Smart Promotions: The Effect of Sales Prices, Advertised Reference Prices and Charm-Pricing on Consumers’ Internal

Smart Promotions: The Effect of Sales Prices, Advertised Reference Prices and Charm-Pricing on Consumers’ Internal

Reference Prices, Perceived Value and Behavioural Intention

University of Groningen Faculty of Economics and Business

Department of Marketing 2017

Vito David Eichwald

Coronastraat 38 9742EG Groningen +49 (0) 176 63157921 vito_eichwald@icloud.com

Student number:

S2983389 1

^st

supervisor:

Prof. Dr. L.M. Sloot 2

^nd

supervisor:

Dr. W. Jager

2

Acknowledgements

During the last few years I have attended four universities in four different countries. It is no exaggeration to state that I learned more at the Rijksuniversiteit Groningen than at all other Universities combined. The learning curve was very steep, the amount of knowledge immense and sometimes overwhelming but it was always a lot of fun. I can say that I learned a lot for my life and my future career. The present thesis is the final part of my ‘education-journey’. I would like to take the opportunity to thank some important people that made it possible. First and foremost, I would like to thank my supervisor Prof. Dr. Laurens Sloot. His enthusiasm about marketing and retail is contagious and spreads to the people around him. This makes it very enjoyable to work with him. His mentoring was the perfect mix between structured guidance and freedom. Second, I would like to thank my survey participants for the time and effort invested. Third, I am grateful for the helpful comments of my fellow thesis students. Last but not least, I would like to thank my second supervisor Dr. Wander Jager for his final evaluation. I don’t know him personally but I know that his guest lecture on social complexity was very inspiring.

Vito Eichwald, Groningen, June 2017

3

Abstract

The present study is built upon previous research in the domain of price-perceived value models within the context of temporary price reductions in a grocery retail setting. The study explains the effect of external reference prices (sales price and advertised reference prices) and charm pricing (i.e.

sales prices ending with the digit 9) on consumers’ internal reference prices, perceptions of value and the subsequent willingness-to-buy. An experimental study using a 2x2x2 design is used to test the conceptual model and has been analyzed in a structural equation modelling framework (SEM). The results show that both external reference prices (sales price and advertised reference price) influence internal reference prices and value perceptions. Furthermore, the analysis shows that charm prices have an influence on consumers’ perceived transaction value, which is connected to the joy and pleasure derived from a deal. Perceived transaction value influences perceived acquisition value, which in turn influences willingness-to-buy. Subsequently, the author discusses the implications for management and academia and suggests future research directions.

Keywords: Perceived Value, Charm Prices, Reference Prices, Structural Equation Model, Retail

Marketing, Pricing

4

T ABLE OF C ONTENTS

1. INTRODUCTION ... 6

1.1

R

EFERENCE

P

RICES

... 7

1.2

C

HARM

P

RICES

... 7

1.3

O

BJECTIVES OF THE

S

TUDY AND

R

ESEARCH

Q

UESTION

... 7

1.4

A

CADEMIC

C

ONTRIBUTION

... 8

1.5

M

ANAGERIAL

I

MPLICATIONS

... 9

1.6

S

TRUCTURE OF THE

P

APER

... 9

2. LITERATURE REVIEW ... 10

2.1

R

EFERENCE

P

RICES

... 10

B

EHAVIOURAL AND

M

ODELLING

A

PPROACHES OF

R

EFERENCE

P

RICE

- R

ESEARCH

... 10

2.1.1 I

NTERNAL AND

E

XTERNAL

R

EFERENCE

P

RICES

... 10

2.1.2 A

DVERTISED

R

EFERENCE

P

RICE

... 11

2.1.3 R

EFERENCE

P

RICES AND

A

NCHORING

... 11

2.1.4 P

SYCHOLOGICAL

A

PPROACHES OF

R

EFERENCE

P

RICE

... 12

2.1.5 G

ENERALIZATIONS

F

ROM

R

EFERENCE

P

RICE

R

ESEARCH

... 13

2.1.6 O

PERATIONALIZATIONS OF

R

EFERENCE

P

RICE

... 13

2.2

C

HARM AND

R

OUND

P

RICING

... 14

2.2.1 L

EVEL

E

FFECTS

... 15

2.2.2 I

MAGE

E

FFECTS

... 16

2.2.3 C

HARM

P

RICING IN

D

UTCH

R

ETAIL

... 17

3. CONCEPTUAL FRAMEWORK AND DEVELOPMENT OF HYPOTHESES ... 21

3.

