The Effect of Odd Price on the Consumers’ Purchase Intentions in the Chinese Market

(1)

The Effect of Odd Price on

the Consumers’ Purchase Intentions

in the Chinese Market

by

Long Chen

University of Groningen

Faculty of Economics and Business

(2)

Abstract

Research into odd pricing indicates a clear effect but strong variance, suggesting that their effects are context dependent. This research explores odd price effects in the Chinese market. The framework is built on price, category, consumer characteristics and cognitive load effects. The empirically supported hypotheses indicate that the framework is well adapted for explaining the effectiveness of odd price usage. The results display that a wide practice of odd pricing in the Chinese market is effective. The findings show that the effect of odd price setting is category dependent. Fair price perception plays an important role in decision-making stage when consumers see an odd price. Consumer segments can be divided based on their characteristics. Lastly, cognitive pressure is proved to affect consumers’ purchase intentions on odd-price-products.

Research theme: Odd price

Key words: Odd pricing; Psychological price strategy; Conjoint analysis

(3)

1. Introduction

The awareness of the complex role of price as a determinant of a purchase intention has always been growing (Monroe, 1973). Among different pricing strategies, it is well known that retailers have a tendency to set prices ending in ‘9’, which is one formulation of odd prices (Mace, 2012). The research on this topic is reaching to a sophisticated level in developed countries. There are plenty of observations that prove how the results of study on odd price should be used in practice. For example, the largest super market retailer in the Netherlands, Albert Heijn (AH) launches a 0.99 (€) price promotion campaign every year. While this paper was being developed, the latest 0.99 (€) price promotion campaign was held since 2nd of April, 2013. Aside this special event rooted on the effect of odd price, more proof could be shown for the practice of using price ending in “9”. For instance, in AH online web shop, among the category of vegetables and fruits, nearly 83.3% of products are labeled with odd prices. Similarly, Chinese retailers and companies are also tended to set odd prices ever since China stepped into the new era of reform and open up in 1979. Lacking of either research or practice, Chinese retailers tend to simply copy the western style even though they do not formulate a clear picture of price strategy yet. One of the reasons is that the empirical research about Chinese consumers in the Chinese Market is rarely conducted.

Bernhard and Winfried (2007) founded that consumers are heterogeneous in their preferences for odd prices. Furthermore, Wedel and Leeflang (1998) argue that consumers are sensitive to round prices up to units of dollar, pound, franc, mark (when euro didn’t exist yet), etc. No such widely accepted results have shown the performance of Chinese consumers on purchase intentions. Therefore, based on the developed theories and methodology, the purpose of this paper is to investigate the effects of odd price within the context of the Chinese market, Chinese consumers and products sold in China under Chinese currency, Yuan. Hence, the main research question would be as follows:

What is the effect of odd price on purchase intentions of Chinese consumers?

(5)

bundles of researchers have proven that the effect of category would be visible (Monroe, 1973; Lichtenstein and Burton, 1989; Anderson and Simester, 2003a; Ngobo et al., 2010; Mace, 2012). In real situation, variances of prices setting exist in different categories in the Chinese market.

Secondly, as well known, consumers hold different fair prices for same products (Friedman, 1968). Helson (1964) and Thaler (1985) use adaptation level theory and transaction utility theory respectively to support the effect of “fair price” on purchase intentions. Chinese consumers’ fair price perceptions are various due to the fact that Chinese retailing market is quite complicated. Therefore, to what extent consumers will treat the odd price as a good price relative to their own fair price is critical in the decision making phase.

Thirdly, Wedel and Leeflang (1998) came up with the idea to divide “price segments” before predicting individual purchase. Later, Bernhard and Winfried (2007) show that heterogeneity does exist in consumers. This is especially the case with China, who has the largest population in the world, namely 1.35 billion people. The heterogeneity is a heat topic in the Chinese retailing market.

Fourthly, a real factor will be introduced into our research given the distinct characteristic of Chinese market and Chinese consumers. Chinese people have very limited daily time for grocery shopping. Single consumers nearly shop everyday after having about 10-12 working hours. Since they are seriously depleted in that situation, the cognitive effort left for involving shopping, such as brand analysis and price comparison is rare (Fennis, Janssen & Vohs, 2009). The exploding of online shopping is a sound ground for this situation. The effect of price digits on such depleted consumers is turning interesting to study on. Brenner and Brenner (1982) observed that processing all of the digits of a price involves considerably more effort than processing only the leftmost digits. Jacoby and Olson (1977) argue that consumers may process all of the digits, and perhaps they often do so when there are no other tasks making demands on the consumers’ limited processing capacities. Then, Hafner and Trampe (2009) present the effect of cognitive load on consumers’ choices. Based on the previous researches, the sub-research questions in this study are as follows:

1. Does the effect of odd price vary across different categories?

(6)

3. Do consumers show characteristic heterogeneity for odd price preference? 4. Do consumers perform different purchase intentions when they have more

cognitive load compared to those having less?

(7)

2. Literature Review

2.1. Odd Price

Monroe (1973) defined “Odd Price” as prices ending with an odd number (e.g., 1, 3, 5, 7, 9), or just under a round number (e.g., 99, 98). This definition has been accepted by other researchers’ studies (Blattberg and Neslin, 1990, p. 349; Wedel and Leeflang, 1998). Based on Friedman’s (1968) work, Wedel and Leeflang (1998) conjectured that the origin of this price setting practice might be as a quantity-discount when fresh products were sold at pound, and the consumers tended to buy half of the standard quantity at a reduced price, namely odd prices.

Among those various formulations, the dominance of 9-ending prices in retailing area is obvious, compared to prices ending in other odd digits (e.g., Anderson & Simester, 2003; Gendall, Fox, & Wilton, 1998; Gedenk & Sattler, 1999; Monroe, 2003; Schindler & Kirby, 1997). In his survey, Friedman (1968) found prices ending in “9” to be the most popular among food retailers: prices ending in 9 capture as much as 80% of retail food prices. Some researchers say that nine-ending pricing is a common, and perhaps overused, marketing technique (Ngobo, Legohérel, and Guéguen 2010; Schindler and Kirby 1997). One of the reasons for this issue happening could be that such nine-ending pricing could increase consumer price sensitivity (Monroe, 1973), as compared to those levels within the surrounding price-interval (Wedel and Leeflang, 1998).

