satisfied about the service delivered, and thus about the returning process. As satisfaction generates free word-of mouth advertising and has a positive influence on a company’s excellence in human capital, customer satisfaction boosts the efficiency of future advertising (Luo and Homburg, 2007).
Although, as discussed, product returns not necessarily mean unsatisfied customers, the returning process can function as a “recovery” point for unsatisfied customers. According to Horvath et al. (2005), service failure has been negatively associated with customer satisfaction, customer loyalty, service quality, trust, and behavioral intentions. It could lead to losing profitable customers which never return and a damaged image. To avoid such undesirable outcomes, service recovery could help. Sousa and Voss (2009), for example, show that failure resolution and satisfaction with the recovery both have a positive effect on customer loyalty intentions. Furthermore, Horvath et al. (2005) conclude that reverse logistics programs, which enable the returning process, represent an opportunity to undertake a sort of “product recovery,” that if handled correctly, gives the retailer a second chance to “get it right” with the customer. Well-handled service recovery significantly and strongly enhances the association between customer satisfaction and both trust and commitment toward the service provider (Tax et al., 1998). Although e-service delivery systems should include effective e-service recovery mechanisms, they should not look at superior recovery as a substitute for error-free e-service (Sousa and Voss, 2009). Sousa and Voss (2009) also state that a service system’s main focus should be with a strong failure-prevention mindset as a service failures occurs much more often than a service recovery paradox, where the loyalty intensions are more favorable than they would be had no failure occurred.
Hoff et al. (1998) states: what brings online customers back, primarily, is a sense of loyalty that comes from an internet company offering better service than the competition. Also Mukhopadhyay and Setaputra (2006) emphasize that the generosity of return policy and customer support, are important tools in building and maintaining customer’s loyalty. They state that a generous return policy, therefore, can be an excellent marketing tool to improve the product’s demand in the market. Furthermore, Mollenkopf et al. (2007) argues that satisfaction about return process is positively associated with loyalty intentions of a customer. So, handling returns well could help to increase customers’ loyalty.
3.3 Customer Loyalty
e-commerce typically experiences higher levels of product return compared to traditional retailing. Also Heim and Sinha (2001) state that product return is one of the three order fulfillment factors that influences customer loyalty (the other two are product availability and timeliness of delivery). Ramanathan (2011) shows that handling product returns plays an important role in shaping customer loyalty for low-risk products and also for high-risk products, but not for products that exhibits medium levels of risk. Ramanathan (2011) base the amount of risk on the price and ambiguity (the degree to which a product could be accurately described) of a product. The ease of return has a positive effect on the customer loyalty for low-risk products, but has a negative effect for high-risk products. As low-risk products are purchased without much detailed analysis since the products are less expensive. Customers are likely to return the purchase because of change of mind and are happy if it can be returned “hassle-free”. He states that, they also tend to buy more from the company if the returns are handled well. On the other hand, customers usually spend much more time and effort in knowing about the product before purchasing high risk products. Hence, the most likely reason for returning such a product would be some real issue and happens because customers are not satisfied with the product.
Mollekopf et al. (2007) state that any return presents a service recovery opportunity, even if the Internet retailer flawlessly delivers the requested product as promised, because the customer was not satisfied with the initial purchase experience. Therefore, firms should try to grasp the opportunity to recapture value for themselves and their customers, and thereby build customer loyalty as loyal customers will return to a company to make other purchases. However, returns represent an often-missed opportunity to manage customer relationships and build customer loyalty to the retailer. Not only are loyal customers more likely to make future purchases, repeat customers are also less likely to cheat than onetime customers.
However, although a stronger attitudinal loyalty results in a higher likelihood that a customer purchases more from a firm, a loyal customer is not necessarily a profitable customer (Reinartz and Kumar, 2002). Also Kumar et al. (2006) found that there is a low correlation between customer loyalty and (future) profitability. In addition to that, Kumar et al. (2009) state that some of the links in the conventional framework which show the path to profitability are weak. The conventional framework suggest that innovation leads to acquisition, acquisition combined with a rich experience leads to satisfaction, satisfaction leads to loyalty and customer retention, and customer loyalty and retention lead to profitability. Instead of using this framework they use the customer lifetime value (CLV) metric of Berger and Nasr (1998). CLV predicts the future profitability of customers. This metric calculates the net present value of a customer which shows how much a customer is worth (in monetary terms) and how much should be spend to acquire and/or retain this customer. Initially this metric did not take product returns into account. Petersen and Kumar (2008) propose to include product returns in the CLV metric. They show that product returns play a significant role in more accurately measuring and maximizing customer lifetime value (CLV) and is a necessary and independent part of any CLV objective function. Including the product returns to the CLV metric makes it possible for companies to determine if customers are profitable despite of the returns they make. Furthermore, Kumar et al. (2006) argue that it is imperative to distinguish between customers making moderate to excessive returns on basis of the CLV metric. When a company allocates customers with excessive returns, it can choose to cut back the marketing operations for these customers and try to discourage them to return products (i.e. by charging a fee for returning products).
