Do Consumers Always Spend More When Coupon Face Value is ...

He (Michael) Jia, Sha Yang, Xianghua Lu, & C. Whan Park

Do Consumers Always Spend More When Coupon Face Value is Larger? The Inverted U-Shaped Effect of Coupon Face Value on Consumer Spending Level Commonly, a coupon can be applied to one of several vertically differentiated products sold at different prices within the same product line of a brand. With such a product-line coupon, consumers need to decide on the specific product to buy, resulting in different levels of consumer spending. One field data set and four lab experiments demonstrate that the relationship between coupon face value and consumer spending level may not always be intuitively positive; under certain circumstances, it could take an inverted U-shape. The authors develop a threshold-based model to explain the inverted U-shaped effect of coupon face value on consumer spending level and show that this effect occurs when the price level of products is high, when consumers have a strong saving orientation, when they experience low information load from processing a small number of products, when they are inclined to engage in thorough product comparison, or when they have a weak preexisting preference for a specific level of product benefit. Keywords: product-line coupon, consumer spending, savings percentage, inverted U-shape, threshold Online Supplement: http://dx.doi.org/10.1509/jm.14.0510

product line of a brand.1 For instance, consumers may get a $50off coupon with which they can enjoy a price discount on any model of a Dell laptop computer series. In another scenario, consumers may receive an unrestricted $10-off coupon from a restaurant, which can be applied to any combination of dishes and drinks. Extensive marketing research has examined the promotional effects of product-specific coupons, wherein a discount is restricted to a specific product (e.g., Alba et al. 1999; Chandran and Morwitz 2006; Chen, Monroe, and Lou 1998; Chen and Rao 2007; Lee and Tsai 2014; Mishra and Mishra 2011; Nunes and Park 2003; Raghubir 1998; Shiv, Carmon, and Ariely 2005). Yet much less attention has been paid to product-line coupons, wherein an unrestricted discount can be applied to any option within the same product line of a brand. In this scenario, the consumer must decide not only whether to redeem the coupon (e.g., Cheema and Patrick 2008) but also what specific product to buy with the coupon, which may lead to different spending amounts. In this research, we focus on product-line coupons in an amount-off format and examine how the face value of a

rice-based promotions have been widely used by retailers to stimulate product sales (DelVecchio, Krishnan, and Smith 2007). In marketing practice, it is quite common for sellers to offer product-line coupons to consumers. A product-line coupon is not restricted to a specific product. Instead, it can be used to buy one of several vertically differentiated products sold at different prices within the same

P

He (Michael) Jia is Assistant Professor of Marketing, Faculty of Business and Economics, University of Hong Kong (email: [email protected]). Sha Yang is Ernest Hahn Professor of Marketing, Marshall School of Business, University of Southern California (email: shayang@marshall. usc.edu). Xianghua Lu is Professor of Information Systems, School of Management, Fudan University (email: [email protected]). C. Whan Park is Robert E. Brooker Professor of Marketing, Marshall School of Business, University of Southern California (email: choong@marshall. usc.edu). The authors thank the JM review team for their constructive comments and gratefully acknowledge the financial support from the Early Career Scheme of the Research Grants Council of Hong Kong (27503517), the National Natural Science Foundation of China (71422006), the HKU-Fudan IMBA Joint Research Fund (16170704), and the HKU Seed Fund for Basic Research (201609159003). The second author acknowledges support from the Jingdong Retail Research Center at Hunan University, with which she is affiliated. Please address correspondence to Sha Yang or Xianghua Lu. Vikas Mittal served as area editor for this article.

© 2018, American Marketing Association ISSN: 0022-2429 (print) 1547-7185 (electronic)

1The context in which the same coupon is applicable to several vertically differentiated products is different from the one in which the same coupon can be used for different package sizes of the same product (Krishna and Shoemaker 1992).

70

Journal of Marketing Vol. 82 (July 2018), 70–85 DOI: 10.1509/jm.14.0510

product-line coupon could influence consumers’ specific product choice and, consequently, their spending level. One would intuitively expect that a larger coupon face value would incentivize consumers to spend more. In contrast, by developing a threshold-based model, the present research identifies conditions in which the relationship between coupon face value and consumer spending level does not take a positive form but, instead, an inverted U-shape. With one field data set and four lab experiments, we demonstrate that the inverted U-shaped effect of coupon face value on consumer spending level occurs when products are expensive, when consumers have a strong saving orientation, when they experience low information load from processing a small number of products, when they have a high tendency to compare options thoroughly in their decision process, or when they have a weak preexisting preference for a specific level of product benefit. These findings contribute to the price-based promotion literature and offer important implications for managing coupon face value.

promotions increase in-store expenditures. These findings suggest that an increase in the face value of a product-line coupon will propel consumers to spend more by choosing a more expensive product that provides greater benefits due to an increase in their mental budget (Allenby and Rossi 1991). According to this budget-increase perspective, the relationship between the face value of a product-line coupon (C) that consumers receive and their likelihood of choosing a high level of spending (LH) over a low level of spending can be expressed as: (1)

LH = a · C + b,

where b is the baseline attractiveness of the high spending level, and a is positive and represents that consumers’ likelihood of choosing a high level of spending increases as coupon face value increases. Nevertheless, another theoretical perspective suggests that coupon face value may have an opposite effect on consumer spending level. Savings Comparison

Conceptual Framework A consumer’s spending decision during a transaction is highly susceptible to the influences of sales promotions (e.g., Park, Iyer, and Smith 1989; Ramanathan and Dhar 2010). As Table 1 summarizes, prior research has mainly examined how the monetary value of a price-based promotion (i.e., promotion depth) influences consumers’ purchase incidence and quantity in the focal promotion period as well as their responses in the postpromotion period, in the case of restricted product-specific promotions. The present research fills a gap in the literature by focusing on the circumstance of unrestricted product-line coupons and by investigating how coupon face value shapes consumers’ spending level under this circumstance. To simplify the theoretical analysis, we begin with a scenario in which consumers make a binary choice between a high-priced, high-benefit option (i.e., a high spending level) and a low-priced, low-benefit option (i.e., a low spending level) and then discuss the boundaries of our analysis. The aim of this research is to examine how consumers’ choice of a high spending level (SH) over a low spending level (SL) varies as the face value of a product-line coupon in an amount-off format (C) increases from zero to the lower bound of potential spending (0 < C < SL < SH). We first discuss two forces that may drive consumers’ spending decision opposingly and then analyze how the relative strengths of these two forces may shift as coupon face value increases. Budget Increase Consumers’ total amount of spending is constrained by their mental budget (Karlsson et al. 2004, 2005; Larson and Hamilton 2012; Stilley, Inman, and Wakefield 2010; Van Ittersum, Pennings, and Wansink 2010). When a coupon is present, consumers may experience an increase in their mental budget for product expenditure and thus are encouraged to spend more, which was documented by Heilman, Nakamoto, and Rao (2002) as a “psychological income effect.” Similarly, Drèze, Nisol, and Vilcassim (2004) found that

Research in marketing and economics has shown that when people evaluate price discounts, they often base their spending decision on the savings percentage associated with a specific discount, which is the ratio of the monetary value of the discount to the original product price (Nunes and Park 2003; Saini, Rao, and Monga 2010; Saini and Thota 2010; Tversky and Kahneman 1981). From this perspective, it is likely that consumers may focus on savings percentages associated with redeeming a product-line coupon for different products in their spending decision. Given that consumers could enjoy a greater savings percentage by applying the same coupon to a less expensive product than to a more expensive one, the less expensive product would become more attractive when consumers base their spending decision on savings percentages. To further illustrate this perspective in a simplified binary choice scenario, consumers may compare the savings percentage associated with choosing a high spending level (C/SH) with that associated with choosing a low spending level (C/SL). Consequently, their likelihood of choosing a high spending level (LH) over a low spending level is determined by the relative savings percentage, which is denoted by C/SH – C/SL, such that: (2)

LH = C=SH - C=SL + b = ðSL - SH Þ=ðSL · SH Þ · C + b;

where b again represents the baseline attractiveness of the high spending level, and (SL – SH)/(SL · SH) is negative because SL is smaller than SH. As Equation 2 indicates, as coupon face value increases, the relative savings percentage related to choosing the high (vs. the low) spending level decreases. Consequently, consumers’ likelihood of choosing the high spending level should also decrease. In summary, the savings-comparison perspective suggests a negative relationship between coupon face value and consumer spending, such that the relative attractiveness of choosing the high spending level is reduced when coupon face value increases.

Coupon Face Value and Consumer Spending Level / 71

TABLE 1 Review of Literature on Influences of Promotion Depth on Consumer Responses Product-Specific Discount

Study

Key Insights into Influences of Promotion Depth on Consumer Responses

Gupta (1988)

Discount depth has a positive influence on purchase incidence and purchase quantity.

Grewal, Marmorstein, and Sharma (1996)

A moderate discount is most likely to increase consumers’ information processing.