1

P

ERCEIVED

Q

UALITY

... 22

3.2

I

NTERNAL

R

EFERENCE

P

RICE

... 22

3.3

P

ERCEIVED

V

ALUE

... 24

3.4

W

ILLINGNESS

-

TO

-B

UY

... 25

4. METHODOLOGY ... 27

4.1

R

ESEARCH

P

LAN

... 27

4.1.1 P

RODUCT

C

HOICE AND

C

ATEGORY

C

HOICE

... 27

4.1.2 P

RICE

M

ANIPULATIONS

... 28

4.1.3 M

EASUREMENT AND

S

CALING

... 29

4.2

C

ONTROL

M

EASURES

... 30

4.2.1 B

RAND

E

QUITY

... 30

4.2.2 I

N

-S

TORE

D

EAL

P

RONENESS

... 30

4.2.3 C

OUNTRY OF RESIDENCE

... 30

4.2.4 S

TORE

C

HOICE

... 31

4.2.5 B

RAND

C

HOICE

... 31

4.3

P

OPULATION AND

S

AMPLING

... 31

4.4

P

LAN OF

A

NALYSIS

... 32

5. RESULTS ... 33

5.1

S

AMPLE

D

ESCRIPTION AND

D

EMOGRAPHICS

... 33

5.2

D

ATA

C

HECKS

... 34

5.2.1 M

ISSING

D

ATA AND

O

UTLIERS

... 34

5.2.2 S

KEWNESS AND

K

URTOSIS

... 34

5.3

I

NITIAL

R

ELIABILITY AND

V

ALIDITY

M

EASUREMENT

... 34

5.3.1 E

XPLORATORY

F

ACTOR

A

NALYSIS

... 34

5

5.3.2 I

NITIAL

C

ONFIRMATORY

F

ACTOR

A

NALYSIS AND

C

RONBACH

’

S

A

LPHA

... 36

5.4

C

ONTROL

V

ARIABLES

... 37

5.5

S

TRUCTURAL

E

QUATION

M

ODELL

... 38

5.5.1 C

ONFIRMATORY

F

ACTOR

A

NALYSIS

... 38

5.5.2 MULTIVARIATE DATA CHECKS ... 41

5.5.3 P

ATH

M

ODEL

F

IT

... 41

5.5.4 T

EST OF HYPOTHESES

... 43

5.5.5 M

ULTI

-G

ROUP

E

FFECTS

... 44

6. DISCUSSION AND IMPLICATIONS ... 46

6.1

I

NTERNAL

R

EFERENCE

P

RICES

... 46

6.2

C

HARM

P

RICING

... 46

6.3

P

RICE

-P

ERCEIVED

-V

ALUE

M

ODEL

... 47

7. LIMITATIONS AND IMPLICATIONS FOR FUTURE RESEARCH ... 48

8. CONCLUSION ... 50

9. REFERENCES ... 51

10. APPENDICES ... 57

6

1. I NTRODUCTION

In recent years, retailers and brand manufacturers increasingly rely on price promotions

¹

to attract customers and raise sales (Srinivasan, Pauwels, Hanssens & Dekimpe, 2002). This shift is caused by several developments in the fast-moving-consumer-goods (FMCG) and retail sector. First, there is a reduced commitment to classic advertising on behalf of brand manufacturers (Buzzell, Quelch &

Salmon, 1990). Second, the success story of hard discounters (e.g. Aldi and Lidl) puts pressure on traditional service retailers. Hard discounters sell a limited assortment, with a fairly high quality, at rock-bottom prices (Steenkamp & Kumar, 2009). Third, during a recession, like the recent recession caused by the American housing bubble and the Euro-crisis, consumers typically become more value oriented and especially temporary price promotions are playing a significant role in retail pricing strategy (Quelch, 2008).

