Coupon Codes
and their effect on
Shopping Cart Abandonment
Page |2
Coupon Codes
and their effect on
Shopping Cart Abandonment
by
Peter
Harsevoord
University of Groningen Faculty of Economics and Business
Department of Marketing Master Thesis Marketing Management
May, 2013 Peter Harsevoord Student number 1918427 van Hasseltstraat 14 8266 DK Kampen p.harsevoord@student.rug.nl University supervisors J.E.M. (Erjen) van Nierop
Page |3 CONTENT ABSTRACT ... 4 1. INTRODUCTION ... 4 1.1 Infomediaries ... 5 1.2 Affiliate Marketing... 5
1.3 Shopping cart abandonment ... 6
1.4 Uitmetkorting.nl ... 8
2. THEORY ... 9
2.1 Coupon codes ... 9
2.2 Shopping cart abandonment ...11
2.3 Retailer switching ...13
2.4 Coupon usage intentions ...14
2.5 Data collection ...16 3. METHODOLOGY ...17 3.1 Data collection ...17 3.2 Measurement scales ...18 3.3 Plan of Analysis ...20 4. RESULTS ...21 4.1 Preliminary Analysis ...21 4.2 Main Analysis ...24 4.3 Hypotheses Overview ...30 5 DISCUSSION ...31 5.2 Conclusion ...31 5.3 Managerial Implications ...33
5.4 Limitations and recommendations for further research ...33
LITERATURE ...37
Page |4 ABSTRACT
Retailers implement coupon codes as a pricing instrument in their marketing strategy to influence consumers in an early stage of the online purchase process. In contrast, consumers are confronted with the existence of a coupon code at the end of the purchase process when a coupon code input field is prompted. This research investigates 1) consumers reaction towards prompting a coupon code at the end of the purchase process, 2) how this subsequently is influenced by the absence or presence of a coupon code at a discount website, 3) how this influences consumers to change retailer choice and finally 4) how this is influenced by a consumer’s attitude or proneness towards coupon codes. Findings indicate that the presence of an input field significantly creates more consumers to abandon the shopping cart. Moreover, the final choice for a web-shop is significantly influenced by the presence or absence of a (competitive) coupon code at a discount website. Contribution of this research lies in the managerial implications in which advice is given how to optimally implement a coupon code strategy.
Keywords: Shopping cart abandonment, discount websites, coupon codes, Theory of Planned Behaviour, coupon proneness, retailer switching.
1. INTRODUCTION
Page |5 1.1 Infomediaries
A direct consequence of this value creation is that the information available online has grown to a gigantic and cluttered proportion, an in an information overkill. Due to the overkill a new online consumer demand arose requesting better ways of translating retailer value creation to clear and understandable formats of clustering information. This is difficult to fulfil by retailers due to the difficulty to collaborate with competitors. It resulted in a demand gap and opened doors for new business ideas. This demand gap created the arise of ‘infomediaries’ (short for information intermediates) who provide consumers a well-structured and clear overview of retailer possibilities and subsequently provide consumers the best retailer option (Papatla & Liu, 2009). Literature also often refers to infomediaries as cybermediaries (Sarkar, Butler, & Steinfield, 1998), digital intermediates or shopbots (Smit, 2002). As a business model an infomediary is referred to as an aggregator (Grewal, Gopalkrishnan, & Levy, 2004).
Intermediates are better suited to facilitate information aggregating services than retailers due to the independency of these intermediates and the absence of prejudice. As a broad definition an infomediary is an online format that facilitates a variety of services in combining retailers or products with customers to provide the customer a clear and organized overview of all information available (King, 1999). In other words infomediaries add value to the online consumption process by reducing search costs (Harrington & Leahey, 2007), increasing market transparency, intensify competition and optimizing customer satisfaction by providing the best buying option in consumer’s perception. To give an example of an infomediary Kieskeurig.nl is one of the most well-known infomediaries in The Netherlands. It adds value to the online consumption process by acting as an intermediate between retailers and consumers in several ways. Kieskeurig.nl collects retailer prices of different types of customer goods and lists these retailers based on price, store reviews, product availability, product reviews and delivery time. Consumers get a transparent market overview in combination with low search costs resulting in a perceived good and informed buying decision.
1.2 Affiliate Marketing
Page |6 not necessarily a sale) are most commonly used. Commission based rewarding in affiliate marketing is seen as a form of online advertising that is performance-based marketing. This means advertisers (retailers) only pay affiliates for measurable results (action) such as a sale or an application. This relatively new form of advertising is therefore attractive for retailers. It resulted in an increase of expenditures in affiliate marketing over the last years (and a further increase forecast in the years to come). In the Online Adspend Study 2013 (Advertising Bureau Nederland (IAB), 2013) the expenditures on affiliate marketing are calculated at 131 million euros in the Netherlands in 2012 with an increase of 3,3% compared to 2011 (figure 1). In other words retailers have paid affiliates and affiliate networks commissions of a total value of € 131 million. To indicate the size of the market, advertisement investments are estimated 10 percent of the total revenues earned by retailers via affiliate marketing. This follows to an estimation of a total market value of € 1.3 billion in 2011, affirming the importance of affiliate marketing. As a brief conclusion, consumers are significantly profiting from these shopping functionalities (Nielson-Online, 2003) resulting that online shopping through infomediaries (affiliate marketing) is subjected to a be a behavioural evolution in online-shopping.
Figure 1.
As can be seen in figure 1 affiliate marketing is a large and rapidly increasing segment of online marketing. It can be broken down in different types of publishers (not shown in figure 1) in which the largest revenue publishers are coupon code websites. This research will focus on these coupon code websites and subsequently their direct effect on shopping cart abandonment.
1.3 Shopping cart abandonment
Page |7 stores. The exposure to the opportunity to use discount vouchers is only when confronted to obtain a voucher. Customers must therefore confront the cashier with the voucher. Consequently cashiers will not prompt the customer to use a discount voucher. Hence, the process is customer initiated.
In the online setting vouchers are often referred to as coupon codes or voucher codes. Coupon codes are in comparison to offline discount vouchers retailer-initiated. In the checkout process customers are prompted to fill in a coupon code regardless the customer having a coupon code or not. Customers who do not own a coupon code are therefore confronted with a limitation in obtaining a lower price. This could lead to customers undergoing dissatisfaction, lesser price fairness, lesser future purchase intentions and lesser positive recommendation intentions towards the online shop (Oliver & Shor, 2003).
Coupon codes where introduced by retailers as an extra promotional channel to attract new customers and stimuli existing customers to do a second purchase or increase order size. As a consequence a new customer demand arose; the facilitation of gathering, bundling and distributing coupon codes to customers. This demand gap is filled by discount websites (discount infomediaries) with a focus on coupon codes. Retailers collaborate with these discount websites in order to influence customers’ retailer choice and measure up to the customer demand of coupon codes. In addition, Dutch online marketing strategists (Twinkle Magazine, 2012), affiliate networks and affiliate infomediaries (Marketingfacts, 2010) describe the additional value of discount websites as opportunities to stimulate upselling.
Page |8 transaction. In the Netherlands the tendency to collaborate with discount websites is therefore negative, resulting in retailers to terminate collaboration with discount infomediaries (Affiliate Blog, 2011a). The presence or absence of a coupon code has an effect on shopping cart abandonment. Presence of a discount possibility - a coupon code- has a positive influence on customer’s likelihood to finalize the purchase (Oliver & Shor, 2005). In contrary, the absence of a coupon code results in a higher amount of customers abandoning the shopping cart. The exposure to competitive web stores at a discount website, and a possible presence of a coupon code for that competitive store, could influence consumers’ retailer choice. In fact 20 percent of all coupon code searchers did not return to the web store (PayPal and comScore, 2009). Thus, shopping cart abandonment has a direct effect on the profitability of a retailer, suggesting the importance of coupon codes if retailers use a coupon code field.