Grewal et al. (1998)

The positive influence of discount depth on purchase intention is stronger for consumers with less domain-specific knowledge.

Alba et al. (1999)

A deep discount leads to lower retrospective estimates of a brand’s price level, particularly when price information is easy to process.

Anderson and Simester In the postpromotion period, (2004) a deep discount increases repeated purchases from first-time customers but decreases repeated purchases from established customers.

Purchase Incidence

Purchase Quantity

√

√

Postpromotion Responses

√

√

√

√

√

√

Thomas, Blattberg, and Discount depth has Fox (2004) a positive influence on reacquired customers’ relationship duration with a firm. DelVecchio, Krishnan, and Smith (2007)

A deep discount reduces price expectation and purchase incidence in the postpromotion period.

√

Biswas et al. (2013)

An exaggerated discount decreases purchase intention by inducing an inference about the poor quality of an unknown product.

√

Andrews et al. (2014)

A moderate discount is the most effective in terms of stimulating product purchase when a charitable cause is paired with the discount.

√

Aydinli, Bertini, and Lambrecht (2014)

The positive influence of discount depth on purchase incidence is stronger for products that are rich in affect.

√

72 / Journal of Marketing, July 2018

Product-Line Coupon

In the Promotion Period

√

Choice of Vertically Differentiated Products

TABLE 1 Continued Product-Specific Discount

Study

Key Insights into Influences of Promotion Depth on Consumer Responses

Fong, Fang, and Luo (2015)

The positive influence of the discount depth of a mobile promotion on purchase incidence is stronger for recipients shopping in an area where a business competitor is located.

Del Rio Olivares et al. (2018)

In a relational setting (e.g., insurance), a moderate initial discount of the first contract is the most effective in terms of retaining customers after the first contract has expired.

This article

Under certain circumstances, the face value of a product-line coupon has an inverted U-shaped influence on consumers’ spending levels associated with choices of vertically differentiated products.

Product-Line Coupon

In the Promotion Period Purchase Incidence

Purchase Quantity

Postpromotion Responses

Choice of Vertically Differentiated Products

√

A Threshold-Based Analysis Ostensibly, the savings-comparison perspective leads to a prediction opposite to that based on the budget-increase perspective. We reconcile these two mechanisms by analyzing when each would be more prevalent. We argue that, consistent with one’s intuition, the budget-increase mechanism is the default mechanism driving the effect of coupon face value because it serves as a straightforward heuristic for consumers. In contrast, the savings-comparison mechanism is more cognitively complex and effortful, and thus its activation should depend on specific conditions. Relevant to the focal issue, previous research has shown that consumers are not responsive to savings percentages of pricebased promotions unless the savings percentages exceed a certain threshold (Chen, Monroe, and Lou 1998; Gupta and Cooper 1992). From this threshold account, we predict that the relative strengths of the budget-increase mechanism and the savings-comparison mechanism may vary as coupon face value increases. This is because the magnitude of the relative savings percentage associated with choosing a high spending level over a low spending level (C/SH – C/SL) is directly determined by coupon face value (C). Consider the following example for a numerical illustration. Suppose that the same coupon can be used for buying either Product A (price = $40) or Product B (price = $60). When coupon face value is as small as $5, the relative savings percentage associated with choosing Product A over Product B is

√

√

only 4.1% (i.e., $5/$40 – $5/$60). Such a difference may be too small to exceed the threshold above which consumers start to use the relative savings percentage as an important basis for decision making (Chen, Monroe, and Lou 1998; Gupta and Cooper 1992). Thus, we expect that when the face value of a product-line coupon is relatively small, the savings-comparison mechanism will not be activated. At this stage, we expect that coupon face value will influence consumer spending level mainly through the budget-increase mechanism and thus will have a positive impact on consumer spending level. In the same numerical example described in the last paragraph, if coupon face value increases to a large amount, such as $25, the relative savings percentage becomes 21% (i.e., $25/ $40 – $25/$60). At this stage, the relative savings percentage (i.e., 21%) may become large enough to exceed a certain threshold (Chen, Monroe, and Lou 1998; Gupta and Cooper 1992) so that it is more likely to be used as an important decision input for consumers. As a result, coupon face value may influence consumers’ spending level mainly through the savingscomparison mechanism that suppresses the budget-increase mechanism. Consequently, consumers could be more attracted by a lower spending level that is associated with a higher savings percentage, and consumer spending will decrease as coupon face value further increases. Building on these threshold-based analyses, we propose that the face value of a product-line coupon in an amount-off format might have an inverted U-shaped effect on consumers’ total amount of spending.


to take advantage of an increase in their mental budget by simply spending more when coupon face value increases or constantly inclined to choose the high-priced option regardless of coupon face value. Thus, we propose:

FIGURE 1 Conceptual Framework Saving Orientation (H1)

Tendency for Comparison (H3)

Coupon Face Value

Information Load (H2)

Spending Level

Preexisting BenefitLevel Preference (H4)

Conditions for the Inverted U-Shaped Relationship The previous threshold-based analyses suggest that the savingscomparison mechanism remains inactive when coupon face value is small because the relative savings percentage is too small to be used as an important decision input. In contrast, at large coupon face values, the relative savings percentage becomes large enough to exceed a certain threshold so that it could potentially serve as a meaningful decision input for consumers. However, consumers’ potential adoption of the relative savings percentage as a decision input could still be inhibited in the following circumstances: (1) when consumers have a weak saving orientation and thus do not eventually base their spending decision on the relative savings percentage (H1), (2) when consumers cannot figure out the relative savings percentage in the first place because of either a low cognitive ability constrained by information overload (H2) or a low tendency to compare savings percentages associated with different options (H3), or (3) when consumers have a preexisting preference for a certain product that provides a specific level of benefit and thus do not base their product choice on external factors at all, including savings percentages (H4). In these cases, the savings-comparison mechanism and, consequently, the inverted U-shaped effect of coupon face value on consumer spending will still not emerge. Figure 1 shows the conceptual framework summarizing these moderators. We develop our hypotheses in the following paragraphs. First, we discuss a circumstance under which consumers differ in the motivation to enjoy a greater savings percentage by choosing a low-priced option. Prior research has shown that consumers can vary substantially in frugality, in terms of being “restrained in acquiring and in resourcefully using economic goods and services to achieve longer-term goals” (Lastovicka et al. 1999, p. 88). In the focal research context, we focus on the saving orientation component of being frugal in product acquisition. When consumers care about saving money and minimizing their product acquisition cost, they should be more attracted by the greater savings percentage associated with choosing a low-priced option. For these consumers, the savingscomparison mechanism should be more likely to emerge at large coupon face values. Therefore, the inverted U-shaped effect should be more likely to occur. In contrast, consumers who have a relatively weaker saving orientation should be less motivated to base their spending decision on a sizable relative savings percentage at large coupon face values. Instead, these consumers may be either more likely


H1: Consumers’ saving orientation moderates the effect of coupon face value on consumer spending level, such that coupon face value has (a) an inverted U-shaped effect for consumers who have a strong saving orientation but (b) a simply positive effect or no effect for consumers who have a weak saving orientation.

Second, while upwardly adjusting one’s spending level according to an increase in his or her available budget is effortless, roughly calculating and comparing savings percentages necessarily requires a greater extent of cognitive processing (Kahneman 2011). This implies that if consumers have already spent a large amount of cognitive resources browsing and processing excessive product information, further calculating and comparing savings percentages would become a highly formidable task because of the depletion of cognitive resources (Shiv and Fedorikhin 1999). In this case, consumers would not be able to figure out the relative savings percentage, which cannot be further used as a decision input as well. Consequently, the savings-comparison mechanism would be unlikely to operate, and the budget-increase mechanism would remain the driver for a positive effect of coupon face value on consumer spending as coupon face value further increases. Thus, we expect that information overload in product presentation will diminish the inverted U-shaped effect and instead foster a positive effect of coupon face value on consumer spending. In contrast, the inverted U-shaped effect should be more likely to occur when product information is not excessive for consumers to process. Thus, we formally propose: H2: Information load moderates the effect of coupon face value on consumer spending level, such that coupon face value has (a) an inverted U-shaped effect when information load is low but (b) a simply positive effect when information load is high.