(8)

Table 1

Consumers’ Distortion to Odd Price

Consumers reaction Literature support

Round down prices Bizer and Schindler 2005

Stiving and Winer 1997

Compare prices from left to right

and ignore right-hand digits Monore 1973

Stiving 2000

Stiving and Winer 1997

Only remember left hand digits Schindler and Kibarian 1993

According to this mechanism, an odd-pricing effect occurs because consumers tend to ignore the rightmost nines and fill their places with “default” values (e.g. zeros) or with words which could be used for any rightmost digits and therefore connote numbers lower than nine. Thus, a consumer may think of $5.98 as “around five dollars,” or may consider $395 as “three hundred and some dollars.”

To sum up, bundles of researches have been done to show that there is a positive effect of odd price setting on sales although the number of increasing percentage is not identified among researchers (Blattberg and Wisniewski, 1988; Schindler and Kibarian, 1996; Kalyanam and Schively, 1998; Anderson and Simester, 2003a). The sales increase is treated as the result of the increase of consumer demand. Anderson and Simester (2003a) showed that odd price can increase in demand of approximately 35% while Ngobo et al. (2010) found that odd price setting has a significant, positive impact on the percentage of the buyers. Therefore, the conjecture of a positive effect would be formulated as follows:

H1: The use of odd price can increase consumers’ purchase intentions.

2.2. Category

(9)

and Burton (1989) supported that the price-perceived quality relationship appears to be stronger for nondurable products than for durable products. Anderson and Simester (2003a) showed that the effect of odd price is stronger for new items that customers have seen less frequently in the past. Ngobo et al. (2010) found that odd prices attract more buyers in more concentrated and higher-promoted categories but fewer buyers in more expensive categories. Recently, Macé (2012) came up with the conclusions that a nine-ending price is more effective for increasing sales of small brands (e.g., low market-share, low price, and new items) that belong to weaker categories (e.g., low price, low budget-share). Table 2 summarized the various results of previous research on category aspects.

Table 2

Previous Research Results on The Relationship Between Categories Factors and Odd Price

Author Year Category factors Relationship with

odd price

Monroe 1973 fast moving consumer goods +

Lichtenstein and Burton 1989 nondurable products opposed to durable products

+ Anderson and Simester 2003a new item opposed to old item +

Ngobo et al. 2010 more expensive categories -

Mace 2012 small brands in weaker categories opposed to big brands

+

In summary, it has been clearly investigated that odd price setting has different effects on different product categories. For example, Ngobo et al. (2010) find that nine-ending prices should result in additional buyers in more concentrated categories. Most of the differences will be exposed to fast moving consumer goods and non-fast moving consumer goods, durable goods and non-durable goods, expensive categories and cheap category. Therefore, there is motivation to set conjecture as follows: H2: The effect of odd price on consumers’ purchase intentions varies across different categories.

(10)

2.3. Fair Price

Researchers have shown in their studies how consumers compare psychological price to fair prices (Friedman, 1968). There are mainly two explaining theories: adaptation level theory (Helson, 1964) and transaction utility theory (Thaler, 1985).

According to adaptation level theory, consumers would form perceptions of new stimuli relative to a standard or adaptation level. Therefore, previous and current stimuli to which a person has been exposed would determine the adaptation level (Helson, 1964). Wedel and Leeflang (1998) then argued that odd prices are perceived by consumers as substantially lower than the nearest round prices, despite of the very small difference. Friedman (1968) showed that consumers assume a discount by odd price relative to the fair price. Thus, it is conjectured that when facing an odd price such as €1.99, consumer will treat the nearest integer price €2.00 as a fair price and they see the odd price €1.99 is a discount obtained relative to the fair price (Wedel and Leeflang, 1998). In other words, people would think the cost of the goods is €2.00 (Hanna and Dodge, 1995, p. 28). Hence, researchers conclude that odd pricing would implicitly present a fair price within which incorporated a price cut (Wedel and Leeflang, 1998). Liefeld and Heslop (1985) found that the implicit presentation of a fair price would contribute to a higher price savings than the explicit presentation. On the other hand, Thaler (1985) proposed that the total utility of a transaction to a consumer is the sum of the acquisition utility and the transaction utility. The value of the item will derive the acquisition utility to the consumer. The transaction utility is determined by the difference between the setting price and the fair price. Therefore, Wedel and Leeflang (1998) hypothesized that relative to a perceived fair price of € 2.00, the perceived price cut at a price of €1.99 increases the transaction utility (at € 1.99). The total utility, and thereby the consumer demand will see an increase according to transaction utility theory.

(11)

still be affected by the odd prices the sellers set. To be specific, the possibility of buying would be larger if consumers treat the round price as fair price for products with an odd price. Therefore, hypothesis is as follows:

H3: The use of an odd price as opposed to the nearest round price has a larger positive effect on purchase intentions for those consumers who perceive the round price as a fair price.

2.4. Segments

There is plenty of literature showing that consumers may perform variety on the sensitivity to perceived price reductions (Blattberg et al., 1978; Blattberg and Neslin, 1990, pp. 77-81; Goenuel and Srinivasan, 1993). The effect of psychological prices consequently may differ among consumers. Zeithaml (1988) had reported that the price-perceived quality relationship to be a variable across consumers and situations. Wedel and leeflang (1998) therefore hypothesized that psychological price effects occur in some consumer segments (market segments) but not in others. Then, different “price segments” can be observed (Wedel and Leeflang, 1998).

Demographic information related with segment, including age, education, women, income, and household size are going to be selected because they are key factors for both retail managers and academics. In addition, they have appeared in recent cross-sectional studies of pricing and promotions and in prior literature relating price elasticity to store, category, and product characteristics (Ailawadi et al. 2006; Bell, Chiang, and Padmanabhan 1999; Fok et al. 2006; Hoch et al. 1995; Macé and Neslin 2004; Narasimhan, Neslin, and Sen 1996; Ngobo et al. 2010; Pauwels, Srinivasan, and Franses 2007).

(12)

Table3

Mace’s (2012) Study Results on Demographic Aspects

Clientele characteristic Hypothesis Supported?