However, shopping cart abandonment through coupon codes need further investigation. Customer responses to the presence of a coupon code prompting field are unknown and thus a better understanding for retailers, affiliate intermediates and discount infomediaries is needed. This research bridges the knowledge gap of customer responses to coupon codes and tries to appease the discussion in the Dutch affiliate marketing branch. A leading discount infomediary in the Netherlands,
Uitmetkorting.nl, initiated this research to influence the tendency in the market. Uitmetkorting.nl is, like similar websites, facing revenue losses due to retailers withdrawing promotional activities through coupon codes with discount infomediaries. This results in the following research question:
“Coupon code shopping cart abandonment amounts 27% of total online shopping cart abandonment behaviour and is therefore an important cause of customer loss. Whilst searching for a coupon code online, to what extent is the absence of a coupon code for a consumer’s current web store choice, or the presence of a coupon code for a competitive web store, an influencer for abandonment behaviour and/or store switching? Besides, are there differences in abandonment behaviour based on consumer’s coupon code usage intentions?”
1.4 Uitmetkorting.nl
Page |9 since by Vakantie.nl (€2.5 million) and Voetbal.nl with 453.780 euro (E-lead, 2011). To put this amount in further perspective, globally Fund.com ($9.99 million), Sex.com ($14 million) and Insure.com ($ 16 million) are the most expensive domain names ever paid for. After the take-over of Uitmetkorting.nl the website professionalized and grown to a broader aggregation of discounts (including coupon codes for web shops). Unique visitors lie between 3,5 and 4,0 million per year. In 2012 besides exploiting a Dutch, Flemish and German website Clansman.nl expended the concept of Uitmetkorting.nl to 7 other countries.
2. THEORY
This chapter will cover literature related to coupon codes, shopping cart abandonment, retailer switching and consumers’ coupon usage intentions. The first paragraph defines coupon codes and discusses coupon market trends and coupon market related literature. The subsequent paragraph connects coupon codes to shopping cart abandonment. The third paragraph focusses on the absence or presence of a coupon code in the buying process for both shop X as a competitive shop (Y), and relates both to shopping cart abandonment. The fourth paragraph discusses the effect of consumers’ coupon code usage intentions on shopping cart abandonment. In the last paragraph all proposed hypotheses are combined to a conceptual model.
2.1 Coupon codes
In probably the first research about online price reductions through coupon codes, then referred to as an electronic coupon or an e-coupon, the author (Fortin, 2000) discusses the potential of offering discount possibilities online as a promotional activity. Then e-coupons where not hyperlinks or codes but vouchers that customers had to print on a local printer and hand in at a brick-and-mortar store. The first step of offering discount possibilities online was set, but still there was a long way to go. Since offline retailers were struggling with low redemption rates on offline discount vouchers, online coupons could tackle this problem due to online coupons being customer-initiated. So when customers request a discount voucher, the probability that the voucher will be cashed increases. Moreover, the anticipated technological advances in the IT industry were expected to create a substantial shopping experience for consumers in the future. Within only a few years the future of Fortin (2000) became the present and the first discount websites as known today (e.g.
Page |10 coupon-driven traffic (i.e. via search engines) and forwarding traffic to retailers in order to collect commissions.
A coupon code is a one of many consumer sales promotion types. It is a promotional tool that provokes consumer reaction and is referred to as a sales incentive (stimuli). Traditional discount vouchers (but also coupon codes) serve to attract new consumers into a product category, influence customers to switch brands within a product category and stimulate loyalty (Blattberg & Neslin, 1990). In a broader sense adjusting prices through coupon codes is applied to influence consumers’ value perception and purchase intention (Grewal & Munger, 2001; Kimiloglu, 2004). All are decisions to create market share and strive for a positive net value by influencing consumer demand (Low & Mohr, 2000). Although monetary promotions (reduced prices) serve more utilitarian benefits (Chandon, Wansink, & Laurent, 2000), coupon promotions serve both utilitarian as hedonic benefits such as the enjoyment of achieving lower prices and the utilitarian benefit of saving money.
Besides the benefits of a coupon code, a consumer’s initial usage probability is also of reasonable importance. According to the Theory of Planned Behaviour (TPB) model of (Kang, Hahn, Fortin, Hyun, & Eom, 2006) the attractiveness of a coupon code is not the only factor influential. Also usage intentions and a person’s inherent desire or proneness to use a coupon code is relevant. It is therefore important to address the effects of proneness and subsequent usage intentions in the course of this research.
To summarize, a coupon code can be defined as a certificate (code or hyperlink) that gives consumers immediate price reductions, serves as an incentive to influence consumer demand, and is likely to be used more if a consumer is more coupon prone towards the act of getting price reductions through coupon codes.
However, the struggle between discount websites and retailers about the additional value of their collaboration remains divergent.
1. First, offering coupon codes through discount websites results in coupon codes being accessible for all consumer types. When applying earlier research price discrimination between price elastic and non-price elastic consumers is therefore ignored (Narasimhan, 1984), resulting in a cannibalization of margins for non-price elastic consumers. Targeting these consumers with a promotional tool would therefore result in superfluous marketing spendings since these consumers could be identified as already loyal or price unconscious.
2. Second, because consumers are often in the last phase of the buying process the intervention of a discount website rather erodes margins than boost sales according to retailers. In this phase of the buying process the consumer can get involved by prompting fields for entering coupon codes, and subsequent start searching for a coupon code online. Arjen Hettinga -online strategist at
Page |11 substantial amount of extra sales for retailers (Twinkle Magazine, 2012). These websites have created large amounts of followers via social networks and have built up substantial brand awareness. Thus this implies the importance of evaluating every discount website on additional value.
3. Third, since the dawn of web analytics the conversions (sales) are assigned to the last source the customer came from, also referred to as Last Cookie Counts (LCC) or cookie overwriting (Affiliate Blog, 2011b). A customer often visits more websites before proceeding to the purchase, which results in only the last cookie getting the reward. The common conception is that discount websites take over the cookie from the original source and therefore ‘steal’ the source its commission. However, research (Hewitson, 2010; Lamers, 2012) proved that the effect of cookie overwriting is only between 10 and 20 percent of all transactions, which confirms the misconception.
4. Fourth and last, the presence or absence of a coupon code influences shopping cart abandonment and/or switching between retailers. Close & Kukar-Kinney (2010) identified the key drivers that influence shopping cart abandonment, but is insufficient on the field of coupon codes since they did not incorporate this influencer in their research. So there is no research to support the assumption that shopping cart abandonment will occur.
All four misconceptions between retailers and discount websites concern the absence of a substantiated and collective vision. The first three misconceptions are managerial mistakes that require a progressive view by online marketers (retailers and affiliate programs).The fourth misconception -shopping cart abandonment- requires more research in order to contribute to current research.
2.2 Shopping cart abandonment
Page |12 Various studies have researched consumer motivations in dropping or abandoning the shopping cart.
Rajamma, Paswan and Hossain (2009) state perceived risk, perceived waiting time and perceived
transaction inconvenience as the three motivators for consumers to abandon the shopping process.
Greenleaf and Lehman (1995) give a more extensive version of this theory by stating a model of four
factors. Cho, Kang and Cheon (2006) worked further on this model, transformed the four factors to four categories with underlying factors, and applied it to the online setting.