Third, given that comparing the savings percentages associated with different spending levels to calculate the relative savings percentage tends to be cognitively demanding (Kahneman 2011), a facilitating condition for the savings-comparison mechanism to dominate the budget-increase mechanism at large coupon face values is that consumers indeed engage in a deep level of comparative information processing. Following this logic, the inverted U-shaped effect should be more evident for consumers who have a high tendency to compare options thoroughly in their decision process. In contrast, if consumers have a low tendency for thorough comparison, they would be less likely to figure out the relative savings percentage. Consequently, they would not further rely on the relative savings percentage as a decision input. For these consumers, the savings-comparison mechanism may not be triggered. Instead, the budget-increase mechanism would be still prevalent and lead to an overall positive effect of coupon face value on consumer spending even when coupon face value increases to a large amount. We formally propose: H3: Consumers’ tendency to compare options in their decision process moderates the effect of coupon face value on consumer spending level, such that coupon face value has (a) an inverted U-shaped effect for consumers who have a high

TABLE 2 Summary of Effects of Coupon Face Value on Consumer Spending Level Studya

Sample Size

1

48,787

2

Coupon Face Valueb Stimulus

Moderator

Low 302.8 171.5

Medium 484.6 395.3

High 452.4 489.8

Dependent Measure

Restaurant spending

Price level

High Low

Spending amount (¥)

300

Mobile hard drive

Saving orientation

High Low

35.5% 67.9%

59.7% 69.8%

44.1% Choice share of the 71.3% high-priced option

3

1,162

Korean barbecue

Information load

Highc High Low

14.7 40.4% 44.2%

15.1 42.2% 57.6%

15.5 Spending amount ($) 43.8% Choice share of the 45.6% high-priced option

4

282

Godiva chocolate

Tendency for comparison

High Low

44.8% 37.9%

69.3% 52.0%


5

202

Mobile hard drive

Preference for a high benefit level

High Low

92.0% .0%

89.5% 52.1%


ES1

403

Mobile hard drive (high vs. midprice)


High Low

61.8% 6.7%

58.3% 23.9%

52.1% Choice share of the .9% high-priced option

Mobile hard drive (high vs. low price)


High Low

66.9% 1.0%

78.2% 6.1%

88.3% Choice share of the .3% high-priced option

ES2

161

Godiva chocolate

Brand liking

High Low

61.2% 24.5%

56.4% 58.6%


NI

300

Korean barbecue

Tendency for comparison

High Low

33.0% 32.0%

51.5% 40.2%


a“ES” indicates two bMeans and choice

extension studies for Study 5; “NI” indicates one study that is not included. Details are available on request. shares are estimated from regression coefficients. “Low” and “high” represent the lower and upper bounds of coupon values; “medium” represents the coupon value at which the inverted U-shaped curve reaches its peak point. cHigh information load due to a relatively larger number of products presented.

tendency for comparison but (b) a simply positive effect for consumers who have a low tendency for comparison.

Finally, another underlying assumption for the inverted U-shaped effect is that consumers have a weak preexisting preference for a specific level of product benefit. These consumers need to make a trade-off between price and benefit for different options within the same product line of a brand. Thus, their product choice is easily susceptible to the influences of situational factors, such as coupons. However, if consumers have a strong preexisting preference for a product that provides a specific level of benefit (e.g., a preference for a certain capacity of a mobile hard drive), their product choice should be less influenced by coupon face value. As a result, for these consumers the inverted U-shaped effect will disappear. Thus, we propose: H4: Consumers’ preexisting benefit-level preference moderates the effect of coupon face value on consumer spending level, such that coupon face value has (a) an inverted U-shaped effect for consumers who have a weak preexisting benefit-level preference but (b) no effect for consumers who have a strong preexisting benefit-level preference.

Overview of Studies Five studies explore conditions for the proposed inverted Ushaped effect and investigate a set of moderators theoretically relevant to the savings-comparison mechanism. Study 1 uses

field data to examine consumer spending at restaurants and provides correlational evidence for the inverted U-shaped relationship between coupon face value and consumer spending when the price level of a restaurant is relatively high, which may induce a strong saving orientation (H1) or a high tendency to compare (H3). Studies 2–5 provide more evidence for causality by experimentally manipulating coupon face value in a controlled lab setting and show that the inverted U-shaped effect would be more likely to occur when consumers have a strong saving orientation (H1), experience low information load (H2), have a high tendency for comparison (H3), or have a weak preexisting benefit-level preference (H4). Table 2 summarizes the results of the aforementioned studies as well as three additional studies.

Study 1: Evidence from Field Data In Study 1, we provide preliminary evidence for the existence of an inverted U-shaped effect of coupon face value in an amount-off format on consumer spending level by using actual consumer spending data from restaurants. When eating in restaurants, consumers can choose different combinations of dishes and drinks with different prices and quantities. Therefore, the same coupon can be linked to different levels of total spending on consumption. Because the dishes and drinks are consumed together while consumers dine in restaurants, they can be regarded as parts of an integrated product, making


Data We obtained the field data from a third-party restaurant review site in China (similar to Yelp.com). Consumers who registered for the review site were provided with a membership card and could download coupons available on the review site onto their accounts. When consumers used their membership cards at participating restaurants, the actual transaction amounts were recorded. The data set contained 48,787 observations from a major city in China on a weekly basis from May 2005 to March 2008, with 26,660 registered consumers who spent money in 106 participating restaurants that posted coupons involving amount-off discounts on the review site.2 We regressed total consumer spending per transaction on a set of variables specified as follows: (3) Spending = b0 + b1 FaceValue + b2 FaceValue

2

+ b3 PriceLevel + b4 FaceValue · PriceLevel + b5 FaceValue2 · PriceLevel + e:

The dependent variable was the total amount of money (in Chinese yuan [CNY]) involved in an individual consumer’s single transaction with a specific restaurant (Spending; i.e., total amount paid plus coupon face value). To test the inverted U-shaped effect of coupon face value (FaceValue; in CNY), we included both the linear and squared terms of this variable. We also examined the moderating role of the price level (PriceLevel; in CNY) of a restaurant in the inverted U-shaped effect. Price level was operationalized as the average spending amount per person consumers reported on the review site for a specific restaurant. We expect that a higher price level will activate a stronger saving orientation (H1) or a higher tendency for thorough comparison (H3) and thus facilitate the inverted U-shaped effect. In contrast, we expect that a lower price level will not trigger a strong saving orientation or a high tendency to compare among consumers. Thus, it would be more likely to foster a simply positive effect of coupon face value on consumer spending. To test this prediction, we entered the firstand second-order interactions between coupon face value and price level in the model.3 Results and Discussion Central to our prediction, the model specified in Equation 3 generated a negative second-order interaction between coupon face value and price level (B = -.0002, t = -7.48, p < .001), suggesting that the price level of a restaurant moderated the nonlinear effect of coupon face value on consumers’ spending amount per transaction in the focal restaurant. We further 2We treat other amount-related coupons (e.g., free dishes) as amount-off coupons by entering their equivalent monetary values into the independent variable because these coupons also reduce consumers’ overall acquisition cost in an amount-off format during restaurant visits. 3The rating and volume of the online reviews for a restaurant may influence consumers’ spending level (Lu et al. 2013). When these variables are entered into the model as covariates, the effects of coupon face value remain unchanged.


FIGURE 2 Coupon Face Value, Price Level, and Spending Level (Study 1) Spending Amount Per Transaction (CNY)

consumption in restaurants an ideal empirical context for our hypothesis testing.

500 450 400 350 300 250 200 150 100 0

25

50

75

100

125

150

175

200

225

Coupon Face Value (CNY) High price level (+1 SD) Low price level (–1 SD) Notes: The curves represent the unstandardized regression coefficients.

decomposed this second-order interaction using a spotlight analysis (Aiken and West 1991; Fitzsimons 2008). When price level was high (1 SD above the mean), we observed a positive linear term (B = 2.27, t = 10.78, p < .001) and a negative squared term (B = -.01, t = 6.67, p < .001) of coupon face value. Initially, consumer spending increased with coupon face value. Yet after coupon face value reached 162 CNY (approximately US$24), a further increase in coupon face value started to decrease consumer spending. In contrast, when price level was low (1 SD below the mean), the model only revealed a simply positive effect (B = 1.39, t = 11.24, p < .001) of coupon face value on consumer spending (for a graphical illustration, see Figure 2; for the detailed statistics for this and other studies, see the Web Appendix). Study 1 demonstrates an inverted U-shaped effect of coupon face value on consumer spending when price level is high but a positive effect of coupon face value on consumer spending when price level is low. This study has two limitations. First, the findings are preliminary given that the nature of the data is correlational. We could not exclude the possibility that consumers who actively look for high-value coupons are less likely to buy high-priced items in the first place. Second, we assume that a higher price level may trigger a stronger saving orientation (H1) or a higher tendency for comparison (H3). These two proposed constructs await further testing. We address these two limitations in the follow-up lab experiments. Specifically, we examine saving orientation in Study 2 and tendency for comparison in Study 4.