3.4 Reverse Logistics
the costs of returns handling can be high (Rogers et al. 2002). According to Guide Jr. et al. (2006) most returns processes in place today were developed for an earlier environment, in which return rates were low and the value of the asset stream was insignificant. Returns processes were typically designed for cost efficiency where collection networks minimized logistics costs and the need for managerial oversight. Such cost-efficient logistics processes may be desirable for collection and disposal of products when return rates are low and profit margins are comfortable, however, this approach can actually limit a firm’s profitability in today’s business environment.Stock et al. (2006) state that within the reserve logistics domain, the product returns process has emerged as a key element that can influence the customers’ purchase decisions and thus, an effective product returns process is viewed as a competitive advantage. The returns process involves crediting the customer, evaluating the returned product condition, and directing the returned product to one of the five options: immediate resale, restock, and sale to third party retailer, donation, or disposal (Stuart et al., 2005). According to Rosen (2001) inbound packages cost two to three times more to handle than outbound ones because employees must inspect the return, repackage it, credit the customer, and put the item back into inventory. By nature, the product return process is more complicated than forward logistics operations due to the presence of individualized returns with small quantities, extended order cycles associated with product exchanges, and a variety of disposition options (Min et al., 2006). Poirier (2004) and Min et al. (2006) emphasize the importance of the logistical process as they found that firms in the optimal (or efficient) supply chain network enjoyed 40% more cost savings, 33% more inventory reductions, and 44% higher customer services than those in the inefficient supply chain network. Also Jack et al. (2010) show that the reverse logistics capabilities are significantly related (positively) to reverse logistics cost saving. These capabilities are negatively affected by customer opportunism as this increases the number of returns and consequently reverse logistics costs. Resource commitments and contractual arrangements affect the reverse logistic capabilities of a firm positively. So, these capabilities could be improved by resource commitment and by making contractual arrangements (which represents the back-end processes with other channel members that involve the development of common goals). This will help to reduce the cost of managing the product return process. Furthermore, Yalabik et al. (2005) sates that an efficient logistics process could either increase total demand by reducing the customer’s cost associated with making a purchase or it could increase the average profit margin by reducing direct costs.
For Internet retailers the reverse logistic process can be even more important as, according to Rabinovich et al. (2011), Internet retailers have a higher number of returns and are likely to have less concentrated returns. E-tailers are not as constrained by limitation in storage and location commonly found in traditional supply chains and, thus, can provide a wide range of choices for consumers. If returns prove to be less concentrated than sales, Rabinovich et al. (2011) suggest that a broad offering exposes the Internet retailer to costly risk because returns are distributed across a relatively wider product assortment. In contrast, a more concentrated returns pattern highlights a more manageable risk relative to the product breadth. As e-tailers compete in an environment in which consumers can easily evaluate and compare their product offerings, they must strike a balance in the distribution of product sales and product returns that optimally matches customers’ preferences for product variety.
direction of what was expected. The research of Cho et al. (2008) shows that outsourcing firms may not consider their logistics capability to be on par with their non-outsourcing competitors. This finding supports the fact that non-outsourcing firms actually do perform poorly compared with non-non-outsourcing firms considering gross and net profit margins. However, it should be noted that their research only compared the performance of non-outsourcing and outsourcing firms for the purpose of investigating the effects of logistics outsourcing. Thus, the accurate effects of logistics outsourcing may not have been investigated. As some retailers might just not have the capabilities to run the whole logistical process by themselves or at least not in an efficient manner, they are better of out-sourcing (a part of) the logistical process. Many e-tailers are outsourcing the shipping and returning process (at least from warehouse to the customer and vice versa) to delivery services which handle this part of the process more efficiently because of, among other things, scale advantages. The true effects of logistics outsourcing might be better analyzed by investigating the change in firm performance via a longitudinal study examining firm performance before and after logistics outsourcing.