Education Nine-ending effect is greater in stores whose area clientele comprises less educated people

Yes Gender Nine-ending effect is greater in stores whose area

clientele comprises more working women

Yes Income Nine-ending effect is greater in stores whose area

clientele comprises more high-income households

Yes Age Nine-ending effect is greater in stores whose area

clientele comprises more older people

n.s. Household size Nine-ending effect is greater in stores whose area

clientele comprises more large families

n.s. Notes: n.s. = not significant

The variety of the sensitivity to price conception is indicated clearly during past years (Blattberg et al., 1978; Blattberg and Neslin, 1990, pp. 77-81; Goenuel, and Srinivasan, 1993). Wedel and Leeflang (1998) incorporated segments as independent variables into their model to investigate the effect of odd price. Afterwards, Mace (2012) presented the results of the investigation for separately clientele characteristics. Bernhard and Winfried (2007) addressed the existence of level effects and investigated the influence of consumer characteristics on preferences for odd prices. Based on Mace’s (2012) job, the conjecture of the difference concerning to segment covariates is going to be set in this study.

H4: Consumers are heterogeneous among segments.

H4a: Consumers are heterogeneous in gender among segments. H4b: Consumers are heterogeneous in age among segments. H4c: Consumers are heterogeneous in education among segments. H4d: Consumers are heterogeneous in household size among segments. H4e: Consumers are heterogeneous in gender among segments.

2.5. Cognitive Load

(13)

More effort could be considerably involved when consumers round an odd price upwards or process all of the digits of a price than processing only the leftmost digits (Brenner and Brenner, 1982). However, based on Jacoby and Olson’s (1977) job, when there are no other tasks making demands on the consumer's limited processing capacities, consumers are intended to round odd prices upward or process all of the digits. Thus, Schindler and Warren (1988) summarized that in situations where there are other stimuli competing for their limited cognitive resources, consumers may fail to take the extra time or make the extra effort to refocus their attention from something else to processing these less important digits, the rightmost nines for odd prices. Instead, they may end up simply ignoring these digits.

In the laboratory studies, it is argued that the respondents usually applied a great deal of attention to each price. In most of the experiments (Dodds & Monroe 1985; Georgoff 1972; Schindler 1984; Schindler & Wiman 1986), the respondents were presented with the prices one at a time in a situation where they were not pressed for time. Thus, those respondents certainly had the time to give their full attention to each price. The respondents in the Lambert (1975) and Alpert et al. (1984) studies were required to deal with four prices within one minute, and thus may have been a bit pressed for time. However, these experiments involved the respondents in a game, and this may have increased their alertness and motivation and, as a result, increased their likelihood of fully attending to each price.

In a study by Hafner and Trampe’s (2009), differences have been shown between respondents who were asked to remember a nine-digit number through their experiment and who were not asked to do that to burden more cognitive load. They argued that consumers who had more cognitive load would respond more impulsively. In the meantime, it is further argued that consumers are lazy so that they just ignore the right-most digits when they are involved in a heavier cognitive load situation. Therefore, just memorizing the left-most digit could increase the purchase intentions given it reflects a lower integer part of the whole price.

(14)

2.6. Conceptual Model

The following conceptual model graphically displays the relationships found in the literature. The focus of this study is on the effect of the odd price on consumers’ purchase intentions. Nevertheless, other relationships will be tested within this model.

Figure 1 Conceptual Model Category Cognitive load Odd price

(15)

3. Empirical Study

3.1. Objectives

This paper is based on the previous research methodology. Following Gendall et al. (1998), a choice-based conjoint study with brand-price stimuli, including both odd and even prices is going to be presented. The main objective is to analyze whether consumers are heterogeneous in their preferences for odd prices, and if so, how they differ in their preference structures. Considering our research scope given real price setting situation in the Chinese market, one of Monroe’s (1973) definition about odd price is identified in our research as the price just below round price, namely 0.99 Yuan and 4999 Yuan in our study.

3.2. Conjoint Design and Data Collection

(16)

product categories. Three well-known brands with comparable quality were set in each category. In category of mineral water, there are two national brands Wahaha and Nongfushanquan, and one international brand Nestle. The relative largest brand is Nongfushanquan. In category of notebook, there are two national brands Lenevo and Isus and one international brand Dell. As Bernhard and Winfried (2007) argue in their work, image effects can be attributed to price levels independent of product quality.

Table 4

Brand Names and Price Levels Used in The Conjoint Study

Attribute Levels Product category

Mineral water Notebook

Brand 1 Nongfushanquan Dell

Brand 2 Nestle Asus

Brand 3 Wahaha Lenevo

Low “anchor” price 0.90 Yuan 4900 Yuan

Price “5” 0.95 Yuan 4950 Yuan

Price “9” 0.99 Yuan 4999 Yuan

Price “0” 1.00 Yuan 5000 Yuan

High “anchor” price 1.10 Yuan 5100 Yuan

(17)

Table 5

Summary of Respondents’ Characteristic Distribution

# of respondents # of respondents All 434 Gender % Household % Female 250 57.6% 1 (single) 107 24.7% Male 184 42.4% 2 96 22.1% Age 3 115 26.5% 16-18 2 0.5% 4 7 1.6% 19-25 227 52.3% 5 14 3.2% 26-30 177 40.8% >5 95 21.9% 31-35 15 3.5% Income >35 13 3.0% <2000 181 41.7% Education 2000-2999 40 9.2% Bachelor 254 58.5% 3000-3999 55 12.7% Master 132 30.4% 4000-4999 58 13.4% PhD 9 2.1% 5000-5999 25 5.8% others 39 9.0% 6000-6999 17 3.9% Cognitive load 7000-7999 15 3.5% 0 215 49.5% >8000 43 9.9% 1 157 36.2% 2 62 14.3%

0: without cognitive load when make choice

1: with cognitive load and succeed recalling the right digits 2: with cognitive load but failed recalling the right digits

The data was collected through an-online research software. The choice sets were presented consecutively on a laptop so that the respondents could not compare the different choice sets with each other. Respondents were informed at the outset that the presented notebook (mineral water) alternatives were identical in other attributes (e.g., processor-type, internal memory capacity, hard disk and screen size for notebooks; package size and mineral formulation for mineral water), and that the three brands were the only ones available in each category.