First, medium and channel innovation factors cause consumers to develop a degree of resistance towards a purchase due to continuous innovative changes. So consumers that have created an habitual buying behaviour will judge changes in that behaviour as negative. Second are three contextual factors that consist of time pressure (other things to do), perceived need (do not actually need the product) or negative experiences towards online shopping (high shipping rates, long delivery times, not fulfilled expectations etc.). Third are characteristics of consumers that include resistance towards online shopping, higher consumption styles (degree of consumer search) or a limitation to cope with the information available. Fourth and last, there are uncertainty factors that causes consumers to perceive negative consequentions towards a purchase. These factors include financial risk (overspending in buying), social risk (feeling of failure in consumers’ social environment), procedural uncertainty (lack of knowledge) and psychological risk (future regrets).
Page |13 financial risk category will be proposed in this research: the ability to achieve lower prices (a broadening of financial overspending). Thus, the awareness of this ability is aroused by alerting the consumer to achieve discount. In summary of this paragraph, the presence of a coupon code field in the purchase process results in a higher degree of shopping cart abandonment.
H1: Prompting the consumer an ability to achieve a lower price through a coupon code causes higher degree of consumers abandoning the purchase process.
The following up step in a consumers’ buying process –if the consumer postpones the process- is the search for a discount possibility. This ‘possibility’ is often found online at discount websites mainly focussing on the distribution of discounts. The consumer turns directly to these websites or enters a retailer and discount related keyword combination into a search engine to find a proper discount website. When the coupon code is found the consumer returns to the shopping cart to finish the order. The buying process is therefore postponed, not abandoned. If no code is available online, the chance of the consumer resumes the purchase process is expected to be lower, and the chance that the consumer terminates their purchase intention is higher.
H2: The absence of a coupon code for shop X at a discount website causes a higher degree of consumers abandoning the purchase process.
2.3 Retailer switching
Another consequence of prompting a coupon code is the possibility that a consumer gets influenced by competitive web-shops whilst searching for a coupon code for the web store that has their current preference. Since most resources that consumers consult for a coupon code facilitate discounts for more retailers, there is a chance of retailer switching. Subsequent consumers become more and more aware of the possibility to reduce their order price, and start searching for other web shops that offer the same product but with –in comparison to their current web store choice- a discount possibility.
Keaveny and Parthasarathy (2001) and Donthu and Garcia (1999) discussed the effect of consumers
switching online from retailer A to retailer B, and found that the online consumer is more likely to switch between brands and stores in comparison to the offline consumer because online consumers are less loyal. This implies that consumers could be more sensitive for competitive discount offers whilst searching for a coupon code at a discount website. Therefore there is the expectation that consumers switch from shop X (when no coupon code is present) to shop Y when there is a coupon code present for shop Y.
Page |14 2.4 Coupon usage intentions
From a consumer’s perspective a coupon code is a way to reduce prices. Depending on the height of the discount the chance of redeeming a coupon code increases positively. Therefore discount height is the most influential coupon usage factor (Chiou-Wei and Inman, 2008). In a more cognitive matter, the decision for consumers to search for a coupon code is seen as a trade-off between the degree of perceived search and time costs and the perceived benefits that are related to the coupon code (King, 1999).
Another more extensive approach to explain and predict consumers’ (e-)coupon redemption behaviour was initiated with the theory of reasoned action (TRA) (Fishbein & Ajzen, 1975). This model allows to observe and predict perceived value from a customer’s viewpoint. It investigates 1) consumers’ attitudes towards a certain behaviour and 2) their perceptions of whether they should execute the behaviour. Notable is that the TRA is only used since 1984 to also predict behaviour instead of only describing behaviour. In 1991 the model improved with a social factor (perceived behavioural control) since social factors are a determinant in real life for individual behaviour (Ajzen, 1991). In this same research the theory of planned behaviour (TPB) was introduced to overcome the imperfection of the TRA concerning social influences. Furthermore, in 2006 the TPB model was extended with 1) the attitude towards searching online and 2) consumers’ past behaviour with using e-coupons (Kang et al., 2006). Moreover, the model became suited for online coupons.
Nevertheless, the TPB model is rarely used in literature to investigate consumers’ intention usage of coupon codes in relation to shopping cart abandonment. Consumers who are prompted with a coupon code field tend to perceive lesser purchase completion intentions (Oliver & Shor, 2003; Rajamma, Paswan, & Hossain, 2009) which can result in abandonment behaviour. Furthermore, when eventually obtained a coupon code, a consumer creates more favourable purchase intentions towards the purchase (Chen & Lu, 2011). So relating this literature to the TPB model the attitude towards using coupon codes is affecting abandonment behaviour if a coupon code is absent. Therefore, a consumers’ intention to use coupon codes will result in a positive effect on shopping cart abandonment.
H4: The effect of shopping cart abandonment is strengthened if the intention to use coupon codes is higher.
Page |15 consumers who are more discount prone tend to judge discount possibilities as ‘good deals’, and therefore do less price comparisons between retailers (Zeithaml, 1988). This implies failure to include coupon proneness as a mediator in the TPB model will result to an incompleteness of describing consumer reaction of using e-coupons (Chen et al., 2011).
H5: The intention of consumers to use coupon codes is higher if a consumer is more prone towards the act of using coupon codes.
Figure 2.
The TPB model of Kang et al. (2006) defines five factors that influence consumers’ proneness towards using coupons. The first factor is the attitude of consumers towards a the act of using e-coupons). This can be defined in consumers’ behavioural beliefs (expectations) of what the outcomes could be while performing a certain behaviour (Fishbein et al., 1975; Chen & Lu, 2011). Other research (Kim, Kim, Im, & Shin, 2003) stated that the intention to execute a certain behaviour is positively influenced by favourable attitudes toward a certain research. Shim, Eastlick, Lotz, & Warrington (2001) give a similar conclusion for online shopping. The expectation therefore is that a consumer is more coupon prone when the consumer has a positive attitude towards using e-coupons.
H6a: A positive attitude towards e-coupons results in a higher proneness towards using coupon codes.
Second is consumers’ attitude towards the efforts that have to be made to acquire a coupon code. In other words the attitude towards the investment of search costs that a consumer has to make (Kang et al., 2006). Consumers have to be willing to invest a certain amount of search and time costs into the acquisition of a discount possibility (King, 1999). If these investments outweigh perceived benefits a consumer could choose not to search for a coupon code and simply finalize or terminate the purchase. An example in which this could happen are consumers that are ‘lazy’, low on computer skills or generally unskilled. Their investment should be significantly higher than other consumer types which could imply a lower prone attitude towards searching for a coupon code.
Page |16 Third are the influences from others that gives direction to a consumer’s expected behaviour to search for, and use, coupon codes. It is both the perceived expectations from relevant people and groups as a person’s own intentions to comply to these expectations that are relevant (Ajzen, 1991; Chen & Lu, 2011). In other words it is defined as the extent to which a consumer wants to comply to the wishes of others (Mathieson, 1991), or social influences (Venkatesh, Morris, Davis, & Davis, 2003). This motivation is higher when social pressure to execute a certain behaviour is also higher (Baker, Al-Gahtani, & Hubona, 2007). Moreover, a consumer’s intention to purchase something online is increased by subjective norms (Pavlou & Fygenson, 2006) and electronic word of mouth (Jalilvand & Samiei, 2012). Therefore, the chance of complying towards subjective norms is high.
H6c: Subjective norms towards coupon codes results in a higher prone attitude towards using coupon codes.