Chen and Bell (2009) suggest that in general firms who are offering a full refund and facing customer returns that are proportional to the quantity of product sold should raise price (reducing the quantity sold) and reduce order quantity to mitigate the loss in profit resulting from the customer returns. On the other hand, according to both Hess and Mayhew (1997) and Anderson et al. (2009) the amount of returns and the return rate increases as the price increases. Therefore Chen and Bell (2009) state that when customer returns increase with price, the firm should reduce price and raise the order quantity which again leads to fewer returns. Higher prices leads to higher likelihood to return, but Anderson et al. (2009) found that the opposite is also true. Products with a lower price have a lower likelihood of return, because of an additional consumer surplus created by the lower price. Furthermore, Yu and Goh (2012) state that it is best to tailor the return policy according to the nature of the products and their condition upon return. Merchandiser should increase the order quantity for products that are recoverable when customers are allowed to return unwanted items within a predetermined period of time unconditionally. Likewise, the mass merchandiser should limit the order quantity for items that can be returned beyond a time window but with no penalty applied if the returned products are prone to deterioration with time. In general, return policies that allow products to be returned beyond a time window should have higher order quantities to cater for such eventualities.
processing such products. So, an effective reverse logistical process could enhance customer service, customer satisfaction, and customer’s perceived value on the one hand and decrease inventory levels, return processing time and costs.
3.5 Return Policy Conditions
Yalabik, et al. (2005) state that a refund policy decreases a customer’s risk associated with making a purchase, and thereby increases the total demand for the product. Furthermore, consumers may allow the retailer to charge higher prices because the reduction in the risk from the product’s being a poor match with their tastes may increase a consumer’s willingness to pay (Davis et al., 1995). Anderson et al. (2009) argue that not only are consumers are willing to pay more for a product if there is a return option, their willingness to pay for a return option also differs per product category. Therefore retailers should determine the willingness to pay per product category and use this to set their prices. Also, Wood (2001) states that for online purchases return policy leniency will decrease deliberation time for the initial decision to order, but will not lead to a subsequent increase in the deliberation time for the secondary decision to keep or return; thus, lenience appears to decrease the total overall amount of purchases decision conflict. Furthermore, it will reduce continued product search during the period between order and receipt. According to Yan (2009) in the e-market the optimal return policy and pricing strategy is to offer a more generous returns policy and to charge a higher price when the product web-fit (the extent of synergy between the characteristics of a product and the internet) is strong. Furthermore, Yan’s results also show that while the returns policy always is valuable for the e-marketer, the value of returns policy increases with the product web-fit.
Moral hazard
Although both a higher willingness to pay and higher purchase intension increase revenues, the possibility of abuse of the returning policy and higher costs should be taken into account. Some consumers may have valid reasons for returning merchandise. However, other consumers may take advantage of a liberal return policy as an opportunity to purchase without paying and to use the retailer as their own closet by using the return policy as a ‘vicious circle’ to purchase a product, use the product, then return the product to begin the process again (Wachter and Vitell, 2012; Longo, 1995). This is the so called moral hazard problem. Wachter and Vitell (2012) researched the relationship between product returns and unethical behavior. They found that the planned/unethical return orientation dimension is negatively correlated with the ethical dimension of ‘doing good’ and imply that such planned approaches to the return behavior process may have origins based on fundamental unethical beliefs. Wirtz and Kum’s (2004) suggest that high levels of satisfaction, morality, and self-monitoring reduce cheating, whereas high levels of Machiavellianism (which means according to the Oxford English Dictionary: "the employment of cunning and duplicity in statecraft or in general conduct") increase cheating. They also state that increasing restrictions may reduce unethical returns or consumer cheating. Furthermore, restrictions may increase consumer effort and therefore help reduce product returns by dissatisfied customers (Davis, et al., 1998). Posselt et al. (2008) argues for example that shortening the duration of the return policy is a way to discourage free renting by opportunistic customers. Shulman et al. (2011) state that if there are higher production costs involved it is likely that this will lead to a return policy that is more restrictive. One of the restrictions retailers can consider decreasing the retailers’ costs and the likeliness of moral hazard is diverting the costs of returns to the customer. To do this, a retailer can choose to give no refund at all or to give a partial refund.