Study 2: The Moderating Role of Saving Orientation In Study 2, we manipulated coupon face value to provide direct evidence for the causal relationship between coupon face value

Design and Procedure Study 2 adopted a 5 (coupon face value: $5, $15, $25, $35, or $45) · 2 (saving orientation: weak vs. strong) between-subjects design, with saving orientation measured as an individual difference variable. Three hundred U.S. residents (145 women; Mage = 35.69 years, SD = 11.76) from Amazon Mechanical Turk (MTurk) participated for monetary compensation. In this study, participants were asked to imagine that they had received a coupon that could be used to buy a mobile hard drive at an online store. We chose this product category because the performances of different mobile hard drives can be unambiguously differentiated. Thus, participants had to make a clear trade-off between product benefit and product price, making their product choice highly susceptible to the influence of coupon face value. Participants imagined that the coupon they received could be used to buy one of two mobile hard drives of the same brand. The two mobile hard drives differed in capacity (1,000 GB vs. 2,000 GB), revolutions per minute (RPMs; 5,400 vs. 7,200), and price ($55.45 vs. $99.95; for a description of product stimuli and a coupon example, see Figure WA1 in the Web Appendix). Participants indicated which of the two mobile hard drives they would like to buy using the coupon they received. They could also choose to not redeem the received coupon, such that they had three options in total (i.e., low-priced hard drive, high-priced hard drive, or no purchase). At the end of the survey, we measured participants’ saving orientation by borrowing items related to saving money in product acquisition from the frugality scale (Lastovicka et al. 1999), including “I am willing to wait on a purchase I want so that I can save money,” “There are things I resist buying today so I can save for tomorrow,” “I believe in being careful in how I spend my money,” and “I discipline myself to get the most

FIGURE 3 Coupon Face Value, Saving Orientation, and Spending Level (Study 2) 1.0 Probability (High-Priced Option)

and consumer spending level. Whereas Study 1 provides preliminary evidence for the moderating role of saving orientation by assuming that it is associated with price level, in Study 2 we directly measured participants’ saving orientation. If the inverted U-shaped effect is indeed driven by the fact that the savings-comparison mechanism dominates the budget-increase mechanism at larger coupon face values, this effect should be more evident when consumers have a strong saving orientation. In contrast, if consumers have a weak saving orientation, the savings-comparison mechanism will not be triggered, and thus the inverted U-shaped effect should not be observed (H1). In this study, we simplified consumer spending level to a binary choice between a low-priced, low-benefit product and a high-priced, high-benefit product from the same product line of a brand. We also provided participants with a third option to not redeem their coupons for the two presented products (i.e., no-purchase option). In this setup, we were able to examine how the face value of a product-line coupon influences both product category purchase incidence (i.e., whether participants would choose to redeem their coupons in the first place) and spending level (i.e., which product participants would choose when they have decided to redeem their coupons). Spending level (rather than product category purchase incidence, which has been extensively examined in prior research) is the primary focus of the present research.

.9 .8 .7 .6 .5 .4 .3 .2 .1 .0 0

5

10

15

20

25

30

35

40

45

50

Coupon Face Value (USD) Strong saving orientation (+1 SD) Weak saving orientation (–1 SD) Notes: The curves represent the unstandardized regression coefficients.

from my money” (1 = “strongly disagree,” and 7 = “strongly agree”). We also more directly asked participants the extent to which they put more emphasis on “lower price” (1) or “better performance” (7) on a seven-point bipolar scale (reverse coded). These items were averaged to form a saving orientation index (a = .75). Results and Discussion Purchase incidence. First, we examined the effect of coupon face value on product category purchase incidence in a logistic regression (0 = “not purchasing any product,” and 1 = “choosing either the low-priced product or the high-priced product”). Although purchase incidence increased from 86.9% to 93.4% as coupon face value increased from $5 to $45, such an increase was not significant (B = .02, Wald(1) = 1.17, p = .28).4 Furthermore, neither the quadratic effect of coupon face value nor the first- or second-order interactions between coupon face value and saving orientation were significant (ps > .78) when they were added to the logistic regression. Given that, in this study, coupon face value did not significantly affect product category purchase incidence, we dropped the “no-purchase” option from the dependent variable and further investigated the effect of coupon face value on participants’ spending level. Spending level. To examine the moderating role of saving orientation, we entered saving orientation, coupon face value, coupon face value’s squared term, and the firstand second-order interactions between coupon face value and saving orientation into a logistic regression with product choice as the dependent variable (0 = “low-priced option,” 4In all of our studies, the significance levels of the regression coefficients are similar when we code the actual value of the coupon face value variable and when we code it in an increasing level.


and 1 = “high-priced option”). Central to our theorization, the second-order interaction between coupon face value and saving orientation was negative and marginally significant (B = -.002, Wald(1) = 2.69, p = .10). This result indicates that participants’ saving orientation moderated the nonlinear effect of coupon face value on spending level. In a further spotlight analysis (Aiken and West 1991; Fitzsimons 2008), when saving orientation was relatively stronger (1 SD above the mean), the linear term of coupon face value was positive (B = .11, Wald(1) = 3.86, p = .05), and its squared term was negative (B = -.002, Wald(1) = 4.25, p = .04), suggesting that the relationship between coupon face value and consumer spending was inverted U-shaped. Participants’ spending level was the highest when coupon face value was between $25 and $35. When saving orientation was relatively weaker (1 SD below the mean), coupon face value had only a directionally positive yet nonsignificant effect on spending level (B = .004, Wald(1) = .09, p = .77; for a graphical illustration, see Figure 3). As Figure 3 shows, a weaker saving orientation encouraged a stronger preference for the high-priced option (i.e., around 70%) in the first place. Thus, room for an increase in the choice share of the highpriced option could be very limited. Taken together, these results support the moderating role of saving orientation (H1). Study 2 establishes the causal effect of coupon face value on consumer spending level by using an experimental approach and provides support for H1’s assertion that coupon face value has an inverted U-shaped effect on spending level only when consumers’ saving orientation is relatively stronger. In the following experiments, we seek further evidence for the savingscomparison mechanism that underlies the inverted U-shaped effect by examining other theoretically relevant moderators. Given that, in Study 2, we found that coupon face value did not significantly influence product category purchase incidence, in the following studies we do not include a “no-purchase” option in the choice set, to simplify the experimental design.

Study 3: The Moderating Role of Information Load According to our theorization, information overload should inhibit the inverted U-shaped effect because such overload depletes consumers’ cognitive resources that are necessary for carrying out a thorough comparison among savings percentages. Instead, under information overload, consumers should follow a less effortful path by simply adjusting their spending level in line with an increase in their mental budget as coupon face value increases (H2). In Study 3, we tested this hypothesis and, specifically, focused on the description (Townsend and Kahn 2014) and number (Scheibehenne, Greifeneder, and Todd 2010) of products as two sources of information overload. Another purpose of this study is to generalize the findings of Study 2 to another product category. Whereas the mobile hard drive stimuli used in Study 2 were primarily utilitarian, we chose a pair of product stimuli that were primarily hedonic for Study 3. Design and Procedure In Study 3, we examined participants’ choice of food menus and investigated the effects of two sources of information overload: (1)


the description of products and (2) the number of products in a choice set. For this purpose, this study adopted a 4 (coupon face value: $2, $4, $6, or $8) · 2 (perceived information load of product descriptions: low vs. high) · 2 (number of presented products: small vs. large) between-subjects design, with the second factor measured at the end of the study. In total, 1,162 U.S. residents (515 women; Mage = 33.63 years, SD = 11.46) were recruited from MTurk and took a survey for monetary compensation. Participants imagined that they received a coupon that could be used in a Korean barbecue buffet restaurant. Then, they were presented with Korean barbecue buffet menus, which differed in price and variety of dishes. In the small-number condition, participants could choose from only two Korean barbecue buffet menus: a five-item menu at $12.99 and a ten-item menu at $19.99. Those in the large-number condition could choose from six Korean barbecue buffet menus, which contained five to ten items each (these included the two menus presented in the small-number condition; for a description of product stimuli and a coupon example, see Figure WA2 in the Web Appendix). After viewing the coupon and menus, participants indicated their menu choice. At the end of the survey, we measured the perceived information load of product descriptions by asking participants the extent to which they thought that “there was too much information in the menu descriptions” and “it was difficult to process all the menu information” on a seven-point scale (1 = “not at all,” and 7 = “very much”), which formed an information load index (a = .82). Compared with the product stimuli used in Study 2, which listed only three key attributes of mobile hard drives, the stimuli employed in this study listed all the dishes contained in the buffet menus and, thus, presented moderately excessive information even when the number of presented products was small. Such amounts of information would create a reasonable variation in perceived information load among participants and thus facilitate our hypothesis testing regarding the moderating role of information load. We expect that when the number of presented products is small, the effect of coupon face value on consumer spending level will be inverted U-shaped when participants’ perceived information load from processing product descriptions is low. Conversely, this effect will become positive when consumers’ perceived information load is high. In contrast, when the number of presented products is large, the effect of coupon face value on consumer spending level will be positive because the presence of multiple products plus the moderate amount of information per product already increases the perceived information load of all participants to a relatively high level (H2). Results and Discussion When the number of presented products was small. We conducted a logistic regression in which level of spending (0 = “low-priced option,” and 1 = “high-priced option”) was regressed on the linear and squared terms of coupon face value, perceived information load, and their first- and second-order interactions. There was a positive second-order interaction (B = .04, Wald(1) = 5.41, p = .02) between coupon face value and perceived information load, suggesting that the nonlinear effect of coupon face value was moderated by perceived information load.