The most frequently used price promotion technique is a temporary price reduction (TPR) (Gedenk, Neslin & Ailawadi, 2006). TPR are usually supported by non-price promotions. For example, comparative price advertising, where the (reduced) sales price is shown with a (usually higher) advertised reference price (ARP). This can be done in-store (e.g. promotional displays) and/or out-of- store (e.g. leaflets). Possible ARP are the price suggested by the manufacturer or the former (i.e. non- reduced) sales price.

As price promotions become more common, it is a top-priority for managers of retailers and brand manufacturers to design smart price promotions to increase promotion effectiveness. However, setting prices right is a very sensitive topic in retail strategy. In general, price is not just an objective measure of value but also has a subjective internal representation in consumers’ minds (Doods, Monroe &

Grewal, 1991) This makes pricing a powerful tool. However, it can be risky if used inappropriately. It can be used to build trust and convince customers to buy products. Conversely, it can damage brand equity, negatively affect customers’ trust and even lead to lawsuits (Anderson & Simester, 2003a).

This is making a retailer’s pricing strategy a powerful means to influence consumers’ perceptions and behaviors.

The present study focuses on two domains in psychological pricing research and investigates how they can be applied in a TPR context. First, it builds upon research on reference prices. Second, it incorporates research on charm pricing. In the following, the concepts of reference prices and charm pricing are briefly presented. Consequently, the main research question and the managerial and theoretical implications of the research are discussed.

1 The current study investigates the effect of certain price manipulations in a price-promotion setting. However, price- promotions are not the prior subject of the study. For a review of the literature on price promotions it is referred to Gedenk, Neslin and Ailawadi (2006).

7

1.1 R EFERENCE P RICES

When consumers shop at a retailer they use reference prices. Reference prices are consumers’

standards that they use to judge purchase prices of products (Monroe, 1973). They can be internal and external. The internal reference price (IRP) is the consumer’s internal, psychological price construct. It is typically a range of plausible prices for a respective product or brand in the mind of the consumers.

External reference prices (ERP) usually refer to extrinsic cues from the store-environment (e.g. the actual price or prices of competing brands) (Mazumdar &Papatla, 2000). Examples are the advertised selling price or advertised reference prices (ARP) (e.g. competitive comparative prices, was-now comparisons). However, there are various conceptualizations of IRP and ERP and the line between the two concepts is not always clear. Chapter 2.1 will provide a more detailed review of the literature on reference prices.

1.2 C HARM P RICES

Prices which end on a 9 (e.g. € 89, € 19.99) seem ubiquitous in modern day retailing. Studies report that in the United States 30% to 65% of retail prices end on a 9 (Anderson & Simester, 2003b;

Schindler & Kirby, 1997). Charm pricing is also commonplace in Dutch retail. However, more recently several Dutch retailers started to use more rounded prices. (see chapter 2.2.3). One anecdotal theory says that charm prices emerged due to the advent of the cash register. Charging one cent less than a full dollar was a way to force the cashier to open the cash register as a tactic to prevent fraud (Arango, 2009; Liang & Kanetkar, 2006).

However, this is hardly the reason why many prices still end on a 9. There is the belief among retailers that prices just below a round number (e.g. € 1.99 or € 199) can influence consumers’ price perceptions. For example, it is assumed that consumers perceive the utility of paying € 1.99 instead of

€ 2 as higher than the perceived utility of a discount from € 2.17 to € 2.16. Even if the discount is equal (i.e. 1 cent) it is assumed that the magnitude in consumers’ price evaluations is higher when it is just below a round number. Intuitively, one might think that the tactic has no effect because it is so common and consumers are likely to ignore it. However, literature finds some support for some psychological effects of charm prices. For example, a field study by Anderson & Simester (2003b) found that raising the price for a dress in a catalogue from $ 34 to $ 39 has been able increase consumer demand.

1.3 O BJECTIVES OF THE S ^{TUDY AND} R ^ESEARCH Q ^UESTION

The present research investigates the effects of several price manipulations in a TPR setting. The study

builds on a price-perceived value model by Grewal, Monroe and Krishnan (1998), from research on

8

price discounts in a durable consumer goods context, and applies it in a grocery retail setting. A special focus lies on the formation of IRP based on ERP (i.e. ARP and sales prices). Furthermore, the research at hands also incorporates possible effects of charm pricing on consumers’ perceptions of value. Hence, the main research question is:

RQ: How can retailers and brand manufacturers utilize psychological price techniques (i.e. reference prices and charm pricing) to influence consumers’ value perceptions of a TPR and ultimately generate sales?