A no refund at all will decrease the number of orders drastically and is not at all favorable for trust, perceived quality, customer loyalty, and willingness to pay. Therefore, it is very unlikely that a no refund policy is the optimal condition for a retailer. Chu et al (1998) state that a no refund policy is too restrictive as too many customers risk dissatisfaction, and as a result, orders will drop. Bonifield et al. (2010) show that a signal, like a lenient product return policy, could therefore help to increase trust and lower the perceived risk of a customer to purchases from an online retailer. By offering a restrictive return policy, customers will thus have a lower level of trust in a retailer. A return penalty, which in case of no refund will be the price paid for a product, reduces consumers’ willingness to pay for initial purchases because of uncertainty about whether or not the product matches with preferences (Shulman et al., 2010). Moorthy and Srinivasan (1995) state a money-back guarantee is a useful supplement to price as a signal of quality. As a low quality seller offering a money-back guarantee will have a higher probability of returns than a high quality seller, and hence also higher transaction costs. This will make it too costly for a low quality seller to mimic the high quality seller’s money-back guarantee. Therefore, a retailer with a no refund policy could be seen as a low quality retailer. However, this kind of policy could work out for a retailer which has only onetime customers and/or customers who are not interested in a return option. Some low-quality e-tailers may position themselves as a value provider, but by using the wrong signal these low-quality e-tailers may create unrealistic expectations to which they cannot live up to (Bonifield et al., 2010). It could also be the case that customers prefer low-service outlets and that a low service e-tailer will try to signal its true service level by, for example, not offering a refund (Mitra and Fay, 2010). Furthermore, Wirtz and Kum (2004) show that people have a lower propensity to take advantage of a guarantee when they are highly satisfied, than when the level of satisfaction is moderate. As satisfied customers tend to be more loyal, it could be assumed that customers with a low or moderate level of satisfaction are more likely to abuse the refund policy.
Partial refund
A partial refund could mean that only a percentage of the price will be refunded or that the shipping costs will not be reimbursed. Chu et al (1998) show that partial compensation policies are more likely to be offered under the following conditions: (a) The probability of dissatisfaction is low, (b) usage rate during product trail is high (that is, there is a serious threat of opportunistic behavior), (c) consumer’ cost of complaining is low, and (d) seller’s salvage value is low. Under these conditions a partial refund is the optimal policy for a retailer as it mitigates the problem of opportunistic abusive returns. Such a return penalty charged to consumers has several effects: it recoups costs associated with returns from consumers and it prevents a number of returns from occurring (Shulman et al., 2011). Moorthy and Srinivasan (1995) suggest that for cost efficiency it could help to give a partial refund instead of a full refund. Whether a partial refund is the optimal policy will depend on the costs of the returns. Also, it plays a role in altering consumer behavior (Shulman et al., 2011). By letting customers pay for a return, retailer can discourage customers to return products both the ones with a valid reason and the ones who misuse the return policy. However, it also reduces consumers’ willingness to pay for initial purchases because of uncertainty about whether or not the product matches with preferences (Shulman et al., 2010). Pasternack (2008) states that in a B2B market, the most profitable and therefore optimal refund policy for a manufacturer is offering a partial refund. Also Huang et al. (2011) suggest using a discount contract, where the amount of refund paid by the manufacturer exponentially decreases with the number of returns a retailer makes, as a form of a partial refund. They state that such a refund contract would also decrease the number of false returns as the retailer puts more effort in trying to reduce the number of returns.
A more lenient return policy offers a greater protection in case the customer has a change of mind regarding the purchases. According to Che (1996) consumers are always better off when the seller adopts the return policy with a full refund. Under such a return policy, the consumers are protected from any loss, so they receive strictly positive expected utility. So, for customers a money-back guarantee is the best option. For retailers though, it can be costly to handle such policy. Moorthy and Srinivasan (1995) state however that although the costs of returns can be high for retailers, high quality sellers should consider absorbing buyers’ transaction costs if there is little danger of them damaging the product or consuming the services of the product and returning it (moral hazard). They also argue that a return option combined with a money-back-guarantee can be used as a signal of quality. A low quality seller offering a money-back guarantee will have a higher probability of returns than a high quality seller, and hence also higher transaction costs. This will make it too costly for a low quality seller to mimic the high quality seller’s money-back guarantee. So, a high quality seller could offer a money-back guarantee to signal their quality. According to Akçay et al. (2010) it also prevents retailers from being associated with product dissatisfaction. Furthermore, a money-back guarantee offers the high quality seller the ability to signal with a price closer to his “complete information price” (the price he would have charged if consumers could observe quality directly and full information is available). Davis et al. (1998) state that a retailer is more likely to offer a low-hassle return policy when: 1) its products’ benefits cannot be consumed during a short period of time; 2) its products provide opportunities for cross-selling; and 3) it can obtain a high salvage value for returned merchandise. Akçay et al. (2010) state that the ability of the retailer to sell returned products as open-box items creates an opportunity to attract more price-conscious consumers and effectively pursue product differentiation. However, the downside of pursuing product differentiation is that open-box items can cannibalize part of new product demand. Furthermore, the result of Wirts and Kum (2004) research on cheating consumer behavior show that people who cheat do so at any amount of payout (small and large), and people who do not cheat at a low payout are also not enticed to do so when the payout increases. Even though the amount of refund in their research was 100 percent, the absolute material gain in dollars may not have been sufficient (i.e., up to $988 in our studies) to tempt people to switch from not cheating at a low payout to cheating at the higher payout. This finding has important managerial implications as it suggests that managers can offer a full-money-back guarantee without fearing that the generous offer would induce more customers to cheat. Furthermore, Suwelack et al. (2011) found that a money-back guarantee reduces performance risk, financial risk, and anticipated regret and increases the positive emotion of liking to significant degrees and Akçay et al. (2010) show that an alternative strategy of complementing the MBG with a reselling practice of returns can be very effective to generate additional profit.