To reveal the effects of fair price, the sample was to be divided into two groups. People taking round price (1.00 Yuan and 5000 Yuan) as fair price are in the first group. On the other hand, people having other price levels as fair prices are in the second group.

(18)

had to remember the number but failed recalling the right number. This setting as mentioned before is following the arguments by Hafner and Trampe (2009).

3.3. Statistical Model

Following the work by Bernhard & Winfried (2007), the Multinomial Logit Model (MNL) will be used in this study to estimate respondents' part-worth utilities (McFadden, 1974). Accordingly, the conditional choice probability 𝑃!"# that respondent i (i = 1,..., I) will choose alternative/stimulus j ( j = 1,..., Jc) from choice set c (c = 1,..., C) is:

𝑃

_!"#

=

𝑒𝑥𝑝 (𝑉

!"

)

exp(𝑉

_!"!

)

!!_∈!

The deterministic utility, 𝑉!", of respondent i for alternative j is an additive part-worth function of the brand names and price levels:

𝑉!" = 𝛽!!,!𝑥!!,!+ 𝛽!!,!𝑥!!,!+ 𝛽!!",!𝑥!!",!+ 𝛽!!!,!𝑥!!!,!+ 𝛽!!!,!𝑥!!!,!+ 𝛽!!,!𝑥!!,!

(19)

4. Results

4.1. Effect of Odd Price

The main effect of odd price setting for each category is determined by latent class model with one class, namely the aggregate model. The importance results (Table 6) show that in water and computer category, the attribute price has relatively more important effect than attribute brand in this research. Therefore, we can assume respondents made choices mainly based on price levels instead of brands even though the attribute brand also shows a significant effect (Table 7).

Table 6a Table 6b

Attributes Importance – Water Attributes Importance - Computer

Figure 2 and Table 7 show the part-worth utility of price for category water and computer respectively. The p-value of price is 1.20E – 122 for water and 3.70E-83 for computer, which states that the attribute price is significantly affecting respondents’ choice in both categories. The results can further display that the odd price has largest effect for both water and computer category.

Fig 2a Table 7a

Price Utility – Water Overall Model - Water

Importance Maximum brand 0.573 price 1.129 Relative brand 33.5% price 66.5% Importance Maximum brand 0.592 price 0.985 Relative brand 37.6% price 62.4%

Attributes Utility p-value

(20)

Fig 2b Table 7b

Price Utility – Computer Overall Model - Computer

In water category, the part-worth utility of odd price (0.99) is higher than that of high anchor price (1.10) by 0.988. In computer category, the part-worth utility of odd price (4999) is also reaching higher than that of high anchor price (5100) by 1.129. The increases of utilities lead to the increases of purchase intentions accordingly. Therefore, H1 is supported: the use of odd price will increase consumers’ purchase intentions.

T-test is employed to test for the equality of the coefficients of the effect of odd price for water and computer. Based on our results, following calculation is formulated:

𝑡 = 𝑏!"#$%_!"#$− 𝑏!"#$%&'(_!"#$ stdev(𝑏_{!"#$%_!"#$}− 𝑏_{!"#$%&'(_!"#$}) =

0.3463

0.096711 = 3.58

Given our sample is 434 which has large degree of freedom, t-value of 3.58 is significant at α = 0.01. Therefore, the two parameters significantly differ from each other and H2 is supported: The effect of odd price on purchase intentions varies across different categories.

4.2. Fair Price

The effect of fair price will be analyzed by the combination of both latent gold class model and T-test for effect equal test. The results are going to be presented on category separately.

(21)

-‐0.6 -‐0.4 -‐0.2 0 0.2 0.4 0.6 0.8 0.9 0.95 0.99 1 1.1 -‐0.6 -‐0.4 -‐0.2 0 0.2 0.4 0.6 0.8 0.9 0.95 0.99 1 1.1 4.2.1. Water

Fig 3 and Table 8 show that attribute price is significantly (p-value = 1.30 E-07 for group A and 4.70E-71 for group B) contributing to respondents’ choices decision for water products. The drops of part-worth utilities from odd price to round price however are different for these two groups.

Price Utility – Group A Overall Model – Group A

Price Utility – Group B Overall Model – Group B

Notes: Group A – fair price = 1.00

Group B – fair price = 0.90, 0.95, 0.99 or 1.10 Yuan

The drop of the utility for group A is 0.649 while the drop of the utility for group B is 0.622. T- test is conducted as follows to test if these drops are equal.

Brand 1.30E-07 Nongfushanquan 0.386 Nestle -0.081 Wahaha -0.305 Price 4.70E-71 0.90 -0.136 0.95 -0.497 0.99 0.730 1.00 0.081 1.10 -0.178

(22)

-‐0.6 -‐0.4 -‐0.2 0 0.2 0.4 0.6 0.8 4900 4950 4999 5000 5100 -‐0.8 -‐0.6 -‐0.4 -‐0.2 0 0.2 0.4 0.6 0.8 4900 4950 4999 5000 5100 𝑡 = 𝑏!"#$_!− 𝑏!"#$_! stdev(𝑏_{!"#$_!}− 𝑏_{!"#$_!}) = 0.0267 0.0055= 4.85

Given our large degree of freedom, t-value of 4.85 is significant at α = 0.01. Therefore, we can conclude that the effects of odd price are various on respondents who hold different fair prices for water products. Furthermore, the positive effect of odd price is larger on those consumers who perceive the round price a fair price.

4.2.2. Computer

Fig 4 and Table 9 show that attribute price is significantly (p-value = 3.70 E-83 for group A and 4.70E-71 for group B) contributing to respondents’ choices decision for computer products.

Price Utility – Group A Overall Model – Group A

Price Utility – Group B Overall Model – Group B

Notes: Group A – fair price = 5000

Group B – fair price = 4900, 4950, 4999 or 5100 Yuan

Brand 6.50E-26 Dell 0.365 Asus -0.557 Lenevo 0.191 Price 3.70E-83 4900 0.113 4950 0.126 4999 0.560 5000 -0.389 5100 -0.409

(23)

The drop of the utility for group A is 0.949 while the drop of the utility for group B is 1.114. T- test is conducted as follows to test if these drops are equal.