Following sub-research questions are derived from the main research question:

1) How do ERP (i.e. ARP and the sales price) influence consumers’ IRP?

2) Does charm pricing influence consumers’ IRP?

3) How do external reference prices (i.e. ARP and the sales price) influence consumers’

perceptions of value?

4) How do consumers’ IRP influence their perceptions of value?

5) Does charm-pricing have an influence on consumers’ perceptions of value?

6) How do consumers’ value perception translate into behavioral intention (i.e. willingness-to- buy)?

1.4 A CADEMIC C ONTRIBUTION

From an academic perspective, the present paper makes several contributions. Firstly, it applies a model that has been established in the domain of durable consumer goods in the FMCG domain. This may help to establish some generalizability of the model across product categories and industries.

Secondly, the study employs several measures from previous research on other product categories,

which have been adjusted to the product and category at hand. Hence, this research demonstrates that

the scales developed in previous research can be applied in other domains as well. Second, the study

contributes to the research domain on reference prices. There is no clear agreement among scholars

whether internal reference prices are best explained solely by past observed prices or if it is more

appropriate to consider price cues in the current shopping environment as well. Hence, the piece of

research adds to the stream of literature that acknowledges the importance of price cues from the

current store environment Thirdly, the study adds to the literature on charm pricing. The study

investigates how charm prices may have an impact on consumers’ IRP. Next, to the knowledge of the

author this is the first study that puts charm pricing in a price-perceived value context. Hence, the

study demonstrates how charm prices affect the consumers’ perceptions of value.

9

1.5 M ANAGERIAL I MPLICATIONS

The study was specifically designed with practitioners in mind. Hence, it is not only investigated how APL, ARP and charm pricing affect IRP and perceived value but it also explains how the perceived value ultimately affects willingness-to-buy (WTB) (i.e. leads to business outcomes). It can give managers from retailers and brand manufacturers valuable advice on how to design effective price promotions. First, as mentioned above and discussed in more detail in chapter 2.2.3, several Dutch retailers recently employ more rounded prices. However, the effects of switching from charm to more rounded prices are not fully clear. Hence, the study demonstrates how switching from charm to round prices affects consumers’ perceptions of value. Next, the study provides retail managers with an understanding of how to make use of ARP to increase consumer perceptions of value. Furthermore, the study can help management understand how IRP influence perceived value in a retail setting.

1.6 S TRUCTURE OF THE P APER

The research paper is organized as followed; After this introduction and a subsequent literature review the conceptual framework is presented along with the hypotheses tested. Then, the research methods and the study design will be discussed. After that the findings will be presented. Next, these findings will be discussed and a conclusion with implications for academics and managers will be given.

Finally, implications for future research will be discussed.

10

2. L ITERATURE R EVIEW

In the following chapter, the author briefly reviews existing research on the topics of reference prices and charm pricing.

2.1 R ^EFERENCE P ^RICES

Reference prices are consumers’ standards that they use to judge purchase prices of products (Monroe, 1973). It is about their price expectations and not necessarily about their willingness-to-pay.

Mazumdar, Raj and Sinha (2005: 85) conceptualize reference prices as “a predictive price expectation that is shaped by consumers’ prior experience and current purchase environment”. Hence, a reference price is not only dependent on prior experience but can be actively influenced by extrinsic cues (e.g.

brands and prices) in the current purchase environment. Moreover, a reference price is a range rather than a specific price. For example, a consumer might be willing to pay € 2-3 for a tub of margarine and not only precisely € 2.73 (Lichtenstein & Bearden, 1989). Importantly, the reference price construct is latent. Thus, it is not directly observable and consumers are usually not aware that they use it.