increased for experience goods (with a money-back guarantee), retailers whose assortment consists of a large proportion of experience goods may have an even greater incentive to offer money-back guarantees compared to retailers selling mainly search goods.
So, the optimal refund policy for retailers regarding their product returns depends highly on the costs and efficiency of the reverse logistic system and the signal they would like to send (i.e., high vs. low-quality). For customers it depends on the costs of the returns, the ease of the returning process and the probability of fit. As different kind of customers have different kind of needs, also the optimal return policy can differ among customers. According Yu and Goh (2012), the merchandiser should not have a standard run-of-the-mill return policy for all product categories. It is best to tailor the return policy according to the nature of the products and their condition upon return. Also, Anderson et al. (2009) state that companies should have different kind of return policies for different product categories and for different customers. Likewise, Suwelack et al. (2011) state that money-back guarantees positively affect consumers’ responses for search and experience goods. Although for experience goods, money-back guarantees should be designed with stricter return conditions as compared to money-back guarantees for search goods.
4. Conceptual Model
Figure 1: Conceptual model
Customers often can evaluate the product only fully after it has been delivered. They need some time to decide whether they should keep or return the product. To reduce the risk of the customer, e-tailers are likely to offer a number of days for reflection. The duration of return policy is also an important dimension of its overall quality and may affect the rate of product returns (Hess and Mayhew, 1997 and Posselt et al., 2008). Suwelack et al. (2011) show that a short duration of a money-back guarantee decreases the money-back guarantee’s credibility. This negative effect of a short duration on money-back guarantee’s credibility is even stronger for experience goods than for search goods. Therefore it is likely that the utility is positively affected by the number of days the customer is allowed to return the product. However, after a certain time it is less likely that a customer has the need of returning a product as the terms and conditions (often) only allow undamaged and unused product to return. After a certain period customers are not likely to extract any extra value from a longer returning possibility because they already made up their mind whether to keep or return the product.
H1: The number of days a customer is allowed to return a product is, until a certain threshold, positively related with the level of utility for customers.
Shipping costs or a returning fee increases the risk of the customer, because he has higher costs in case of returning an unwanted product. Charlton (2007) argues that shoppers are not happy when they have to pay postage costs to return goods and Heim and Field (2006) show that charging a restocking fee when the customer returns an item is one of the primary e-service process attributes associated with decreased customer ratings for ease of returns (and refunds). Also, according to Harnos (2000) free shipping and returns are cited as one of the top reasons online consumers say they shop at a particular site and Suwelack et al. (2011)
H1: Duration of the right to return
H2: Returning fee
H3: Convenience location of the delivery point
H4: Convenience opening hours of the delivery point Utility - + + + H5: The effect of a) H3 and b) H4 are larger for high income households
states that by reducing the financial risk, MBG credibility increases consumers’ purchase intentions. Therefore it is likely that returning costs has a negative effect on the utility.
H2: Shipping costs or fees for returning a product have a negative effect on the level of utility for the customers.
If a customer can choose where, how and when to return a product, this will reduce the risk of wasting time on complex returning processes. Here, the convenience of the delivery point is dependent on the location where the customer can return the product and the opening hours of the returning service. Charlton (2007) states that shoppers are unhappy about the hassle of going to the Post Office to send return packages. Furthermore, Boyer and Hult (2006) state that one of the most commonly cited reasons for shopping for products on the Internet is to save time. By increasing the convenience for customers, they are able to save more time and thus retrieve a higher value from their interaction with the e-tailer. Reinartz and Kumar (2003) state that high-income households have high opportunity costs of time. They tend to substitute time by buying goods that will save time and are willing to pay for the added convenience. This makes that high-income households tend to spend more money for the same bundle of products than low-income households. Therefore, it is likely that the degree of convenience when returning a product is positively related to the utility, especially for high-income households.
H3: The convenience of the delivery point (location) for returning products has a positive effect on the level of utility for customers.