FIGURE 4 Coupon Face Value, Information Load, and Spending Level (Study 3)

Probability (High-Priced Option)

A: The Small-Number Condition 1.0 .9 .8 .7 .6 .5 .4 .3 .2 .1 .0 0

2

4 6 Coupon Face Value (USD)

8

10

High information load (+1 SD) Low information load (–1 SD)

Study 4: The Moderating Role of Tendency for Comparison

B: The Large-Number Condition 18

Spending Level (USD)

17

16

15

14

13

12 0

2

4 6 Coupon Face Value (USD)

8

When the number of presented products was large. In line with our theorization, a large number of presented products also created higher information load (Msmall = 2.41, SD = 1.40 vs. Mlarge = 3.31, SD = 1.77; F(1, 1,160) = 91.47, p < .001). Because the dependent variable in this condition was no longer a binary choice and instead had six levels, we treated it as continuous and ran an ordinary least squares regression. We found that coupon face value increased participants’ spending level (B = .12, t = 2.53, p = .01; for a graphical illustration, see Figure 4, Panel B) when they experienced higher information load as a result of a larger number of presented products. Taken together, the results support H2 and demonstrate that information load plays an important role in determining the effect of coupon face value on consumer spending, such that the effect takes an inverted U-shape only when consumers experience low information load from processing a small number of presented products. Yet when a larger number of presented products, with a moderate amount of information per product, imposes information overload on consumers, the effect becomes simply positive.

10

Notes: The curves represent the unstandardized regression coefficients.

In a further spotlight analysis (Aiken and West 1991; Fitzsimons 2008), for participants who perceived low information load in product descriptions (1 SD below the mean), there was a positive, marginally significant linear effect (B = .54, Wald(1) = 3.16, p = .08) and a negative, marginally significant quadratic effect (B = -.05, Wald(1) = 3.24, p = .07) of coupon face value, suggesting that the effect of coupon face value was inverted U-shaped. Participants’ choice share of the high-priced menu was the highest when coupon face value was between $4 and $6. In contrast, for participants who perceived high information load (1 SD above the mean), coupon face value had a directionally positive yet nonsignificant effect on spending level (B = .02, Wald(1) = .15, p = .70; for a graphical illustration, see Figure 4, Panel A).

We propose that the occurrence of the savings-comparison mechanism results in the inverted U-shaped effect. In Study 4, we further demonstrate the savings-comparison mechanism by examining the moderating role of consumers’ tendency to compare options in their decision process. Consumers who are more inclined to compare options thoroughly should be more likely to engage in savings percentage calculation and comparison, such that the savings-comparison mechanism will emerge at large coupon face values and result in an inverted Ushaped effect. In contrast, consumers who are less inclined to compare options would not bother to engage in savings calculation and comparison. Thus, for these consumers, the budgetincrease mechanism should always be dominant, resulting in an overall positive effect of coupon face value on consumer spending level (H3). We tested this hypothesis in Study 4. Study 4 also aims to exclude two alternative explanations. The face value of a product-line coupon could affect consumers’ inferences about product quality (Biswas et al. 2013; Chandran and Morwitz 2006; Darke and Chung 2005) and marketers’ persuasion attempts (Hardesty, Bearden, and Carlson 2007). To rule out these two alternative explanations, we directly measured these variables to control for their possible influences in Study 4. Design and Procedure Two hundred eighty-two undergraduate students (174 women; Mage = 20.03 years, SD = 1.42) from a large U.S. West Coast university participated in Study 4. This study adopted a 4 (coupon face value: $5, $10, $15, or $20) · 2 (tendency for thorough comparison: low vs. high) between-subjects design, with the second factor measured as an individual difference variable. We used Godiva chocolate boxes as our focal stimuli. Participants imagined that they received a coupon that could be



FIGURE 5 Coupon Face Value, Tendency for Comparison, and Spending Level (Study 4) 1.0 Probability (High-Priced Option)

used to buy either a Godiva spring chocolate gift box ($28.98, 16 pieces) or a Godiva signature chocolate truffle gift box ($38.98, 18 pieces) for their own consumption. Then, they indicated which gift box they would like to buy (for a description of product stimuli and a coupon example, see Figure WA3 in the Web Appendix). These two boxes differed in price, quantity, and flavor and thus represented two different Godiva products. After participants made their product choice, they further answered a set of ancillary questions regarding the chocolate boxes. First, we measured quality inference by asking participants to rate the two Godiva chocolate gift boxes on two bipolar scales anchored at “very poor quality” (1) versus “very good quality” (7) and “very poor condition” (1) versus “very good condition” (7), which formed a quality inference index (a = .85). Second, we assessed participants’ inference about the persuasion attempt associated with the coupon by asking them the extent to which the promotional offer seemed “to be a sales gimmick used to get consumers to buy,” “to have strings attached,” and to be “too good to be true” (1 = “not at all,” and 7 = “very much”), adapted from Hardesty, Bearden, and Carlson (2007). Given that the reliability of the persuasion knowledge scale was too low (a = .46) in our study, we conducted analyses on the three individual items separately. At the end of the survey, to measure participants’ tendency to compare options thoroughly in their decision process, we followed Parker and Schrift (2011) by combining the items from two subdimensions of the shortened maximization scale (Nenkov et al. 2008), including “When I am in the car listening to the radio, I often check other stations to see if something better is playing, even if I am relatively satisfied with what I’m listening to”; “When I watch TV, I channel surf, often scanning through the available options even while attempting to watch one program”; “No matter how satisfied I am with my job, it’s only right for me to be on the lookout for better opportunities”; “I often find it difficult to shop for a gift for a friend”; “When shopping, I have a hard time finding clothing that I really love”; and “Renting videos is really difficult. I’m always struggling to pick the best one” (1 = “strongly disagree,” and 7 = “strongly agree”). We averaged these six items to form a tendency for comparison index (a = .67), which represents both the breadth and the depth of option comparison in people’s decision process. Prior research has shown that a maximization mindset predicts the effort that people exert in their decision-making process across various domains (Iyengar, Wells, and Schwartz 2006; Levav, Reinholtz, and Lin 2012). In particular, Nenkov et al. (2008) show that the items used in our tendency for comparison index strongly predicted the amount of information participants processed, the time participants took to make decisions, and the number of options participants considered in a decision task. Parker and Schrift (2011) further demonstrate that the tendency for comparison index represents a comparative thinking style. Taken together, the tendency for comparison index captures people’s chronic inclination to perform a thorough comparison among different options in their decision process. As a supplementary measure for tendency for comparison, participants’ decision time was also recorded in this study (i.e., how many seconds they spent before their last clicks on the web

.9 .8 .7 .6 .5 .4 .3 .2 .1 .0 0

5

10

15

20

25

Coupon Face Value (USD) High tendency for comparison (+1 SD) Low tendency for comparison (–1 SD) Notes: The curves represent the unstandardized regression coefficients.

page on which they made their product choices). Because a higher tendency for comparison should lead to longer decision time (Nenkov et al. 2008), we expect that decision time will moderate the effect of coupon face value in a similar way tendency for comparison does. Results and Discussion The role of tendency for comparison. We entered tendency for comparison, coupon face value, its squared term, and its first- and second-order interactions with tendency for comparison into a logistic regression with product choice as the dependent variable (0 = “low-priced option,” and 1 = “highpriced option”). The second-order interaction between coupon face value and tendency for comparison was negative and significant (B = -.01, Wald(1) = 4.14, p = .04), indicating that the nonlinear effect of coupon face value was moderated by tendency for comparison. We further conducted a spotlight analysis to understand this moderation (Aiken and West 1991; Fitzsimons 2008). When tendency for comparison was relatively high (1 SD above the mean), the linear term of coupon face value was positive (B = .49, Wald(1) = 7.46, p = .006), and the squared term of coupon face value was negative (B = -.02, Wald(1) = 7.80, p = .005). The highest choice probability of the high-priced option occurred when coupon face value was between $10 and $15, demonstrating an inverted U-shaped effect of coupon face value on consumer spending level. In contrast, when tendency for comparison was relatively low (1 SD below the mean), there was a simply positive effect of coupon face value on consumer spending level (B = .08, Wald(1) = 5.77, p = .02), confirming H3 (for a graphical illustration, see Figure 5). Moreover, adding product quality or items of persuasion knowledge as a covariate in the logistic regression did not change the results reported

previously regarding the effects of coupon face value and tendency for comparison, which suggests that inferences about product quality and persuasion attempt from different coupon face values could not explain our results. The role of decision time. As predicted, participants’ decision time as a supplementary measure for tendency for comparison was indeed correlated with their tendency for comparison (r = .19, p < .001). Further moderated logistic regression analyses show that although the second-order interaction between coupon face value and decision time did not reach significance (B = -.001, Wald(1) = 1.77, p = .18), the effects of coupon face value at the high and low levels of decision time were similar to those at the high and low levels of tendency for comparison. For participants who spent more time making their product choices (1 SD above the mean), the effect of coupon face value was inverted U-shaped (linear term, B = .44, Wald(1) = 5.74, p = .02; quadratic term, B = -.02, Wald (1) = 5.23, p = .02). In contrast, for those who spent less time making their product choices (1 SD below the mean), the effect of coupon face value was directionally positive (linear term, B = .04, Wald(1) = 1.54, p = .21). Overall, the effects of decision time were consistent with those of tendency for comparison. A comparison of the results of the two moderated regressions also suggests that the tendency for comparison scale is a more sensitive measure than the decision time measure in revealing when the inverted U-shaped effect would emerge, given that for some participants, longer decision time might result from the difficulty in calculating the savings percentage rather than from a thorough product comparison. By demonstrating that consumers’ tendency to thoroughly compare options determines whether the effect of coupon face value on consumer spending level is inverted U-shaped or simply positive, Study 4 further supports the role of the savingscomparison mechanism in shaping the inverted U-shaped effect.