Fourth is the perceived behavioural control that includes a consumer’s perceived capability to control the use of discount vouchers (Ajzen, 1991), or in other words a person’s beliefs how difficult or easy it will be to execute a certain behaviour. This behavioural control is based on control beliefs that stop or stimulate individuals to use coupon codes. Since the effect of perceived behavioural control is higher in a technology-based environment (Rust & Lemon, 2001), the chance that a consumer will use coupon codes is due to the Internet environment therefore higher (Teo & Beng Lee, 2010). So if a consumer has a higher perceived capability to use coupon codes, the chance that a consumer is more coupon prone increases.
H6d: A higher perceived behavioural control of coupon codes results in a higher prone attitude towards using coupon codes.
Fifth and last, a consumer’s past behaviour can give more clarity to a future behaviour (Kang et al., 2006). In other words past behaviour is a meaningful predictor of coupon usage behaviour (Bagozzi, Baumgartner, & Yi, 1992). It is together with perceived behavioural control the two most influential factors of coupon code usage (Chen & Lu, 2011). The expectation is that a consumer with a positive experience in the past will more likely to be coupon prone.
H6e: A positive past experience (behaviour) towards coupon codes results in a higher prone attitude towards using coupon codes.
2.5 Data collection
Page |17
Figure 3.
3. METHODOLOGY
The purpose of this research is to identify consumers’ behaviour when exposed to a coupon code field in the check-out process. In the methodology part the method that is used to research this behaviour is explained.
3.1 Data collection
The total population consist of all Dutch consumers that have purchased something online in 2012 and is based on information of the Central Bureau of Statistics (CBS). The population amounts a total of 10.500.000 consumers. The sample data was collected by executing an online survey (Appendix A) hosted by ThesisTool.com. A total of 3.000 Twitter followers and 170.000 Facebook followers of
Page |18 2 2 2 2 2 2
(1
)
10.500.000 1,96 0,5(1 0,5)
385
(1
) (
1)
1,96 0,5(1 0,5) (10.500.000 1) 0, 05
N z
n
z
N
F
ρ
ρ
ρ
ρ
⋅ ⋅
−
⋅
⋅
−
>=
=
=
⋅
−
+
− ⋅
⋅
−
+
− ⋅
Figure 4 - Finite population calculation.
3.2 Measurement scales
All measurement indicators will be measured using a seven point Likert scale. The dependent variable ‘shopping cart abandonment for shop X’ will be measured by presenting the four constructs as listed in table 1. The participant will be shown one of two shopping situations by presenting different screenshots of the last page before the consumer definitively confirms an order online. By using two different screenshots the variable ‘prompting a coupon code for shop X’ can be tested. The first screenshot lacks a coupon code prompting field, and the second screenshot does contain a coupon code prompting field. Using Thesistools.com the respondents will be directed to one of the two versions of the survey randomly, in which each survey version contains one of the two screenshots. Constructs Measurement indicators Adapted from Abandonment
behaviour shop X.
SCA1 I will abandon the shopping cart to search for the best possible price.
Close & Kukar-Kinney, (2010) and Negra & Mzoughi (2012).
SCA2 I will delay the purchase process to achieve the best possible price.
Table 1.
To measure the two Independent variables ‘absence coupon code shop X’ and ‘availability coupon code shop Y’, the participant will be asked the four measurement indicators as listed in table 2. To test SX the participant will be simulated in one of two situations. The first situation is a discount website lacking a coupon code, and the second situation is the presence of a coupon code at a discount website.
Constructs Measurement indicators Adapted from Absence of a coupon
code for shop X
SX1 I will not return to the shopping cart to finish the purchase.
Close & Kukar-Kinney, (2010) and Negra & Mzoughi (2012).
SX2 I will abandon the shopping cart to search for a coupon code.
Availability of a coupon code for shop Y
SY1 I will not return to the web-shop of my first choice to finish the purchase.
SY2 I will continue the purchase at the web-shop I obtained a coupon code for.
Table 2.
Page |19 Coupon code prompting field.
Absent Present
Coupon code availability at a discount website.
Absent Situation 1 Situation 2
Present Situation 3 Situation 4
Table 3.
To test hypotheses H5 and H6a until H6e, the measurement indicators and constructs as listed in table 4 will be questioned to the research population. All indicators are based on previous TPB and coupon code research.
Constructs Measurement indicators Adapted from Coupon code usage
intention
CU1 I intend to use a coupon code to conduct product purchases in the future.
Chen & Lu (2011), Huang, Wu, Wang, & Boulanger (2011) and Jalilvand & Samiei (2012). CU2 I predict I will purchase through coupon
codes in the future.
CU3 I usually search for coupon codes on the Internet.
Coupon code proneness
CP1 Redeeming coupon codes makes me feel good.
Chen & Lu (2011) and Wirtz & Chew (2002).
CP2 When using coupon codes, I feel that I am getting a good deal.
CP3 I enjoy using coupon codes, regardless the amount I can save from doing so.
CP4 I am more likely to buy brands or products that have promotional deals.
CP5 Coupon codes and promotional deals have caused me to buy things I normally would not buy.
CP6 Beyond the money I save, redeeming coupons and taking advantage of promotional deal give me a sense of joy. CP7 I have favourite brands, but most of the time
I buy brands I have a coupon code for. Attitude towards using
coupon codes.
AC1 I find using coupon codes useful. Chung, Stoel, Xu, & Ren (2012), Chen & Lu (2011) and Lee & Jun (2007).
AC2 I think using coupon codes is beneficial for me.
AC3 Using coupon codes would improve my performance on the purchase.
Page |20 Attitude towards
internet searching.
AS1 It is easy for me to find a coupon code online.
Chen & Lu (2011) and Lee & Jun (2007). AS2 I find it clear and understandable how to
search for a coupon code.
AS3 It is easy for me to become skilful at searching for a coupon code.
AS4 I find it fun to search for coupon codes. Subjective norms about
coupon codes
Other people close and important to me probably consider my coupon code usage
SN1 foolish/wise Baker, Al-Gahtani, & Hubona (2007) and Chen et al. (2011) SN2 harmful/beneficial
SN3 wise us of time/waste of time SN4 worthless/valuable
SN5 bad/good Perceived behavioural
control
PBC1 I would be able to find a coupon code online. Baker, Al-Gahtani, & Hubona (2007), Chen & Lu (2011), Huang, Wu, Wang, & Boulanger (2011), Jalilvand & Samiei (2012) and Kang, Hahn, Fortin, Hyun, & Eom (2006). PBC2 I have the resources, the knowledge and the
ability to find coupon codes.
PBC3 As far as the coupon code I need is available online, it is easy for me to find it.
PBC4 There are no obstacles for me to use e-coupons.
PBC5 Using coupon codes is entirely within my control.
Past behaviour PB1 I have experience with searching for coupon codes through Internet within the past six months.
Chen & Lu (2011) and Huang, Wu, Wang, & Boulanger (2011).
PB2 I have used a coupon code in the past.
Table 4.
3.3 Plan of Analysis
Before the main survey was executed a pre-test was done to test if the measurement indicators as listed in table 1, 2 and 3 are suited for aggregation. Validating is done by applying the Cronbach’s Alpha test to every indicator group. Since a Cronbach’s Alpha higher than α=.6 is generally accepted, this condition is also applied in this research. The Cronbach’s Alpha of the pre-test are calculated and presented per construct in table 5. The results give sufficient Cronbach’s Alpha for all eleven constructs.
Page |21 SCA Abandonment behaviour shop X. ,844
SX Absence of a coupon code for shop X. ,915 SY Availability of a coupon code for shop Y. ,665 CU Coupon code usage intention. ,752 CP Coupon code proneness. ,919 AC Attitude towards using coupon codes. ,961 AS Attitude towards internet searching. ,690 SN Subjective norms about coupon codes. ,975 PBC Perceived behavioural control. ,942
PB Past behaviour. ,627
Table 5.