𝑡 =_stdev(𝑏𝑏!"#$_!− 𝑏!"#$_!

!"#$_!− 𝑏!"#$_!) =

0.165

0.168= 0.984

Thus, t-value of 0.984 given our degree of freedom is not significant. Therefore, we can conclude that the effects of odd price are not various on respondents who hold different fair prices for computer products. The positive effect of odd price is not distinct on those consumers who perceive the round price a fair price.

To conclude for overall hypothesis, H3 is partially supported: The use of an odd price as opposed to the nearest round price, has a larger positive effect on purchase intentions for those consumers who perceive the round price as a fair price. However, this larger effect is category dependent.

4.3 Heterogeneity Among Segments

Using latent gold class model, models from one-class to 6-class were specified to compare. The following section shows models and segmentation for each category.

4.3.1. Water

(24)

Table 10

Model Summary Outputs – Water

BIC(LL) AIC(LL) AIC3(LL) CAIC(LL) p-value Class.

Err. R²

Prediction. Err.

1-Class Choice 6906 6881 6887 6912 9,5e-1100 0 0.090 47.3%

2-Class Choice 5893 5734 5773 5932 1,5e-873 0.008 0.321 33.2%

3-Class Choice 5317 5023 5095 5389 5,5e-736 0.020 0.470 23.6%

4-Class Choice 5184 4756 4861 5289 1,5e-687 0.026 0.546 19.0%

5-Class Choice 5162 4600 4738 5300 1,3e-661 0.042 0.568 19.1%

6-Class Choice 5174 4478 4649 5345 1,3e-642 0.045 0.600 18.9%

In table 11, Wald shows attributes brand and price significantly contribute to explaining respondents’ choices for water products and Wald(=) means significant difference among segments on each attribute. Segment 1 contains 40.1% of the sample and displays (Fig 5a) the significant preference for odd price (0.99). Segment 2 contains 29.7% of respondents and shows (Fig 5b) the significant preference for round price (1.00). Segment 3 and segment 4 include 16.3% and 13.9% of the sample respectively. Both of them have significant preference for low anchor price (0.90). The difference between segment 3 and 4 lies in the brand preference.

Table 11

Model for 4-class Choice – Water

Class1 Class2 Class3 Class4

Class size 40.1% 29.7% 16.3% 13.9%

Attributes Wald Wald(=)

(25)

Fig 5a: Price Effect for Segment 1 Fig 5b: Price Effect for Segment 2

Fig 5c: Price Effect for Segment 3 Fig 5d: Price Effect for Segment 4

Based on Table 12, the results can be observed that Gender (p = 0.00091) and Income (p = 0.05) display significant difference among four segments. Specifically speaking, segment 1 (58.8%), segment 2 (59.9%) and segment 3 (72.8%) have more female than male respondents. The most respondents belonging to segment 4 are male (68.6%). With income, 46.4% of respondents in segment 1 and 36.5% of respondents in segment 2 have lower income than 2000 Yuan per month. 35.1% of respondents in segment 3 have below 2000 Yuan per month income, however, 19.8% of this segment capture the highest level of income in this research, namely more than 8000 Yuan per month. Segments 4 spreads the biggest 3 parts to income lower than 2000 Yuan per month (47.1%), 4000-4999 Yuan per month (16.2%) and income more than 8000 Yuan per month (10.1%). Hence, H4a1 and H4e1 are supported that: consumers are heterogeneous in gender and income for water category. But, there is no statistical evidence to support H4b1(p = 0.65), H4c1(p = 0.34) and H4d1(p = 0.68): consumers are not heterogeneous in age, education, or household size for water category.

(26)

Table 12

Profile for 4-class Model – Water

Covariates Class1 Class2 Class3 Class4 Wald p-value

Gender 16.48 0.00091 Female 0.588 0.599 0.728 0.314 Male 0.412 0.401 0.273 0.686 Age 9.64 0.65 16-18 0.012 0 0 0 19-25 0.566 0.507 0.463 0.502 26-30 0.382 0.399 0.481 0.414 31-35 0.029 0.047 0.028 0.033 >35 0.012 0.047 0.027 0.051 Education 10.07 0.34 Bachelor 0.604 0.619 0.433 0.636 Master 0.289 0.287 0.437 0.227 PhD 0.019 0 0.044 0.046 others 0.088 0.093 0.088 0.092 Household 12.06 0.68 1 (single) 0.255 0.230 0.271 0.228 2 0.207 0.250 0.225 0.197 3 0.214 0.339 0.263 0.256 4 0.188 0.126 0.169 0.169 5 0 0.054 0.072 0.032 >5 0.137 0 0 0.118 Income 32.66 0.05 <2000 0.464 0.365 0.351 0.471 2000-2999 0.123 0.100 0.017 0.073 3000-3999 0.119 0.177 0.100 0.069 4000-4999 0.114 0.169 0.092 0.162 5000-5999 0.048 0.047 0.073 0.092 6000-6999 0.011 0.048 0.112 0.015 7000-7999 0.034 0.031 0.057 0.017 >8000 0.086 0.062 0.198 0.101 Fair price 26.65 0.0088 0.90 -1.013 -0.626 0.618 1.021 0.95 -0.671 -1.021 0.936 0.756 0.99 -0.229 -0.424 0.786 -0.132 1.00 2.286 -0.091 0.302 0.163 1.10 -0.373 2.162 -2.641 -1.607

(27)

4.3.2. Computer

As table 13 shows, based on the argumentation of Bodzogan (1987), model 4 with 4-class is going to be used due to its CAIC(LL) is the lowest with 4855. The R2 increases from 0.06 of model1 to 0.62 of model4 while the prediction error decreases from 48.7% of 1-class model to 16.8% of 4-class model.

Table 13

Model Summary Outputs – Computer

BIC(LL) AIC(LL) AIC3(LL) CAIC(LL) p-value Class. Err. R² Prediction. _Err.