Study 5: The Moderating Role of Preexisting Benefit-Level Preference Study 5 examines consumers’ preexisting preference for a specific level of product benefit as a boundary condition for the inverted U-shaped effect (H4). To provide further evidence for the savings-comparison mechanism, we also measured participants’ consideration of savings percentages in the decisionmaking process. In our theorization, only at large coupon face values do consumers start to base their spending decision on savings percentages. At this stage, the savings-comparison mechanism occurs, resulting in the inverted U-shaped effect. According to such theorization, we expect that participants’ consideration of savings percentages mediates the negative effect (i.e., the “downward” side) of coupon face value on spending level at large coupon face values, but it does not mediate the positive effect (i.e., the “upward” side) of coupon face value on spending level at small coupon face values. Design and Procedure Study 5 used a 5 (coupon face value: $5, $15, $25, $35, or $45) · 2 (preexisting benefit-level preference: weak vs. strong)

between-subjects design, with preexisting preference measured as an individual difference variable. Two hundred two U.S. residents (92 women; Mage = 34.94 years, SD = 10.77) from MTurk participated in exchange for monetary compensation. The procedure and stimuli of this study were the same as those of Study 2, except for the following three differences. First, in Study 5 participants made a binary choice between two mobile hard drives, with the more expensive mobile hard drive providing greater benefits (for a description of product stimuli and a coupon example, see Figure WA1 in the Web Appendix). Second, after participants made their product choice, they indicated the extent to which they thought about the percentage of the list price that they could save by using the coupon when making their choice (1 = “not at all,” and 7 = “very much”). Participants’ rating on this scale serves as a measure for their consideration of savings percentages. Third, at the end of the survey, participants rated the extent to which they had a clear preference for the capacity of a mobile hard drive (1 = “no preference,” and 7 = “a strong preference for 2,000GB over 1,000GB”). A higher score represents a stronger preexisting preference for a higher level of product benefit. Results and Discussion The role of preexisting preference. We entered preexisting benefit-level preference, coupon face value, the squared term of coupon face value, and the first- and second-order interactions between coupon face value and preexisting preference into a logistic regression with product choice as the dependent variable (0 = “low-priced option,” and 1 = “highpriced option”). We obtained a positive second-order interaction between coupon face value and preexisting preference (B = .01, Wald(1) = 11.47, p < .001), indicating that the quadratic effect of coupon face value was moderated by preexisting preference. We further conducted a spotlight analysis (Aiken and West 1991; Fitzsimons 2008). When preexisting preference for the 2,000 GB capacity was relatively weaker (1 SD below the mean), the positive linear term (B = 1.30, Wald(1) = 13.71, p < .001) and the negative squared term (B = -.02, Wald(1) = 13.47, p < .001) of coupon face value confirmed the inverted U-shaped relationship between coupon face value and consumer spending. Participants’ spending level was the highest when coupon face value was between $25 and $35. When preexisting preference for the 2,000 GB capacity was relatively stronger (1 SD above the mean), the effect of coupon face value became nonsignificant because the choice share of the 2,000 GB hard drive was above 88% regardless of coupon face value (B = -.01, Wald(1) = .20, p = .65; for a graphical illustration, see Figure 6, Panel A). These results establish preexisting preference for a specific level of product benefit as a boundary condition for the inverted U-shaped effect (H4). The role of savings percentage consideration. In this study, we also found that coupon face value had an inverted Ushaped effect on consumer spending level in the overall sample even when the moderating role of preexisting benefit-level preference was not modeled. Thus, we further tested the mediating role of savings percentage consideration in the overall sample.


FIGURE 6 Coupon Face Value, Preexisting Preference, and Spending Level (Study 5) A: Preexisting Preference as a Moderator (Main Study)

Probability (High-Priced Option)

1.0 .9 .8 .7 .6 .5 .4 .3 .2 .1 .0 0

5

10

15 20 25 30 35 40 Coupon Face Value (USD)

45

50

Strong preexisting preference (+1 SD) Weak preexisting preference (–1 SD)

Extension Studies for Study 5

B: Brand Liking as a Moderator (Extension Study 2) Probability (High-Priced Option)

1.0 .9 .8 .7 .6 .5 .4 .3 .2 .1 .0 0

5

10 15 20 Coupon Face Value (USD)

25

Strong brand liking (+1 SD) Weak brand liking (–1 SD) Notes: The curves represent the unstandardized regression coefficients.

Consistent with our theorizing, a regression analysis shows that participants’ consideration of savings percentages increased as coupon face value increased, indicating that savings percentages indeed became a more important decision input for participants as the percentages increased (B = .03, t = 3.38, p = .001). Given that the effect of coupon face value on spending level was initially positive but started to turn negative when coupon face value was between $25 and $35, we further examined the differential effects of savings percentage consideration in two data ranges, from $5 to $25 and from $35 to $45, respectively.


We conducted mediation analyses using a bootstrapping approach (Hayes 2013). When coupon face value was from $5 to $25, neither the direct effect of savings percentage consideration on consumer spending (B = -.08, Wald(1) = .80, p = .37) nor the mediating effect of savings percentage consideration between coupon face value and consumer spending (95% confidence interval = [-.011, .002]) was significant. In contrast, at larger coupon face values from $35 to $45, both the direct effect of savings percentage consideration on consumer spending (B = -.31, Wald(1) = 4.27, p = .04) and the mediating effect of savings percentage consideration between coupon face value and consumer spending (95% confidence interval = [-.029, -.001]) were significant. These results further support the activation of the savings-comparison mechanism at large coupon face values that underlies the inverted U-shaped effect. In summary, Study 5 identifies preexisting preference for a specific level of product benefit as a boundary condition for the inverted U-shaped effect. This study also provides convergent evidence for the proposed savings-comparison mechanism by demonstrating the mediating role of savings percentage consideration at large coupon face values but not at small coupon face values.

We further conducted two extension studies that assessed the moderating role of consumers’ preexisting benefit-level preference. We elaborate on these next. Three-option choice set. In the first extension study (U.S. online sample, N = 403), we generalized the findings of Study 5 from a two-option choice set (i.e., low price vs. high price) to a three-option choice set (i.e., low price, midprice, or high price; for a description of product stimuli, see Figure WA4 in the Web Appendix). Replicating Study 5, a multinominal logistic regression shows that both (1) the relative choice share of the high-priced hard drive over the midpriced hard drive and (2) the relative choice share of the high-priced hard drive over the lowpriced hard drive had an inverted U-shaped relationship with coupon face value only when consumers’ preexisting preference for a larger hard drive capacity was relatively weaker (for detailed results, see the Web Appendix). Brand liking. In the second extension study (U.S. undergraduate sample, N = 161), we used brand liking to operationalize preexisting preference for a specific level of product benefit. Consumers with a strong brand liking may also have a strong preexisting preference for the product that provides greater benefits and thus is more expensive, regardless of the face value of the product-line coupon they receive, given that a strong brand liking reduces consumers’ price sensitivity (Ailawadi, Lehmann, and Neslin 2003). As a result, for these consumers, the inverted U-shaped effect of coupon face value on consumer spending on this brand should disappear as well. The second extension study using the Godiva stimuli from Study 4 (for a description of product stimuli, see Figure WA3 in the Web Appendix) supports the moderating role of brand liking. Specifically, the inverted U-shaped effect was revealed only when consumers had a relatively weaker liking for Godiva (for a graphical illustration, see Figure 6, Panel B; for detailed results, see the Web Appendix).