The second step of analysis -after conducting the total survey- is factoring every indicator group to the appropriate constructs. Finally, the data is ready for testing the hypotheses. In table 6 the statistical tests that are appropriate for rejecting or approving the hypotheses are listed.
Hypothesis Appropriate statistical test H1 Independent samples t-test. H2 Independent samples t-test. H3 Independent samples t-test. H4 Moderator regression analysis.
H5 + H6a/e Regression analysis (Baron and Kenny method).
Table 6.
4. RESULTS
In this chapter the research results will be discussed. First, a preliminary analysis is done in which is explained how all measurement indicators are transformed to constructs. Second, the main analysis is done in which the tests as listed in table 6 are done in order to reject or approve the hypotheses. Finally an overview of all hypotheses are given to summarize the findings of the entire chapter.
4.1 Preliminary Analysis
Page |22 4.1.1 Descriptive statistics
The analysis starts with some general characteristics of the sample population and a representation check (table 7). The research sample (n) is reflected to the research population (N), and significant similarities are checked by applying Chi-Square analyses. Both the Chi-Square of gender and age
result in an insignificant group representation (p=.652 and p=.112), indicating that there are no significant differences between sample and population. Unfortunately the third Chi-Square analysis results in a significant difference (p=<.001) between both groups. Deviations in the sample population (in which relatively more higher educated people filled in the questionnaire) could have created this difference resulting in a limitation of the results extracted from this research.
n % (n) % (N) Gender Male 185 46,4 % 51,2 % Female 214 53,6 % 48,8 % Age 16 to 25 years 108 27,1 % 18,7 % 26 to 35 years 83 20,8 % 19,0 % 36 to 45 years 76 19,0 % 21,3 % 46 to 55 years 67 16,8 % 20,8 % 56 to 65 years 41 10,3 % 13,7 % 66 years and elder 24 6,0 % 6,5 % Education
Primary Education 0 0 % 8 % Secondary Education 90 22,6 % 28 % Intermediate Vocational Education 123 30,8 % 32 % Higher Vocational Education 144 36,1 % 21 %
University 42 10,5 % 11 %
Table 7.
4.1.2 Factor analysis
Since H6a up to H6e are taken together into one regression analysis a factor analysis is done to test if the variables as discussed in the literature part are also correct variables according to the survey data. Because not all data is suited for factoring, the Kaiser-Meyer-Olkin (KMO) method and the Bartlett’s test of Sphericity is therefore applied to the data to check the adequacy. In statistics a KMO higher than .5 and a Bartlett’s test of p=<.05 is generally accepted as sufficient. The Bartlett’s test resulted in p=<.001 and KMO=.887; thus the data is appropriate for factoring (table 8).
Nevertheless the outcomes of a test factor analysis gave an insignificant factor for the construct Past
Page |23 (table 8) resulting in also a sufficient KMO (.074) and a significant Bartlett’s test (p=<.001). The two indicators belonging to PB will be tested further on in this paragraph with the use of an internal consistency check.
KMO and Bartlett’s Test Factor Analysis 1 Factor Analysis 2 Kaiser-Meyer-Olkin Measure of Sampling Adequacy ,887 ,874
Bartlett’s Test of Sphericity Sig. ,000 ,000
Table 8.
The output in table 9 represent the eigenvalues (column ‘Total’) associated with each component (factor). In statistics all eigenvalues larger than one represents appropriate components for factoring. Since the input of the factor analysis includes five constructs is an output of five factors most desirable. Unfortunately factors 5 (eigenvalue of .753) or optionally factor 6 (eigenvalue of .580) have insufficient eigenvalues. Therefore, based on the components in table 9 and the used literature input, four factors seem to be appropriate for extracting.
Total Variance Explained after Extraction and Rotation
Component Total % of Variance Cumulative %
1 3,979 20,944 20,944
2 3,915 20,603 41,547
3 3,908 20,568 62,115
4 2,610 13,735 75,850
Table 9.
In table 10 all factor loadings are listed and sorted on size of their loadings. High loadings indicate large significance within a component. In the output al loadings lower than .5 have been supressed, resulting in a clear overview of which indicators fall into which component. The output is identical to the literature input; therefore all separate measurement indicators are computed and transformed to the four components.
Component
Constructs Indicators 1 2 3 4 Attitude towards using coupon codes. AC1 ,885
AC2 ,809
AC3 ,748
AC4 ,892
AC5 ,861
Attitude towards internet searching. AS1 ,784
AS2 ,843
Page |24
AS4 ,676
Subjective norms about coupon codes. SN1 ,814 SN2 ,847 SN3 ,796 SN4 ,897 SN5 ,896
Perceived behavioural control. PBC1 ,850
PBC2 ,876
PBC3 ,891
PBC4 ,808
PBC5 ,721
Table 10.
4.1.3 Internal consistency check
The third step of the preliminary analysis is checking the reliability of the measurement indicators and constructs that not have been checked in the factor analysis. As also with the pre-test data, a Cronbach’s Alpha (CA) test is used to check the internal consistency. To recall, a CA higher than α=.6 is generally accepted. As to be seen in table 11 all CA’s are higher than .6, so therefore all measurement indicators are suited to be transformed to constructs.
Construct Cronbach’s Alpha
SCA Shopping cart abandonment situation A. ,803 SX Absence of a coupon code for shop X. ,659 SY Availability of a coupon code for shop Y. ,660 CU Coupon code usage intention. ,837 CP Coupon code proneness. ,843
PB Past behaviour. ,672
Table 11.
4.2 Main Analysis
This chapter includes the analysis of the hypotheses. The first subparagraph covers the hypotheses related to consumers behaviour towards shopping cart abandonment and prompted discount availabilities. The second subparagraph covers the hypotheses about consumer’s usage intention and proneness towards coupon codes.
4.2.1 Shopping cart abandonment and retailer switching This subparagraph covers the first three hypotheses.
Page |25 H1: Prompting the consumer an ability to achieve a lower price through a coupon code
causes higher degree of consumers abandoning the purchase process.
The corresponding survey question was divided into two groups; one group that was shown a screenshot of a shopping cart without a coupon code prompting field (group 1) and a second group that was presented with a screenshot of a shopping cart that included a coupon code prompting field (group 2). All respondents were asked the extent to which they would think to abandon the shopping cart (construct SCA). An Independent samples t-test is used to analyse the data (table 12).
N Mean Std. Deviation Abandonment behaviour
shop X (SCA)
Coupon code field present. 213 4,380 1,9023 No coupon code field present. 186 3,954 1,9064
Table 12.
Based on a significance level of p smaller than ,05 (p=,026), and a larger mean of shopping cart abandonment in the prompting group (4,380 versus 3,954), hypothesis H1 can therefore be accepted.
Hypothesis 2
H2 The absence of a coupon code for shop X at a discount website causes a higher degree of consumers abandoning the purchase process.
This survey question is also divided into two groups; one group that was simulated a situation without
a discount possibility available for shop X (group 1) and a second group that was simulated a situation
with a discount possibility for shop X (group 2). All respondents were asked the extent to which they would think to return to the shopping cart of shop X to finish their purchase (construct SX). An Independent samples t-test is used to analyse the data (table 13).
N Mean Std. Deviation Returning behaviour to
Shop X
No coupon code available. 210 4,579 1,6651 Coupon code available. 189 5,212 1,4605
Table 13.