1-Class Choice 7047 7023 7029 7053 4,6e-1137 0 0.064 48.7%

2-Class Choice 5827 5668 5707 5866 1,4e-868 0.018 0.334 31.6%

3-Class Choice 5125 4832 4904 5197 5,3e-706 0.018 0.504 21.2%

4-Class Choice 4750 4323 4428 4855 2,8e-609 0.023 0.616 16.8%

5-Class Choice 4747 4185 4323 4885 1,5e-586 0.023 0.635 14.8%

6-Class Choice 4797 4101 4272 4968 6,8e-575 0.028 0.659 15.0%

In table 14, Wald shows attributes brand and price significantly contribute to explaining respondents’ choices for computer products and Wald(=) means significant difference among segments on each attribute. Segment 1 contains 35.6% of the sample and displays (Fig 6a) the significant preference for odd price (4999). Segment 2 and segment 4 include 25.7% and 17.9% of the sample respectively. Both of them have significant preference (Fig 6b and 6d) for low anchor price (4900). Segment 3 counts 20.8% and most respondents significantly prefer price of 4950 (Fig 6c).

Table 14

Model for 4-class Choice – Computer

Class1 Class2 Class3 Class4

Class size 35.6% 25.7% 20.8% 17.9%

Attributes Wald Wald(=)

(28)

Fig 6a: Price Effect for Segment 1 Fig 6b: Price Effect for Segment 2

Fig 6c: Price Effect for Segment 3 Fig 6d: Price Effect for Segment 4

Based on Table 15, the results can be observed that Household (p = 0.018) display significant difference among four segments. Segment 1 is the largest segment with respondents from each household level. However, most of respondents with household size larger than 5 are divided into segment 1. Segment 2 contains most of the respondents with household size 3. Segment 3 and segment 4 display similarity on this covariate.

Table 15

Profile for 4-class Model – Computer

Covariates Class1 Class2 Class3 Class4 Wald p-value

(29)

Education 2.49 0.98 Bachelor 0.601 0.543 0.592 0.607 Master 0.280 0.352 0.298 0.291 PhD 0.021 0.009 0.035 0.021 others 0.098 0.095 0.075 0.081 Household 28.60 0.018 1(single) 0.242 0.244 0.239 0.269 2 0.215 0.198 0.239 0.249 3 0.211 0.331 0.292 0.250 4 0.173 0.103 0.194 0.198 5 0.012 0.081 0.024 0.014 >5 0.147 0.043 0.011 0.019 Income 18.43 0.62 <2000 0.446 0.465 0.320 0.396 2000-2999 0.133 0.064 0.057 0.089 3000-3999 0.123 0.114 0.156 0.119 4000-4999 0.114 0.144 0.149 0.144 5000-5999 0.041 0.071 0.046 0.088 6000-6999 0.017 0.054 0.071 0.028 7000-7999 0.025 0.036 0.057 0.028 8000-8999 0.101 0.054 0.144 0.109 Fair price 12.03 0.44 4900 -0.994 0.235 0.386 0.374 4950 -0.871 -0.028 0.259 0.639 4999 -1.054 0.663 0.256 0.134 5000 -1.153 0.485 0.691 -0.023 5100 4.071 -1.355 -1.593 -1.123

To conclude generally, H4 is partially supported. The summary for sub hypotheses will be shown at the end of this results section.

4.4 Cognitive Load Effect

(30)

4.4.1. Water

Our research shows that price significantly contributes to all respondents’ choice. However, the effect of odd price differs from different cognitive load groups for water product. First of all, t-test is used to test the equality of the effect between each two groups. T-value between group 0 and group 1 is 17.86 (p = 0.00), t-value between group 0 and group 2 is 12.52 (p = 0.00) and t-value between group 1 and group 2 is 0.76 (p = 0.45). This result means the effect of odd price significantly differs between consumers without cognitive load and consumers with cognitive load. However, the effect has no significant difference between consumers who recalled the number right or not.

Then, results (Fig 7 and Table 16) show that the part-worth utilities of odd price 0.99 is much larger on respondents who have cognitive load job than those who do not have. Within groups, respondents without cognitive load are affected by price of odd price 0.99 (U = 0.03) much less than other price levels. The utility of price 0.90 reaches to 0.28 for group 0. In the meantime, the utility of odd price 0.99 (U = 1.29 and U = 1.22) is outstandingly higher than other price levels when consumers have cognitive load.

Price Utility – Group 0 Overall Model – Group 0

(31)

Fig 7c Table 16c

4.4.2 Computer

Respondents are divided as the same way with water product. The effect of odd price is tested for each group as figure 8 and table 17 show. The results display again that for all three groups, price is a significant variable. T-test is used to test the equality of the effect between each two groups. T-value between group 0 and group 1 is 6.16 (p = 0.00), t-value between group 0 and group 2 is 8.10 (p = 0.00) and t-value between group 1 and group 2 is 1.08 (p = 0.28). As same as the results for water products, the

Brand 0.41 Nongfushanquan 0.066 Nestle 0.017 Wahaha -0.083 Price 1.40E-131 0.90 0.155 0.95 -0.425 0.99 1.297 1.00 -0.291 1.10 -0.735

(32)

effect of odd price on purchase intentions significantly differs between consumers without cognitive load and consumers with cognitive load. However, the effect has no significant difference between consumers who recalled the number right or not. Furthermore, the effect of odd price (4999) is significantly higher (U = 1.218 and U = 0.992) than other price levels on consumers’ purchase intention when they have cognitive load. The utility of odd price (U = 0.054) is in the relative middle place of five price levels for group that has no cognitive load.

Brand 3.30E-13 Dell 0.194 Asus -0.347 Lenevo 0.126 Price 0.021 4900 0.153 4950 0.115 4999 0.054 5000 0.006 5100 -0.328

(33)

Fig 8c Table 17c

To conclude for the effect of cognitive load, H5 is completely supported: the effect of odd price results to larger consumers’ purchase intentions when consumers have more cognitive pressure.

Table 18 displays the summary for whether each hypothesis is supported or not.

Table 18

Summary of Hypothesis Testing Results

Hypothesis Supported?

H1: The use of odd price will increase consumers’ purchase intentions. Yes

H2: The effect of odd price on consumers’ purchase intentions varies across different

categories.

Yes H3: The use of an odd price as opposed to the nearest round price has a larger positive

effect on purchase intentions for those consumers who perceive the round price as a fair price.