H4: The convenience of extended opening hours of the delivery point for returning products has a positive effect on the level of utility for customers.
H5: The positive effect of the convenience of the a) location and b) opening hours of the delivery point on the level of utility is larger for high-income households.
5. Research Design
Conjoint analysis is a multivariate technique developed especially to understand how respondents develop preferences for any type of object and is based on the simple premise that consumers evaluate the value of an object by combining the separate amounts of value provided by each attribute (Hair et al., 2010). The value is determined by the utility that is provided by an attribute and this utility is unique to each individual. To define the utility, all attributes that provide utility and form a basis for preference and choice must be identified (Hair et al., 2010). Also, these attributes should have different attribute levels. By combining these attributes and attribute-levels into profiles respondents’ preferences can be determined.
In a conventional conjoint analysis respondents have to rate or rank the dependent variable, but this research uses an online choice-based experiment to measure the level of utility respondents get from the different returning policies. Conjoint choice experimentation involves the design of product profiles on the basis of product attributes specified at certain levels and requires respondents repeatedly to choose one alternative (stimulus) from different sets of profiles offered to them (Lecture notes 8 RUG, AMR 2010-2011). As the respondents have to evaluate all profiles based on the combinations of all attributes simultaneously, the choice-based conjoint model is more realistic because in real markets customers also choose a product out of different alternatives. Choice choice-based models predict choice probabilities directly, while rating-based models convert preference ratings to choices through rules like maximum utility that has rather arbitrary assumptions (Moore, 2004). The choice-based conjoint model estimates the structure of customer’s preferences by decomposing overall evaluations for a specified set of products into utilities for the different attributes (or in this case, conditions). It is important to know how the customer’s choice behavior is influenced by different attribute levels to understand the customer’s preferences better. In a choice-based conjoint analysis the choice of the respondents is the dependent variable, instead of with a regular conjoint analysis where the dependent variable is a rating or ranking, the willingness to pay can be determined if there is a “none option” included. However, this research does not use a “none option” as it investigates hypothetical situation where the respondents already decided to buy the product, but have to choose between different (returning) conditions. Although the willingness to pay therefore cannot be measured, the equalization prices show how much more respondents are willing to pay for a better/higher attribute level.
For a company it is important to know how the customer’s choice behavior is influenced by the different attributes and the attribute-levels, because this way they can evaluate if certain attributes and their levels are preferred by their customers. As Pindyck and Rubinfeld (2005) state, the utility is a numerical score that represents the satisfaction that a consumer gets from a given product. The higher the utility, the more satisfaction a consumer gets out of the product. In this research five different returning conditions (attributes) and levels of returning conditions (attribute-levels) are investigated. In a survey (appendix 1) 206 respondents were asked to choose under which condition he/she would purchase the product. Every respondent has to evaluate nine choice sets which all contain three different stimuli. The five conditions differ in how many days after the purchase the customer is allowed to return the product, if the customer has to pay a fee for returning a product/has to pay (a part of) the shipping costs, if a customer should return it at a service point or if it is possible to be picked up at home, the opening hours of the return service, and in price. All these conditions have two or three possible levels (table 1).
Opening hours pick-up service or SP office hours (9 AM - 17:30 PM on Mo-Fr)
OH + evening hours till 9PM and on Saturday (9AM-17:30PM)
Price €29.95 €32.95 €34.95
Table 1: Attributes and attribute levels
The conditions and the level of those conditions are based on what is most common among different retailers in the Netherlands that sell apparel online. As shown in table 1, the condition for the number of days wherein a customer is allowed to return the product has three different levels: 14 days, 30 days, and 100 days. For shipping costs of or a fee for the returned goods there are also three different levels: no shipping costs/fee, a service charge of €1.00, or shipping costs of €4.95. The delivery point has two different possibilities: the customer can choose to deliver it to a service point (which could be a postal office or a store of the retailer) or to be picked up at any given address (i.e. home). The opening hours of the pick-up service and/or service points also have two different levels: only during office hours (9 AM - 17:30 PM on Monday till Friday) or extended opening hours in the evening (till 9PM) and on Saturday (9AM - 17:30 PM). The price attribute has three levels: €29.95, €32.95, and €34.95.