Based on a significance level of p smaller than ,05 (p=<,001) and a larger mean of returning behaviour in the coupon code present group (5,212 versus 4,579), there is a significant difference in returning behaviour if a coupon code for shop X is present or absent. Therefore H2 is accepted, implying that significantly lesser consumers are returning to the shopping cart when no coupon code is found for shop X at a discount website.
Hypothesis 3
Page |26 This survey question uses the same 2 groups as with hypothesis 2 because we want to know how retailer switching is involved by the presence or absence of a coupon code for shop X. All respondents were asked the extent to which they would think to switch to another shop (shop Y) if they are confronted with a coupon code for this competitive shop whilst searching for a coupon code for shop X (construct SY). An Independent samples t-test is used to analyse the data (table 14).
N Mean Std. Deviation Switching behaviour to
shop Y
No coupon code available. 210 5,683 1,4280 Coupon code available. 189 5,312 1,5160
Table 14.
Based on a significance level of p smaller than .05 (p=.012), and a larger mean of switching behaviour in the coupon code absent group (5,683 versus 5,312), there is a significant difference in switching behaviour to shop Y if a coupon code for shop X is unavailable. Therefore H3 is accepted, implying that significantly more consumers are switching from web shop if no coupon code is found or present for shop X at a discount website.
4.2.2 Coupon code usage intentions
This subparagraph covers hypotheses four to six. Hypothesis four will be tested using a moderator-regression-analysis, and hypothesis five and six will be taken together into one regression-mediation-analysis.
Hypothesis 4
H4 The effect of shopping cart abandonment is strengthened if the intention to use coupon codes is higher.
The analysis is done using a moderator regression in which the coupon usage intention is applied as an interaction effect to the relationship of shopping cart situation on shopping cart abandonment.
Regression Model Summary
R R Square Adjusted R Square Std. Error of Estimate
,325a ,106 ,099 1,8166
Table 15. The correlation (table 15) between the two variables and shopping cart abandonment is .099, and explains .099 of the total variance of shopping cart abandonment. This is a rather small effect that could be clarified by consulting the regression coefficient table (16).
Regression coefficients
Page |27 CU Coupon code usage intention. ,369 ,058 ,000 SCA_SITUATION*CU Combined. -,045 ,115 ,693
Table 16. Only coupon code usage intention is significant in table 16. It suggests that the higher a consumer’s intention to use coupons, the more the chance that a consumer abandons the shopping cart to search for a coupon code. With an effect of .369 (with p=<.001) hypothesis four is accepted.
Hypotheses 5 and 6
H5 The intention of consumers to use coupon codes is higher if a consumer is more prone towards the act of using coupon codes.
H6a to H6e Attitude towards internet searching, perceived behavioural control and attitude, past experience and subjective norms towards coupon codes result in a more prone attitude towards using coupon codes.
The five coupon code usage indicators (X) have to be tested on coupon usage intention (Y) with coupon proneness (M) in a mediating role. Thus, the Baron and Kenny regression mediator analysis is therefore an appropriate method (Baron & Kenny, 1986).
This Baron and Kenny method is a four-step method in which the second and third step must be done in one regression. The first step investigates the effect of X on M, the second step investigates the effect of X on Y and simultaneously the effect of X and M on Y (step three). All three steps are done using a regression analysis. The fourth step is done by applying the Sobel test.
Step 1: The effect of the 5 usage indicators on coupon proneness.
The first regression analysis tests the effect of X on M. If the outcome does not result in a significant regression coefficient, the mediation effect is absent. In table 17 the regression summary is presented.
Regression Model Summary
R R Square Adjusted R Square Std. Error of Estimate
,740a ,547 ,541 ,8251
Table 17. The correlation between X and Y is .541. Thus, the indicators (X) explain around 54 percent of the total variance of Y.
Regression coefficients
1. Effect of X on M. Coefficients Collinearity B (a) Std. Error Standardized B Sig. Tolerance VIF (Constant) 1,622 ,205 ,000
Page |28 coupon codes.
AS Attitude towards
internet searching. -,047 -,057 -,057 ,255 ,464 2,153 SN Subjective norms
about coupon codes. ,100 ,104 ,104 ,008 ,749 1,335 PBC Perceived behavioural
control. -,067 -,081 -,081 ,066 ,596 1,677 PB Past behaviour. ,043 ,065 ,065 ,189 ,480 2,083 Table 18. First, the multicollinearity of the data is tested with the use of the Tolerance and VIF variables (table 18). Since all Tolerance variables are larger than .10 and all VIF ‘s are smaller than 10, no multicollinearity takes place. Second, the B coefficients in table 18 indicate that only AC (p=<0.001) and SN (p=.008) can be present in a mediating role with M and Y, in which AC has the largest effect on CP with a standardized B coefficient of .695. This however concludes little since AC and SN only satisfy the first condition of Baron and Kenny. In step four, the significant B coefficients in table 18 will be referred to as Baron and Kenny’s a value.
Step 2 and 3: The effect of the 5 usage indicators and coupon proneness on coupon usage intentions. These two steps involve a two-step regression. First a regression with X as predictor and Y as the dependent variable is done to prove there is no direct relationship between X and Y. The second step involves a regression analysis with X and M as predictors of the dependent variable Y to prove that a mediating effect is present. The first outcomes of this regression is the model summary as displayed in table 19 below.
Regression Model Summary
R R Square Adjusted R Square
Std. Error of Estimate The effect of X on Y. ,660a ,436 ,429 1,2039 The effect of X and M on Y. ,725b ,526 ,519 1,1054
Table 19. The correlation between X and M is ,429 and increases to ,519 when M is incorporated in the regression. Thus, the variance of X explains 43 percent of the variance of Y. Furthermore, if X is taken together with M the total explained variance of Y is 52 percent.
Page |29 Regression coefficients
Coefficients Collinearity 2. The effect of X on Y. B (c) Std. Error Std. B Sig. Tolerance VIF
(Constant) 1,355 ,299 ,000 AC Attitude towards using coupon
codes. ,475 ,050 ,442 ,000 ,659 1,518 AS Attitude towards internet
searching. ,136 ,061 ,125 ,025 ,464 2,153 SN Subjective norms about
coupon codes. -,015 ,055 -,012 ,780 ,749 1,335 PBC Perceived behavioural control. -,017 -,053 -,016 ,752 ,596 1,677 PB Past behaviour. ,222 ,048 ,253 ,000 ,480 2,083 3. The effect of X and M on Y. B (c’) Std. Error Std. B Sig. Tolerance VIF
(Constant) ,412 ,295 ,164 AC Attitude towards using coupon
codes. ,143 ,060 ,133 ,018 ,387 2,586 AS Attitude towards internet
searching. ,164 ,056 ,150 ,003 ,463 2,160 SN Subjective norms about
coupon codes. -,074 ,051 -,059 ,150 ,736 1,359 PBC Perceived behavioural control. ,022 ,049 ,021 ,650 ,591 1,692 PB Past behaviour. ,197 ,044 ,225 ,000 ,478 2,092 CP (M) Coupon code proneness ,582 (b) ,068 ,445 ,000 ,453 2,209 Table 20. In regression 2, AC (B=.442; p=<0.001), AS (B=.125; p=.025) and PB (B=.253; p=<.001) have a significant effect on Y, meaning that only these three variables satisfy the second condition of Baron and Kenny. Regression 2 gives a significant effect (B=.445; p=<.001) for the mediator CP (Baron and
Kenny’s b value) on the dependent variable Y, indicating that mediation condition three is satisfied.
Thus, merely AC satisfies both condition one and two, and with condition three also significant AC is the only X variable that could be present in a mediation situation.
Step 4: Sobel test.