Partially Yes

H4a: Consumers are heterogeneous in gender among segments. Partially Yes

H4b: Consumers are heterogeneous in age among segments. No

H4c: Consumers are heterogeneous in education among segments. No

H4d: Consumers are heterogeneous in household size among segments. Partially Yes

H4e: Consumers are heterogeneous in income among segments. Partially Yes

H5: The effect of odd price results to larger consumers’ purchase intentions when

consumers have more cognitive pressure.

Yes

(34)

5. Conclusions and Implications

In this study, the effect of odd price on consumer’s purchase intention is investigated in the Chinese market. By employing choice based conjoint analysis, the influence of odd price on consumer’s choice in water and computer category is clearly shown. Particular attention is going to be given to our research questions as table 19 summarizes.

Table 19 Research Questions

Research Questions

1. What is the effect of odd price on purchase intentions of Chinese consumers? 2. Does the effect of odd price vary across different categories?

3. Do consumers show various purchase intentions when they have different fair price perceptions? 4. Do consumers show characteristic heterogeneity for odd price preference?

5. Do consumers perform different purchase intentions when they have more cognitive load?

(35)

categories, for research question 1. Based on this ground, it is obvious to suggest the Chinese retailers to set odd prices in certain product areas, namely the water and computer categories.

Second, as many researchers identified, nine-ending pricing indicates a clear effect on sales but strong variance, suggesting that their effects are context dependent (Mace, 2012; Baumgartner & Steiner, 2007; Anderson & Simester, 2003). The various contexts are defined in our study as different categories. Many researchers argue that the effect of odd price is different according to the categories (Baumgartner & Steiner, 2007; Anderson & Simester, 2003a; Lichtenstein & Burton, 1989; Monroe, 1973). Our study is supportive to these authors, showing that odd pricing would yield different influences on different categories. The later part of results about fair price, heterogeneity and cognitive load effect confirmed to our conclusion that odd price has category dependent effect on consumers. The positive answer for research question 2 is shedding light on the fact that category is an important aspect when Chinese retailers consider setting odd prices for specific products. We reason that due to the properties of the products and the retail context for different categories, the effect of odd prices can be really differentiated from each category. Especially for products that consumers usually buy a multiple amount, for instance the water product in our study. Chinese consumers are paying large attention if they receive a precise change fund.

(36)

consumers who hold a nearest round price as a fair price do not show larger purchase intentions when they see an odd price based on the different values given to their needs. This result could be further generalized to other durable goods.

Fourth, research question 4 would elaborate on the diversity of clientele characteristics. As Georgoff (1972) found, we confirm that in the Chinese market, men are less responsive to odd pricing than women for water products. We argue that Chinese women are mainly responsible to do grocery shopping. Therefore, Chinese women focus more on family budget using. Odd price stands for a good bargain for them. After that, we confirmed the income heterogeneity among segments for water category as Mace (2012) found. Consumers always favor high quality goods if they can afford it. Respondents with lower income are trying harder than those with higher income to find the balance point between qualities pursue and budget limitation. The reason we argue that most respondents with lowest income in our research favor the odd price for water products is that an odd price setting is appealing to consumers especially with relative lower income. These people will highlight that they can get a discount by odd prices (Friedman, 1968). Furthermore, we found that consumers present heterogeneity in household size for computer category. Higher purchase intentions by odd prices are clearly observed in families that have more than 5 people. Consumers, who have high budget-constraints, including low- income households and large families, are more price conscious (Hoch et al. 1995). Lastly, age and education have no effect on the consumers’ purchase intention either for water or for computer products. Mace (2012) found that age does not affect people’s purchase intention due to the context dependence. We believe education to be biased at this point because the most respondents have an education level that is equal to a university level and not lower. This is one of our limitations for this research.

(37)

and read the whole price of all the items (Mace, 2012). Chinese consumer are likely to be attracted to any sign that can help them speed up their shopping visit.

Managerial Implication

China, as the second largest retailing market in the world, has kept drawing attentions of international brands and retailers. To use effective price strategy is critical to make success in the Chinese market. This research provides evidence of the Chinese context dependence of odd price effects and clearly demonstrates that Chinese manufacturers and retailers could practice odd pricing strategy. Odd price is effective to increase Chinese consumer’s purchase intentions. This strategy could be used in different categories or brands. When consumers make purchase decision, price is relative more important than brands. However, retailers should be cautious to copy odd price setting across categories because the effects are not the same for different categories.

Odd price can easily attract consumer’s attention because consumer will perceive the nearest round price as the fair price for certain goods with the odd price. Brand managers should further explore their targeting group. Female consumers and low-income group would more prefer odd price than other groups for non-durable goods. In the meantime, household size does affect the durable goods purchase intention of Chinese. Collecting customer data would help retailer to target effectively on accordingly consumers. We recommend brand managers could cooperate with Chinese supermarkets to gather sales and clientele data.

Lastly, brand managers could distribute advertisement in supermarkets when people are shopping. The distractions by advertisement for consumers will accordingly produce remembering load on consumers. Therefore, the effect of odd price on consumer’s purchase intention will be strengthened by the effectiveness of advertisement.

Limitations

(38)

(39)

Reference

Anderson, E. T., and Simester, D. I. (2003), “Effects of $9 price endings on retail sales: Evidence form field experiments,” Quantitative Marketing and Economics, 1(1), 93−110. Ailawadi, Kusum L., Bari A. Harlam, Jacques César and David Trounce (2006), “Promotion profitability for a retailer: The role of promotion, brand, category, and store characteristics,”

Journal of Marketing Research, 43 (November), 518–35.

Bernhard, Baumgartner., and Winfried, J. Steiner (2007), “Are consumers heterogeneous in their preferences for odd and even prices? Findings from a choice-based conjoint study,”

Journal of Research in Marketing, 24, 312–323.

Bell, David R., Jeongwen Chiang, and Venkata Padmanabhan (1999), “The decomposition of promotional response: An empirical generalization,” Marketing Science, 18 (4), 504–26. Bizer, G. Y., and Schindler, R. M. (2005), “Direct evidence of ending-digit drop-off in price information processing,” Psychology and Marketing, 22(10), 771-783.