B

EHAVIOURAL AND

M

ODELLING

A

PPROACHES OF

R

EFERENCE

P

RICE

- R

ESEARCH

There are generally two streams in pricing literature that aim at making sense of the reference price concept. First, the modelling approach gets its insights from testing and comparing the statistical fit of various reference price conceptualizations based on consumer panel data or scanner data. Second, the behavioral approach aims at measuring the impact of external stimuli on consumers’ internal reference prices (Mazumdar et al., 2005). Hence, the first focuses on a measurement on an aggregate level while the second focuses on a measurement on an individual level. An advantage of a modelling approach clearly is its realism. Consumers shop in a natural environment (e.g. the supermarket) while the behavioral stream bases its observations mainly on artificial environments, like labs. Nevertheless, an advantage is the increased ability to better control for specific extrinsic cues and its impact on consumer behavior (Aronson, Wilson & Brewer, 1998). The present study aims at uncovering the impact of specific extrinsic cues. Thus, the present study follows a behavioral approach in that sense.

2.1.1 I

NTERNAL AND

E

XTERNAL

R

EFERENCE

P

RICES

One can categorize the reference price concept into internal and external reference prices. The IRP is the consumer’s internal, psychological price construct while ERP usually refer to extrinsic cues from the store-environment (e.g. the actual price, prices of competing brands).

However, Mazumdar and Papatla (2000) distinct ERP and IRP differently. According to their

definition, IRP assume that consumers reference price is based purely on past purchase occasions and

people enter the store with a clear idea of an appropriate price in mind. Contrary, their definition of

11

ERP assumes that reference prices are formed solely during the shopping occasion based on observed prices in the store (Mazumdar &Papatla, 2000). Hence, the ERP varies on the store level while the IRP varies on the individual level (Kumar, Hurley, Karande, & Reinartz, 1999).

Some studies concluded that IRP can better explain consumer behavior (e.g. Briesch, Krishnamurthi, Mazumdar & Raj, 1997) while others found a better fit for ERP price-models (e.g Hardie, Johnson &

Fader, 1993). However, the distinction is not as clear as it seems and both views are not mutually exclusive. Extrinsic cues (i.e. ERP) at the store level can get internalized and therefore become part of consumers’ IRP (Mazumdar et al., 2005). The definition of Grewal et al. (1998: 47) recognizes that and defines IRP more broadly as “a price (or price scale) in buyers’ memories that serves as a basis for jugging or comparing actual prices”.

ERP can be various price cues available to a consumer. Thus, the external price cue does not have to be the purchase price for the respective product and does not even have to be from the same product category. For example, it is possible that consumers subconsciously integrate prices from other product categories. Consequently, they may incidentally use price information to form their IRP (Mazumdar et al., 2005; Bolton, 1989).

2.1.2 A

DVERTISED

R

EFERENCE

P

RICE

A special form of ERP is the advertised reference price (ARP), which is a reference price provided by the retailer (e.g. a higher suggested retail price, competitive prices or dealer invoice prices). Research suggests that ARP can affect buyers’ perceptions of value and purchase intention. Hence, IRP are influenced by perceived product quality, the selling price and an ARP (Grewal et al., 1998). However, its effect is largely dependent on the plausibility of the ARP. Thus, very high (and implausible) ARP have a weaker impact than moderately discrepant ones. Furthermore, retailers can use semantics to increase the effect of ARP (e.g. “was at”, “compare to”) (Mazumdar et al., 2005). In a meta-analysis of comparative price advertising Compeau and Grewal (1998) assessed the effectiveness of ARP and found large support for the effectiveness of comparative price advertising (i.e. offering external ARP) as a powerful advertising tool.

2.1.3 R

EFERENCE

P

RICES AND

A

NCHORING

When people make decisions under uncertainty, they are prone to using available information sources as anchors to adjust their estimates. Thus, if people are asked to judge a stimulus and they do not have a specific value in mind they use an anchor and adjust upwards or downwards until they arrive at a plausible value. (Tversky & Kahneman, 1974; Einhorn & Hogarth, 1986).

Anchoring describes the process during which consumers retrieve an initial value from their short-

term memory which then functions as the anchor value (Wansink, Kent & Hoch, 1998). Anchors can

12

be internal or external. In a pricing context, internal anchors refer to knowledge structures of prices observed in the past.

External anchors are information present in the decision-making environment. Thus, this view is consistent with the concept of ERP. Prices of competitive products, prices of products from other categories and, of course, ARP can serve as external anchors (Chandrashekaran & Grewal, 2006).