Page |30 to do this. As a short summary, the analysis of step 1 resulted in an a value, step two in a c value and step three with a c’ and b value. All together the previous three steps have to meet the following three conditions; 1) a is significant, 2) b is significant and 3) c minus c’ is significant. The first two conditions are only met for AC, so the other four usage indicators will be dropped out of this step of the regression. In table 21 the outcomes of the Sobel test is displayed.
Input Output
A Sa B Sb Sobel Sig.
AC ,572 ,034 ,582 ,068 7,6283 ,000 Table 21. The Sobel test gives a significant result for consumers attitude towards coupon codes (S=7,63; p=,000) indicating that the effect from this variable on coupon usage intentions is mediated by coupon proneness. Besides, since c’ is significant (p=,018) the mediation is partial.
4.3 Hypotheses Overview
In table 25 below the outcomes of the statistical tests of paragraph 4.2 are clearly and briefly listed.
Hypothesis Result
H1 Prompting the consumer an ability to achieve a lower price through a coupon code causes higher degree of consumers abandoning the purchase process.
Accepted. H2 The absence of a coupon code for shop X at a discount website causes a
higher degree of consumers abandoning the purchase process.
Accepted. H3 The presence of a coupon code for competitor shop Y results in a higher
effect of retailer switching from shop X to Y.
Accepted. H4 The effect of shopping cart abandonment is strengthened if the intention to
use coupon codes is higher.
Accepted. H5 X on Y: The intention of consumers to use coupon codes is higher if a
consumer is more prone towards the act of using coupon codes.
Partially accepted. H6a AC: A positive attitude towards e-coupons results in a higher proneness
towards using coupon codes.
Accepted. H6b AS: A positive attitude towards the efforts of internet searching results in a
higher prone attitude towards using coupon codes.
Rejected. H6c SN: Subjective norms towards coupon codes results in a higher prone
attitude towards using coupon codes.
Accepted. H6d PBC: A higher perceived behavioural control of coupon codes results in a
higher prone attitude towards using coupon codes.
Rejected. H6e PB: A positive past experience (behaviour) towards coupon codes results in a
higher prone attitude towards using coupon codes.
Rejected.
Page |31 5 DISCUSSION
The final chapter will conclude this research by discussing its main conclusions. Next the implications and suggestions to managers from the research are given. The chapter concludes with the limitations of this research and subsequently the suggestions for further research.
5.1 Conclusion
The
main question of this study is to what extent is the absence of a coupon code for a consumer’s current web store choice, or the presence of a coupon code for a competitive web store, an influencer for abandonment behaviour and/or store switching? The findings will be an important contribution to the question of how retailers should deal with discount websites and vice versa. Moreover, this study covers the attitude and proneness of consumers towards coupon codes in order to give more substantiation and interpretation to the findings.Abandonment behaviour
The first hypothesis covers consumers perceiving overspending (a financial risk or uncertainty ; Cho et al., 2006) due to acquiring new knowledge in the check-out process about the possibility to achieve a lower price. It suggest that the presence of a coupon code field results in more consumers abandoning the purchase process. This research acknowledges (thus accepting H1) that the chance of abandoning the shopping cart is significantly lesser in situations where a prompting field is absent versus situations in which the prompting field is present. These findings support earlier research that found that 27 % of all consumers that abandon the shopping cart do so to search for a coupon code (PayPal and comScore, 2009). Although different studies identified the motivations of dropping or abandoning the shopping cart (Rajamma et al., 2009; Greenleaf et al., 1995; Cho et al., 2006) this research is the first to insert the ‘aroused awareness to achieve lower prices’ to the model.
Page |32 The third hypothesis covers the effect of retailer switching to a competitive shop when a consumer is confronted with a coupon code for a competitive shop at a discount website. This research found a significant difference in switching behaviour to a competitive shop (Y) if a coupon code for shop X is unavailable. Therefore H3 is accepted, implying that significantly more consumers are switching from web shop if no coupon code is found or present for shop X at a discount website. This is plausible since Keaveny et al. (2001) and Parathasarathy (2001) found that online consumers are less loyal and therefore more likely to switch retailer when confronted with a better deal.
All three hypotheses confirm the importance of consistency. If a consumer is pointed at the availability of reducing the price by entering a coupon code, this follows in abandonment behaviour if the occurred need for a coupon code is not satisfied at a discount website. Thus, a retailer should not make expectations that it cannot live up to. Consequently, applying a coupon code pricing strategy requires an implementation at every phase of a consumers purchase process (including discount websites). Or vice versa, coupon code prompting fields should be customer-initiated instead of retailer-initiated.
Coupon code usage intentions
The fourth hypothesis links a consumers intention to use coupon codes to shopping cart abandonment. It states that the effect of shopping cart abandonment is strengthened if the intention to use coupon codes is also higher. Coupon code usage intentions is therefore tested (and significantly accepted) as a moderator on H1, indicating that a higher degree of usage intentions strengthens abandonment when a coupon code is prompted. This is in line with research of Chen et al. (2011) and Oliver et al. (2003) both confirming the relevance of a consumers intention to the relationship of coupon codes and abandonment behaviour.
In the literature chapter the intention of consumers to use coupon codes is further explained by the Theory of Planned Behaviour by i.a. Khang et al. (2006). It states that five factors are the determinants of coupon code usage intentions and that the relationship is mediated by a consumers inherent desire or proneness to use coupon codes. Unfortunately four of the five factors reject any indirect effect on coupon code usage intention, which therefore is contradictory to the proposed literature in this research . Only a consumers attitude towards using coupon codes is of significant positive influence on usage intention with coupon proneness in a mediating role. This is in line with research that state that consumers are more prone to execute a certain behaviour when having favourable attitudes towards that behaviour (Kim et al., 2003; Shim et al., 2001). Nevertheless the mediation is not a full mediation. The direct effect of attitude towards coupon codes on coupon code usage intentions is also significant indicating that the relationship is a partial mediation.
Page |33 coupon usage behaviour is of influence on the likability that a consumer has an intention to use a coupon code. So together with a consumer's attitude towards coupon codes these three factors have a positive effect on consumers abandonment behaviour to search for a coupon code.
Finally, this paragraph will end by summarizing the answer to the research question as formulated in the introduction chapter. Hence, coupon codes prompting fields and coupon codes itself are important influencers of shopping cart abandonment. It arouses consumer’s awareness of the ability to achieve lower prices and subsequently increases abandonment behaviour. Furthermore, the degree of abandonment is accelerated by a consumer’s intention to use coupon codes. In more detail, a positive attitude towards coupon codes and positive subjective norms towards coupon codes are the initial drivers of this higher usage intention.
5.2 Managerial Implications
This research set out to provide more clearness in the Dutch online affiliate marketing branch. Thus, the outcomes of this study tries to give managerial handles on how to properly implement a coupon code pricing strategy. Based on the results of this study it is proven that 1) coupon code prompting fields strengthen abandonment behaviour, 2) the presence and absence of a coupon code strengthen abandonment and switching behaviour, and finally 3) a consumer’s intention to use coupon codes is an accelerator for the first two conclusions. This leads to two conclusions. First, by confronting consumers with a prompting field in the check-out process retailers stimulate consumers to abandon the shopping cart. Second, coupon codes at discount websites do affect consumers choice for a web-shop. So both parties are cause of the tendency in the market.
Today retailers put a large portion of their online marketing activities in the hands of an affiliate network. By outsourcing these contacts it is important to set clear boundaries. Moreover, doesn’t it distract a consumer from making the sale?