General Discussion The present research explores conditions in which the face value of a product-line coupon has an inverted U-shaped effect on consumer spending level. Our findings are robust for both hedonic products (e.g., food and snack) and utilitarian products (e.g., mobile hard drive) and across well-known brands (e.g., Godiva chocolate), less famous brands (e.g., Touro mobile hard drive), and fictional brands (e.g., Korean barbecue buffet). We provide evidence for the savings-comparison mechanism by showing that the inverted U-shaped effect is contingent on the price level of products (Study 1), consumers’ saving orientation (Study 2), information load (Study 3), and consumers’ tendency for thorough comparison (Study 4), and this effect is driven by consumers’ savings percentage consideration (Study 5). We also identify a preexisting preference for a specific level of product benefit as a boundary condition for the inverted U-shaped effect (Study 5). Theoretical Contributions Our research contributes to the literature on price-based sales promotions. Whereas most existing research in marketing has focused on a price discount or coupon that is restricted to a specific product with a fixed price (e.g., Alba et al. 1999; Chandran and Morwitz 2006; Chen, Monroe, and Lou 1998; Chen et al. 2012; Chen and Rao 2007; Lee and Tsai 2014; Leone and Srinivasan 1996; Mishra and Mishra 2011; Nunes and Park 2003; Raghubir 1998; Shiv, Carmon, and Ariely 2005; Thomas and Morwitz 2009), our research examines a popular marketing practice in which a coupon can be used for different products within the same product line of a brand. When a product-line coupon is redeemed, the spending level associated with a specific product choice becomes a very crucial issue—one that, to the best of our knowledge, has not been studied in the literature. Our research makes an important first attempt to examine how coupon face value influences consumer spending level in this context. More importantly, contrary to what the conventional wisdom would predict, we show that an increase in coupon face value does not always lead to an increase in consumer spending level. Instead, for product-line coupons, the effect of face value on consumer spending level could be inverted U-shaped under some circumstances. There are two streams of research on sales promotions that lead to opposite predictions on the effects of coupon face value on consumer spending level. Although the budget-increase perspective suggests a positive effect, the savings-comparison perspective predicts the opposite. Our research reconciles these two opposing predictions by proposing a threshold-based account and by showing that the magnitude of coupon face value could determine the effect of coupon face value on consumer spending level. Managerial Implications It is quite common for retailers to offer deep discounts, especially during major holidays. Prior research has suggested that deep discounts can be detrimental when they dilute brand images (Dodson, Tybout, and Sternthal 1978) or lower consumers’ future price expectations (DelVecchio, Krishnan, and Smith 2007), which may further negatively affect future product

sales in the long run. Our research suggests that the negative effects of deep discounts can take place more immediately. When a deep discount is offered in the format of an amount-off coupon that can be used to buy different products within the same product line of a brand, it may hurt sales because a larger coupon face value may motivate consumers to choose less expensive options under certain circumstances. Fortunately, our findings offer a contingency approach for effectively managing the face value of a product-line coupon to avoid this negative consequence. Drawing on our findings, marketers can determine when they should offer product-line coupons with either a large or a moderate face value, depending on whether the relationship between coupon face value and consumer spending is simply positive or inverted U-shaped. Specifically, marketers can use a large coupon face value to encourage consumer spending when the large coupon face value eventually leads to a high spending level. We identify the positive influence of coupon face value on consumer spending when products are less expensive, when product-line coupons can be applied to a large number of products, or when a firm is able to deliver customized product-line coupons to a target group of consumers who are less inclined to engage in thorough product comparison (e.g., when the firm is able to identify these consumers by tracking their product browsing history). In these cases, marketers could be confident in the power of a large coupon face value to increase consumers’ spending level. In contrast, marketers might consider utilizing a moderate coupon face value when the effect of coupon face value on consumer spending takes an inverted U-shape because a large coupon face value may backfire in terms of inducing consumers to spend less. Our findings suggest that a firm should offer product-line coupons with a moderate face value when the firm’s products are expensive or when product-line coupons can be applied to only a small set of products, because in these cases the relationship between coupon face value and consumer spending is inverted U-shaped. Moreover, we advise marketers to choose a moderate coupon face value if they are able to identify consumers who care more about saving money, who are motivated to conduct thorough product comparison, who do not have a strong preexisting preference for a specific level of product benefit, or who do not have a strong liking for the focal brand. These consumers’ spending amount can be maximized when the face value of a product-line coupon is at a moderate level. Although the present set of studies focuses on the research context in which a coupon can be applied to different products sold at different prices within the same product line of a brand, we expect that the findings derived from this context would still hold when a retail store provides consumers with a coupon that can be used for vertically differentiated products offered by different brands within the same product category. This is because, in the latter context, consumers need to make a similar trade-off between price and benefit for different products, just as participants in our experiments did. Our findings are also applicable to scenarios in which the same coupon can be linked to different combinations of products that are consumed together, such as food and drink items during a restaurant visit. Given that these interrelated items can be regarded as components of an integrated product package, our


theoretical framework can also predict the relationship between consumers’ total spending on these items and coupon face value. Limitations and Future Research Directions The underresearched context of unrestricted product-line coupons provides several interesting avenues for further exploration. First, our studies examine a scenario in which face values of a product-line coupon are below the lower bound of consumers’ potential spending level. What would happen when coupon face values fall between the lower and upper bounds of the retail price range of products to which a coupon can be

applied is a subject for future research. Second, future studies could build more sophisticated models to more accurately specify the range of the turning point of the inverted U-shaped effect. Finally, in the present research the operationalizations of some moderators may have limitations. For example, we operationalized information overload by increasing the number of presented products in Study 3. Yet a large number of products could activate other constructs, such as consumers’ expected choice regret (Iyengar and Lepper 2000). Future studies could address such limitations to provide more evidence for the savings-comparison mechanism that underlies the inverted Ushaped effect.

REFERENCES Ailawadi, Kusum L., Donald R. Lehmann, and Scott A. Neslin (2003), “Revenue Premium as an Outcome Measure of Brand Equity,” Journal of Marketing, 67 (4), 1–17. Aiken, Leona, and S. West (1991), Multiple Regression: Testing and Interpreting Interactions. Newbury Park, CA: Sage Publications. Alba, Joseph W., Carl F. Mela, Terence A. Shimp, and Joel E. Urbany (1999), “The Effect of Discount Frequency and Depth on Consumer Price Judgments,” Journal of Consumer Research, 26 (2), 99–114. Allenby, Greg M., and Peter E. Rossi (1991), “Quality Perceptions and Asymmetric Switching Between Brands,” Marketing Science, 10 (3), 185–204. Anderson, Eric T., and Duncan I. Simester (2004), “Long-Run Effects of Promotion Depth on New Versus Established Customers: Three Field Studies,” Marketing Science, 23 (1), 4–20. Andrews, Michelle, Xueming Luo, Zheng Fang, and Jaakko Aspara (2014), “Cause Marketing Effectiveness and the Moderating Role of Price Discounts,” Journal of Marketing, 78 (6), 120–42. Aydinli, Aylin, Marco Bertini, and Anja Lambrecht (2014), “Price Promotion for Emotional Impact,” Journal of Marketing, 78 (4), 80–96. Biswas, Abhijit, Sandeep Bhowmick, Abhijit Guha, and Dhruv Grewal (2013), “Consumer Evaluations of Sale Prices: Role of the Subtraction Principle,” Journal of Marketing, 77 (4), 49–66. Chandran, Sucharita, and Vicki Morwitz (2006), “The Price of ‘Free’-dom: Consumer Sensitivity to Promotions with Negative Contextual Influences,” Journal of Consumer Research, 33 (3), 384–92. Cheema, Amar, and Vanessa M. Patrick (2008), “Anytime Versus Only: Mind-Sets Moderate the Effect of Expansive Versus Restrictive Frames on Promotion Evaluation,” Journal of Marketing Research, 45 (4), 462–72. Chen, Haipeng (Allan), Howard Marmorstein, Michael Tsiros, and Akshay R. Rao (2012), “When More Is Less: The Impact of Base Value Neglect on Consumer Preferences for Bonus Packs over Price Discounts,” Journal of Marketing, 76 (4), 64–77. Chen, Haipeng (Allan), and Akshay R. Rao (2007), “When Two Plus Two Is Not Equal to Four: Errors in Processing Multiple Percentage Changes,” Journal of Consumer Research, 34 (3), 327–40. Chen, Shih-Fen S., Kent B. Monroe, and Yung-Chein Lou (1998), “The Effects of Framing Price Promotion Messages on Consumers’ Perceptions and Purchase Intentions,” Journal of Retailing, 74 (3), 353–72. Darke, Peter R., and Cindy M.Y. Chung (2005), “Effects of Pricing and Promotion on Consumer Perceptions: It Depends on How You Frame It,” Journal of Retailing, 81 (1), 35–47.