With the attributes known the function for determining utility can be determined. The formula is as follows:
where,
U = Utility
ј = Segment 1,…..,n
X₁ = Duration of right to return X₂ = Returning fee
X₃ = Delivery point X₄ = Opening hours X₅ = Price
designs are generated with the complete enumeration method conform Sawtooth’s principles of minimal overlap, level balance, and orthogonality. The minimal overlap makes sure that each attribute level is shown as few times possible in a single task, the level balance makes sure that each level of an attribute is shown approximately an equal number of times, and the orthogonality helps to choose attribute levels independently of other attribute levels, so that each attribute level’s effect (utility) may be measured independently of all other effects (Sawtooth CBC v6.0, 2008). The complete enumeration strategy considers all possible concepts (except those indicated as prohibited) and chooses each one so as to produce the most nearly orthogonal design for each respondent, in terms of main effects. The concepts within each task are also kept as different as possible (minimal overlap); if an attribute has at least as many levels as the number of concepts in a task, then it is unlikely that any of its levels will appear more than once in any task (Sawtooth CBC v6.0, 2008). The design efficiency of the first design (108 respondents) has a minimum efficiency of 96.47% for the attribute returning fee on level three (returning fee of €4.95). Table 2 gives an overview of the design efficiency per attribute for design 1.
Attribute Freq. Actual Ideal Efficiency
Duration of right to return: 14 days 14
(this level has been
deleted)
Duration of right to return: 30 days 14 0.3817 0.3780 98.08%
Duration of right to return: 100 days 14 0.3840 0.3780 96.87%
Shipping costs/fee for returning: €0.00 14
(this level has been
deleted)
Shipping costs/fee for returning: €1.00 14 0.3826 0.3780 97.62%
Shipping costs/fee for returning: €4.95 14 0.3848 0.3780 96.47%
Delivery point: Picked-up at any given address (i.e. Home) 21
(this level has been
deleted) Delivery point: Service Point (i.e. post office) 21 0.3308 0.3273 97.90%
Opening hours delivery point: Office Hours (Mo-Fr from 9:00 till 17:30) 21
(this level has been
deleted) Opening hours delivery point: Extended Hours (Mo-Fr from 9:00 till 21:00
and Sat 9:00-17:30) 21 0.3308 0.3273 97.90%
Price: €29.95 14
(this level has been
deleted)
Price: €32.95 14 0.3817 0.3780 98.05%
Price: €34.95 14 0.3789 0.3780 99.49%
The second design (98 respondents) has a minimum of 96.28% for the attribute price at level two (price of €32.95). Table 3 gives an overview of the design efficiency per attribute for design 2. Table 2 and 3 show that, although the designs have not as many profiles as they should have, the designs efficiency shows that the design is nearly orthogonal and balanced (see table 2 and 3). Therefore these designs can be used to gather information about the different attributes and attribute-levels.
Attribute Freq. Actual Ideal Efficiency
Duration of right to return: 14 days 14
(this level has been
deleted)
Duration of right to return: 30 days 14 0.3799 0.3780 98.97%
Duration of right to return: 100 days 14 0.3802 0.3780 98.83%
Shipping costs/fee for returning: €0.00 14
(this level has been
deleted)
Shipping costs/fee for returning: €1.00 14 0.3827 0.3780 97.52%
Shipping costs/fee for returning: €4.95 14 0.3797 0.3780 99.10%
Delivery point: Picked-up at any given address (i.e. Home) 21
(this level has been
deleted) Delivery point: Service Point (i.e. post office) 21 0.3299 0.3273 98.44%
Opening hours delivery point: Office Hours (Mo-Fr from 9:00 till 17:30) 21
(this level has been
deleted) Opening hours delivery point: Extended Hours (Mo-Fr from 9:00 till 21:00
and Sat 9:00-17:30) 21 0.3299 0.3273 98.44%
Price: €29.95 14
(this level has been
deleted)
Price: €32.95 14 0.3852 0.3780 96.28%
Price: €34.95 14 0.3797 0.3780 99.10%
Table 3: Design efficiency design 2
6. Results
6.1 Descriptive variables
Of the 206 respondents who participated in this study 173 (84%) is female and 33 (16%) is male. This may not seem a representative sample because it includes far more women than men. However, the customer base for most online apparel companies exists approximately of 75% of women. So, it is logical to have more female respondents in this case. The age of the respondents varies between 14 and 70 years with an average of 31 years. The households of the respondents contain 1 till 15 people where the most live in a household that contains one till four persons (94%). Of these households 77.2% are located in a city and 22.8% are located in a rural area. Of the respondents 41.3% has an income that is less than the average Dutch income in 2011 (€33,000), 17% has an income that is about average and 28.2% has an income that is more than average (table 4).