First of all it is wise to discuss the negatives surrounded by prompting a coupon code at the end of the purchase process, which indirectly also raises the question what its additional value is. In that, it encourages consumers to do a different action than the completion of an order. It is an extra input field whilst the golden web-shop rule emphasizes that a consumers effort of making a purchase should be as minimal as possible. It creates unease and uncertainty in such a way that at the moment of being confronted with the input field a consumer is also willing to fulfil (pay) the order. And finally, it is discussed as an erosion of the margin.
Page |34 simply no other affiliate’s cookie being overwritten. Secondly, almost all transactions take place within 24 hours of the click, and of those the vast majority take place within the hour. Thus, the statement of discount websites stealing commission is untrue.
Since abandonment in different situations is acknowledged in this research, it is advised that coupon codes needs to be part of a retailer’s overall business strategy in order for it to make sense to work with discount affiliates. Thus, an implementation of coupon codes with a proper goal or vision (e.g. market growth, new customers) is most desirable. Avoid consumers to abandon the shopping cart. Create specific landing pages for consumers that are being redirected from discount websites. Only let the input field be retailer-initiated when it is supported by good marketing tactics (for example distribute a coupon code that stimulates to increase the order value). Also, additional value per discount website is divergent. Some discount websites are sales increasers due to their own reach (social media, newsletters, search engines, brand etc.). Therefore, apply conversion attribution to assign value to traffic sources that are used before a consumer proceeds to a conversion. Furthermore, better agreements have to lead in a more fertile cooperation. For example avoid consumers being confronted at a discount page for shoe store A about the availability of a discount for shoe store B. Other examples are to remove expired actions, influence consumers to do upselling, create a fan base to expand a coupon code’s reach or create extra exposure using divergent off and online possibilities (e.g. TV-commercials and blogs).
Finally, a consumer is more and more being conscious of the presence of coupon codes. Therefore cannibalisation of margins will always remain a present factor when prompting a coupon field.
5.3 Limitations and recommendations for further research
This research encounters some limitations regarding its choice, scope, data and interpretation.
Choice
Page |35 the prejudices and branch sentiments would eventually result in a shared vision among affiliate networks and retailers.
Scope
The desired and absent shared vision between affiliate networks and retailers is an aspect of the Dutch affiliate market that is not transferable across countries and cultures. This makes it extremely difficult to apply the findings of this research transboundary. This is particularly true since there are great differences within the degree of maturity. A lot of counties are unfamiliar with coupon codes. Especially eastern European, Southern American, African and Asian countries are regions in which coupon codes are still in its infancy. On the other hand forerunners are North America and the United Kingdom, where coupon codes are an established part of the online shopping process. Both retailers, affiliate publishers, affiliate networks and consumers embrace the use of coupon codes. This results in an extensive collaboration between all parties, and combined with progressive market innovations in significantly more sales. Examples of these innovations are for example Internet browser plugins which display the amount of discount or coupon codes currently available for a certain web-shop whilst shopping at that web-shop.
Data
The data obtained from the questionnaire is in accordance with the research population on age and gender. Unfortunately, on education level the distribution is significantly different from the research population. Therefore, abnormal conclusions can derive from this research that could result in a deviant interpretation of conclusions.
Interpretation
Page |37 LITERATURE
Advertising Bureau Nederland (IAB). (2013). Digitale advertising markt blijft groeien. Retrieved from http://www.iab.nl/2013/03/21/omzetgroei-online-advertising-markt-stagneert/
Affiliate Blog. (2011a). Falend Affiliate Management. Retrieved from http://www.affiliateblog.nl/falend-affiliate-management/
Affiliate Blog. (2011b). De Cookie Win Ratio. Retrieved from http://www.affiliateblog.nl/de-cookie-win-ratio/
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision, 50(2), 179-211.
Bagozzi, R., Baumgartner, H., & Yi, Y. (1992). State versus action orientation and the theory of reasoned action: an application to coupon usage. Journal of Consumer Research, 18(4), 505-518.
Baker, E., Al-Gahtani, S., & Hubona, G. (2007). The effects of gender and age on new technology implementation in a developing country: Testing the theory of planned behavior (TPB). Information Technology & People, 20(4), 352-375.
Baron, R., & Kenny, D. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality(51), 1173-1182.
Blattberg, R., & Neslin, S. (1990). Sales Promotion Concepts, Methods and Strategies. Prentice Hall Inc.
Chandon, P., Wansink, B., & Laurent, G. (2000). A benefit congruency framework of sales promotion effectiveness. Journal of Marketing, 64(10), 65-81.
Chang, C., & Chin, Y. (2011). Comparing consumer complaint responses to online and offline environment.
Internet Research, 21(2), 124-137.
Chen, M., & Lu, T. (2011). Modeling e-coupon proneness as a mediator in the extended TPB model to predict consumers' usage intentions. Internet Research, 21(5), 508 - 526.
Chiou-Wei, S., & Inman, J. (2008). Do shoppers like electronic coupons? A panel data analysis. Journal of Retailing, 84(3), 297-307.
Cho, C., Kang, J., & Cheon, H. (2006). Online shopping hesitation. CyberPsychology & Behavior, 9(3), 261-274.
Chung, J., Stoel, L., Xu, Y., & Ren, J. (2012). Predicting Chinese consumers' purchase intentions for imported soy-based dietary supplements. British Food Journal, 114(1), 143-161.
Page |38
Close, A., & Kukar-Kinney, M. (2010). The determinants of consumers’ online shopping cart abandonment.
Journal of the Academy of Marketing Science, 38(2), 240-250.
Donthu, N., & Garcia, A. (1999). The Internet Shopper. Journal of Advertising Research, 39(3), 52-58.
E-lead. (2011). Goedkoper.nl gekocht voor €185.000. Opgehaald van http://www.elead.nl/2011/03/elead-goedkoper-nl/
Fishbein, M., & Ajzen, I. (1975). Belief, Attitude, Intention, and Behavior: An Introduction to Theory and Research.
Fortin, D. (2000). Clipping Coupons in Cyberspace: A Proposed Model of Behavior for Deal-Prone Consumers.
Psychology & Marketing, 17(6), 515-534.
Greenleaf, E., & Lehman, D. (1995). Reasons for Substantial Delay in Consumer Decision Making. Journal of Consumer Research, 22(2), 186-199.
Grewal, D., & Munger, J. (2001). The effects of alternative price promotional methods on consumers’ product evaluations and purchase intentions. Journal of Product & Brand Management, 10(3), 185-197.
Grewal, D., Gopalkrishnan, I., & Levy, M. (2004). Internet retailing: enablers, limiters and market consequences.
Journal of Business Research, 57(7), 703-713.
Grover, V., Lim, J., & Ayyagari, R. (2006). The dark side of information and market efficiency in e-markets.
Decision Sciences, 37(3), 297-324.
Harrington, J., & Leahey, M. (2007). Equilibrium pricing in a (partial) search market: the shopbot paradox.
Economics Letters, 94(1), 111-117.
Hewitson, O. (2010). The truth about affiliates and cookie overwriting. Opgehaald van http://econsultancy.com/nl/blog/6748-the-truth-about-affiliates-and-cookie-overwriting
Huang, Y., Wu, Y., Wang, Y., & Boulanger, N. (2011). Decision making in online auctions. Management Decision, 49(5), 784-800.
Jalilvand, M., & Samiei, N. (2012). The impact of electronic word of mouth on a tourism destination choice: Testing the theory of planned behavior (TPB). Internet Research, 22(5), 591-612.
Kang, H., Hahn, M., Fortin, D., Hyun, Y., & Eom, Y. (2006). Effects of perceived behavioral control on the consumer usage intention of e-coupons. Psychology & Marketing, 23(10), 841-864.