Blattberg, R.C., Buesing, T., Peacock, P., and Sen, S.K. (1978), “Identifying the deal prone segment,” Journal of Marketing Research, 25, 369-377.

Blattberg, R.C., and Neslin, S.A. (1990), “Sales Promotion, Concepts, Methods and Strategies,” Prentice-Hall, Englewood Cliffs, NJ.

Bozdogan, H. (1987), “Model selection and Akaike’s information criterion (AIC): The general theory and its analytical extensions,” Psychometrika, 52, 345-370.

Brenner, G.A., and Brenner, R. (1982), “Memory and markets, or why are you paying $ 2.99 for a widget?” Journal of Business, 55 (1), 147-158.

Erickson, G.M., and Johansson, J.K. (1985), “The role of price in multi-attribute product evaluations,” Journal of Consumer Research, 12, 195-199.

Fennis, B.M., Janssen, L., and Vohs, K.D. (2009), “Acts of benevolence: A limited-resource account of compliance with charitable requests,” Journal of Consumer Research, 35, 906-924. Fok, Denis, Csilla Horvath, Richar Paap and Philipp Hans Franses (2006), “A Hierarchical Bayes Error Correction Model to Explain Dynamic Effects of Price Changes,” Journal of

Marketing Research, 43 (3), 443–61.

Friedman, L. (1968), “Psychological pricing in the food industry. In: Phillips, A., Williamson, O.E. (Eds.), Prices, Issues in Theory, Practice and Public Policy,” University of Pennsylvania Press, Philadelphia.

Gabor, A., and Granger, C.W.J. (1961), “On price consciousness of consumers,” Applied

Statistics, 10, 170-188.

Gabor, A., and Granger, C.W.J. (1964), “Price Sensitivity of the Consumer,” Journal of

Advertising Research, 4 (December), 40-4.

(40)

Gedenk, K., and Sattler, H. (1999), “The impact of price thresholds on profit contribution — Should retailers set 9-ending prices?” Journal of Retailing, 75(1), 33-57.

Gendall, P., Fox, M. F., and Wilton, P. (1998), “Estimating the effect of odd pricing,” Journal

of Product and Brand Management, 7(5), 421-432.

GoÈnuÈl, F., and Srinivasan, K. (1993), “Modelling multiple sources of heterogeneity in multinomial logit models: Methodological and managerial issues,” Marketing Science, 12, 213-229.

Hanna, N., and Dodge, H.R. (1995), “Pricing, Policies and Procedures,” MacMillan Business, London.

Helson, H. (1964), “Adaptation Level Theory,” Harper & Row, New York.

Hoch, Stephen J., Byung-Do Kim, Alan L. Montgomery and Peter E. Rossi (1995), “Determinants of Store-Level Price Elasticity,” Journal of Marketing Research, 32 (1), 17–29. Jackson, T., and Marks, N. (1999), “Consumption, sustainable welfare and human needs— with reference to UK expenditure patterns between 1954 and 1994,” Ecological Economics (28), pp. 421–441.

Lichtenstein, D.R., and Burton, S. (1989), “The relationship between perceived and objective price quality,” Journal of Marketing Research, 26, 429-443.

Liefeld, J., and Heslop, L.A. (1985), “Reference prices and deception in newspaper advertising,” Journal of Consumer Research, 11, 868-876.

Mace, Sandrine. (2012), “The Impact and Determinants of Nine-Ending Pricing in Grocery Retailing,” Journal of Retailing, 88 (1, 2012), 115-130.

Macé, Sandrine and Scott A. Neslin (2004), “The Determinants of Pre- and Postpromotion Dips in Sales of Frequently Purchased Goods,” Journal of Marketing Research, 41 (3), 339– 50.

Monroe, K.B. (1973), “Buyers' subjective perceptions of price,” Journal of Marketing

Research, 10, 70-80.

Monroe, K. (2003), “Pricing: Making profitable decisions,” Burr Ridge, IL: McGraw-Hill/ Irwin.

M. Wedel, and Leefang, P.S.H. (1998), “A model for the effects of psychological pricing in Gabor Granger price studies,” Journal of Economic Psychology, 19, 237-260.

Narasimhan, Chakravarti, Scott A Neslin and Subrata K. Sen (1996), “Promotion Elasticities and Category Characteristics,” Journal of Marketing, 60 (2),17–30.

Ngobo, Paul-Valentin, Patrick Legohérel and Nicolas Guéguen (2010), “A Cross-Category Investigation into The Effects of Nine-Ending Pricing on brand Choice,” Journal of Retailing

and Consumer Services, 17 (5), 374–85.

(41)

Schindler, R. M., and Kibarian, T. M. (1993), “Testing for perceptual underestimation of 9-ending prices,” Advances in Consumer Research, 20, 580-585.

Schindler, R. M., and Kirby, P. N. (1997), “Patterns of rightmost digits used in advertised prices: Implications for nine-ending effects,” Journal of Consumer Research, 24(2), 192-201. Schindler, Robert M. and Alan R. Wiman (1989), “Effects of Odd Pricing on Price Recall,”

Journal of Business Research, 19 (3), 165–77.

Schindler, R. M., and Warren, L. S. (1988), “Effect of odd pricing on choice of items from a menu,” Advances in Consumer Research, 15, 348-353

Steenkamp, J.B.E.M. (1989), “Product Quality,” Van Gorcum, Assen.

Stiving, M. (2000), “Price-endings when prices signal quality,” Management Science, 46(12), 1617-1629.

Stiving, M., and Winer, R. S. (1997), “An empirical analysis of price endings with scanner data,” Journal of Consumer Research, 24(1), 57-67.

Thaler, R. (1985), “Mental accounting and consumer choice,” Marketing Science, 4, 199-214. Wedel. M., and Leefang. P. S. H. (1998), “A model for the effects of psychological pricing in Gabor-Granger price studies,” Journal of Economic Psychology, 19 (1998) 237-260.

Zeithaml, V.A. (1988), “Consumer perceptions of price, quality and value: A means end model and synthesis of evidence,” Journal of Marketing, 52 (July), 2-22.

The Effect of Odd Price on the Consumers’ Purchase Intentions in the Chinese Market