Frequency Percentage Less than average (<€33.000) 85 41.3% About average (€33.000) 35 17.0% More than average (>€33.000) 58 28.2% I prefer not to answer this question 28 13.6%
Total 206 100.0%
Table 4: Gross income respondents Table 5: How often respondents buy clothing per year
Table 5 shows the number of times respondents purchase clothing per year with an average between 11-15 times a year. Although the average is about once a month, 26.21% purchases clothing only 5 till 10 times a year and 26.21% buys clothing more than 20 times per year. The two respondents who answered that they purchase no clothing at all had a budget per month for clothing and spent money on clothing last year. An explanation could be that someone else bought their clothing or they gave this answer by accident.
Monthly budget Frequency Percentage 0-50 euro 82 39.80% 51-100 euro 60 29.10% 101-150 euro 30 14.60% >150 euro 34 16.50% Total 206 100.00%
Table 6: Monthly budget for clothing Table 7: Channel Preference for purchasing clothing
Table 6 gives an overview of the monthly budget of the respondents. Almost 70% has a budget not higher than 100 euro per month. Table 7 shows that most respondents still prefer to buy their clothes in brick and mortar stores (63.6%) instead of online (36.4%).
Purchase apparel ... times per year Frequency Percentage
0 2 0.97% 1-4 37 17.96% 5-10 54 26.21% 11-15 37 17.96% 16-20 22 10.68% >20 54 26.21% Total 206 100.00%
Table 8: Spent on clothing offline last year.
Table 9: Spent on clothing online last year
Table 8 and 9 show that more money is spent in offline than in online stores. The average spending offline on clothing per year is €906, where the average online spending per year is €385. Almost 79% spent between the 0 and 500 euro online, where offline only 41.8% of the respondents spent between 0 and 500 euro last year. Table 10 shows that most respondents enjoy shopping for clothing in brick and mortar stores (56%). Although most respondents prefer to shop offline for clothing, table 11 shows that vast majority of the respondents also enjoys online shopping (49%).
Enjoy shopping
offline Frequency Percent Strongly Disagree 8 3.88% Disagree 11 5.34% Somewhat Disagree 9 4.37% Neither Agree nor Disargee 18 8.74% Somewhat Agree 44 21.36% Agree 75 36.41% Strongly Agree 41 19.90% Total 206 100.00%
Money spent on clothing last year Frequency Percentage 0-250 euro 50 24.3% 251-500 euro 36 17.5% 501-750 euro 24 11.7% 751-1000 euro 37 18.0% 1001-1500 euro 29 14.1% >1500 euro 30 14.6% Total 206 100.0%
Money spent on clothing
Table 10: Enjoy offline shopping for clothing Table 11: Enjoy online shopping for clothing
In the survey the respondents were also asked to rate the importance of the attributes used for this research. The results show that all five attribute are rated as important by the respondents (table 12). A money-back guarantee is rated as most importance (87.8%), followed by free returns (78.1%), price (70.7%), and convenience (63.3%) during the returning
process. Duration of the right to return is rated as somewhat less important (50%).
Enjoy shopping
Price MBG Free Returns
Duration right to
return Convenience Freq. Percentage Freq. Percentage Freq. Percentage Freq. Percentage Freq. Percentage Not Important at All 0 0.0% 0 0.0% 1 0.5% 1 0.5% 2 1.0% Not Important 3 1.5% 2 1.0% 1 0.5% 8 3.9% 3 1.5% Somewhat
Unimportant 9 4.4% 1 0.5% 2 1.0% 6 2.9% 5 2.4% Neither Important nor
Unimportant 10 4.9% 4 1.9% 6 2.9% 17 8.3% 16 7.8% Somewhat Important 38 18.4% 18 8.7% 35 17.0% 71 34.5% 49 23.8% Important 104 50.5% 79 38.3% 87 42.2% 77 37.4% 85 41.3% Very Important 42 20.4% 102 49.5% 74 35.9% 26 12.6% 46 22.3% Total 206 100.0% 206 100.0% 206 100.0% 206 100.0% 206 100.0%
6.2 Choice based conjoint Specification of the model
The attributes used in this choice based conjoint analysis are: duration of right to return, returning fee, delivery point, opening hours, and price. For every attribute parameters are calculated and these parameters result in different utilities for different attributes (and attribute-levels). To calculate the parameters the preference function of every attribute has to be determined. The levels of a certain attribute could have a linear, quadratic or part-worth relation. A part-worth relation means that every change in a level of a certain attribute needs a specific parameter, where a linear model has one parameter which is multiplied by the level’s value to arrive at a part-worth value for each level. The attributes delivery point and opening hours are nominal variables with no rank order between the attribute-levels, so it can be concluded that their preference functions will be part-worth. Duration of right to return, returning fee, and price are scale variables. These variables could have a part-worth or linear relation.
Figure 1: Relationship part-worth parameters returning fee Figure 2: Relationship part-worth parameters price