DelVecchio, Devon, H. Shanker Krishnan, and Daniel C. Smith (2007), “Cents or Percent? The Effects of Promotion Framing on Price Expectations and Choice,” Journal of Marketing, 71 (3), 158–70. Del Rio Olivares, M.J., Kristina Wittkowski, Jaakko Aspara, Tomas Falk, and Pekka Mattila (2018), “Relational Price Discounts: Consumers’ Metacognitions and Nonlinear Effects of Initial Discounts on Customer Retention,” Journal of Marketing, 82 (1), 115–31. Dodson, Joe A., Alice M. Tybout, and Brian Sternthal (1978), “Impact of Deals and Deal Retraction on Brand Switching,” Journal of Marketing Research, 15 (1), 72–81. Drèze, Xavier, Patricia Nisol, and Naufel J. Vilcassim (2004), “Do Promotions Increase Store Expenditures? A Descriptive Study of Household Shopping Behavior,” Quantitative Marketing and Economics, 2 (1), 59–92. Fitzsimons, Gavan J. (2008), “Death to Dichotomizing,” Journal of Consumer Research, 35 (1), 5–8. Fong, Nathan M., Zheng Fang, and Xueming Luo (2015), “GeoConquesting: Competitive Locational Targeting of Mobile Promotions,” Journal of Marketing Research, 52 (5), 726–35. Grewal, Dhruv, Ram Krishnan, Julie Baker, and Norm Borin (1998), “The Effect of Store Name, Brand Name and Price Discounts on Consumers’ Evaluations and Purchase Intentions,” Journal of Retailing, 74 (3), 331–52. Grewal, Dhruv, Howard Marmorstein, and Arun Sharma (1996), “Communicating Price Information Through Semantic Cues: The Moderating Effects of Situation and Discount Size,” Journal of Consumer Research, 23 (2), 148–55. Gupta, Sunil (1988), “Impact of Sales Promotions on When, What, and How Much to Buy,” Journal of Marketing Research, 25 (4), 342–55. Gupta, Sunil, and Lee G. Cooper (1992), “The Discounting of Discounts and Promotion Thresholds,” Journal of Consumer Research, 19 (3), 401–11. Hardesty, David M., William O. Bearden, and Jay P. Carlson (2007), “Persuasion Knowledge and Consumer Reactions to Pricing Tactics,” Journal of Retailing, 83 (2), 199–210. Hayes, Andrew F. (2013), Introduction to Mediation, Moderation, and Conditional Process Analysis: A Regression-Based Approach. New York: Guilford Press. Heilman, Carrie M., Kent Nakamoto, and Ambar G. Rao (2002), “Pleasant Surprises: Consumer Response to Unexpected InStore Coupons,” Journal of Marketing Research, 39 (2), 242–52. Iyengar, Sheena S., and Mark R. Lepper (2000), “When Choice Is Demotivating: Can One Desire Too Much of a Good Thing?” Journal of Personality and Social Psychology, 79 (6), 995–1006.

Iyengar, Sheena S., Rachael E. Wells, and Barry Schwartz (2006), “Doing Better but Feeling Worse: Looking for the ‘Best’ Job Undermines Satisfaction,” Psychological Science, 17 (2), 143–50. Kahneman, Daniel (2011), Thinking, Fast and Slow. New York: Farrar, Strauss, Giroux. Karlsson, Niklas, Peter Dellgran, Birgitta Klingander, and Tommy Gärling (2004), “Household Consumption: Influences of Aspiration Level, Social Comparison, and Money Management,” Journal of Economic Psychology, 25 (6), 753–69. Karlsson, Niklas, Tommy Gärling, Peter Dellgran, and Birgitta Klingander (2005), “Social Comparison and Consumer Behavior: When Feeling Richer or Poorer Than Others Is More Important Than Being So,” Journal of Applied Social Psychology, 35 (6), 1206–22. Krishna, Aradhna, and Robert W. Shoemaker (1992), “Estimating the Effects of Higher Coupon Face Values on the Timing of Redemptions, the Mix of Coupon Redeemers, and Purchase Quantity,” Psychology and Marketing, 9 (6), 453–67. Larson, Jeffrey S., and Ryan Hamilton (2012), “When Budgeting Backfires: How Self-Imposed Price Restraints Can Increase Spending,” Journal of Marketing Research, 49 (2), 218–30. Lastovicka, John L., Lance A. Bettencourt, Renee Shaw Hughner, and Ronald J. Kuntze (1999), “Lifestyle of the Tight and Frugal: Theory and Measurement,” Journal of Consumer Research, 26 (1), 85–98. Lee, Leonard, and Claire Tsai (2014), “How Price Promotions Influence Postpurchase Consumption Experience over Time,” Journal of Consumer Research, 40 (5), 943–59. Leone, Robert P., and Srini S. Srinivasan (1996), “Coupon Face Value: Its Impact on Coupon Redemptions, Brand Sales, and Brand Profitability,” Journal of Retailing, 72 (3), 273–89. Levav, Jonathan, Nicholas Reinholtz, and Claire Lin (2012), “The Effect of Ordering Decisions by Choice-Set Size on Consumer Search,” Journal of Consumer Research, 39 (3), 585–99. Lu, Xianghua, Sulin Ba, Lihua Huang, and Yue Feng (2013), “Promotional Marketing or Word-of-Mouth? Evidence from Online Restaurant Reviews,” Information Systems Research, 24 (3), 596–612. Mishra, Arul, and Himanshu Mishra (2011), “The Influence of Price Discount Versus Bonus Pack on the Preference for Virtue and Vice Foods,” Journal of Marketing Research, 48 (1), 196–206. Nenkov, Gergana Y., Maureen Morrin, Andrew Ward, Barry Schwartz, and John Hulland (2008), “A Short Form of the Maximization Scale: Factor Structure, Reliability and Validity Studies,” Judgment and Decision Making, 3 (5), 371–88. Nunes, Joseph C., and C. Whan Park (2003), “Incommensurate Resources: Not Just More of the Same,” Journal of Marketing Research, 40 (1), 26–38. Park, C. Whan, Easwar S. Iyer, and Daniel C. Smith (1989), “The Effects of Situational Factors on In-Store Grocery Shopping Behavior: The Role of Store Environment and Time Available for Shopping,” Journal of Consumer Research, 15 (4), 422–33.

Parker, Jeffrey R., and Rom Y. Schrift (2011), “Rejectable Choice Sets: How Seemingly Irrelevant No-Choice Options Affect Consumer Decision Processes,” Journal of Marketing Research, 48 (5), 840–54. Raghubir, Priya (1998), “Coupon Value: A Signal for Price,” Journal of Marketing Research, 35 (3), 316–24. Ramanathan, Suresh, and Sanjay K. Dhar (2010), “The Effect of Sales Promotions on the Size and Composition of the Shopping Basket: Regulatory Compatibility from Framing and Temporal Restrictions,” Journal of Marketing Research, 47 (3), 542–52. Saini, Ritesh, Raghunath Sing Rao, and Ashwani Monga (2010), “Is That Deal Worth My Time? The Interactive Effect of Relative and Referent Thinking on Willingness to Seek a Bargain,” Journal of Marketing, 74 (1), 34–48. Saini, Ritesh, and Sweta C. Thota (2010), “The Psychological Underpinnings of Relative Thinking in Price Comparisons,” Journal of Consumer Psychology, 20 (2), 185–92. Scheibehenne, Benjamin, Rainer Greifeneder, and Peter M. Todd (2010), “Can There Ever Be Too Many Options? A MetaAnalytic Review of Choice Overload,” Journal of Consumer Research, 37 (5), 409–25. Shiv, Baba, Ziv Carmon, and Dan Ariely (2005), “Placebo Effects of Marketing Actions: Consumers Get What They Pay For,” Journal of Marketing Research, 42 (4), 383–93. Shiv, Baba, and Alexander Fedorikhin (1999), “Heart and Mind in Conflict: The Interplay of Affect and Cognition in Consumer Decision Making,” Journal of Consumer Research, 26 (3), 278–92. Stilley, Karen M., J. Jeffrey Inman, and Kirk L. Wakefield (2010), “Planning to Make Unplanned Purchases? The Role of In-Store Slack in Budget Deviation,” Journal of Consumer Research, 37 (2), 264–78. Thomas, Jacquelyn S., Robert C. Blattberg, and Edward J. Fox (2004), “Recapturing Lost Customers,” Journal of Marketing Research, 41 (1), 31–45. Thomas, Manoj, and Vicki Morwitz (2009), “The Ease-ofComputation Effect: The Interplay of Metacognitive Experiences and Naive Theories in Judgments of Price Difference,” Journal of Marketing Research, 46 (1), 81–91. Townsend, Claudia, and Barbara E. Kahn (2014), “The ‘Visual Preference Heuristic’: The Influence of Visual Versus Verbal Depiction on Assortment Processing, Perceived Variety, and Choice Overload,” Journal of Consumer Research, 40 (5), 993–1015. Tversky, Amos, and Daniel Kahneman (1981), “The Framing of Decisions and the Psychology of Choice,” Science, 211 (4481), 453–58. Van Ittersum, Koert, Joost M.E. Pennings, and Brian Wansink (2010), “Trying Harder and Doing Worse: How Grocery Shoppers Track In-Store Spending,” Journal of Marketing, 74 (2), 90–104.