Johnson and Cui (2013) found providing ERP (e.g. a minimum, maximum or suggested price) in a pay-what-you-want setting to have been able to bias consumers chosen price towards the external anchor. Kopalle and Lindsey-Mullikin (2003) found that providing consumers with an ERP in form of a higher ARP as an anchor affects consumers price expectations. They propose that the effect of an ARP on price expectations is quadratic (u-shaped). Hence, if the difference between the ARP and subjects IRP (i.e. their initial price expectations) increases, consumers tend to update their IRP (i.e.

increase price expatiations) up to a certain point when they start to decrease.

Chandrashekaran and Grewal (2006) investigated the effect ARP and advertised sales prices on consumers’ IRP. They found that an ARP changes the way consumers evaluate a sales price. Hence, when a sales price exceeds consumers’ IRP the advertised reference price, as an external anchor, puts upward pressure on IRP. Contrary, when it falls below the IRP it has a negative impact. Furthermore, Adaval and Wyer (2011) propose that priming consumers with an external price anchor can have an impact on consumers’ willingness to pay for a certain product. They found that the effect can go across product categories.

In summary, the availability of an external anchor (i.e. an ERP in various forms) can have an impact on consumers’ IRP.

2.1.4 P

SYCHOLOGICAL

A

PPROACHES OF

R

EFERENCE

P

RICE

Past studies used classic psychological theories to make sense of the reference price concept. First, adaption-level theory (Helson, 1964) proposes that people judge stimuli by comparing it to an internal norm. The internal norm is the level they have adapted over the course of time. All judgements are made relatively to the current adaption level. In a pricing context, this suggests that people compare current price stimuli with past stimuli (Helson, 1964; Crompton, 2011). Hence, a consumer’s adaption level is “a function of the frequencies of different values for that category i.e. the distribution of values […] and the magnitude of the series of stimuli, the range of stimuli, and the dispersion of stimuli from the mean” (Kalyanaram & Winer, 1995:163).

Second, psychological assimilation-contrast theory explains how consumers integrate external

information into their reference price (Mazumdar et al., 2005; Sherif & Hovland, 1961). Hence,

usually, a consumer has a distribution of prices that is acceptable. Consumers assimilate price

13

information if the external price belongs to this distribution (Mazumdar et al., 2005). Consumers have

“a latitude of acceptance around their price beliefs that influence their reaction to an external reference price” (Biswas, Krishnan & Burton, 1999: 54). Therefore, consumers may assimilate an external reference price (e.g. actual price information and/or advertised reference prices) into their internal reference price or contrast it against their internal reference price range (Biswas et al., 1999).

Kalyanaram and Little (1994) carried out research in the domain of consumer packaged goods and found that consumers tend to assimilate prices if they fall within .75 times the price variability of the product. Hence, findings based on adaption-level theory and assimilation-contrast theory suggest that consumers have a zone of tolerance. Thus, “for a given service and quality level, people have a range of prices that are considered acceptable” (Crompton, 2011: 208). Consumers with a higher average internal reference price have a wider zone of tolerance and consumers that shop more frequently have a narrower range because they have more knowledge about the actual price distributions. Furthermore, reference prices can be either stored in specific numbers or in ranks (Mazumdar et al., 2005). For example, a consumer may think blue band margarine costs about € 2 or she/he can think that it is cheaper than Bertolli but more expensive than the store brand.

2.1.5 G

ENERALIZATIONS

F

ROM

R

EFERENCE

P

RICE

R

ESEARCH

Previous research on the antecedents of IRP led to some generalizable conclusions (Mazumdar et al., 2005; Kalyanaram & Winer, 1995). First, the strongest determinant of an IRP is the prior price the consumer has observed (Mazumdar et al., 2005). Mazumdar et al. (2005) elaborate that the last two to three shopping occasions have the highest impact on IRP. Furthermore, Rajendran and Tellis (1994) compared different models of reference price formation. They concluded that a model that assumes that IRP are formed out of contextual prices (i.e. observed directly at the point-of-sale) and temporal prices (i.e. prices observed during recent shopping trips) can better explain consumers IRP than models that are only built on either temporal prices and contextual prices. Furthermore, previous promotions also shape the formation of reference prices. For example, if a certain item is often on sale this influences consumers’ promotion expectations, where promotion expectation is the extent to which consumers are conditioned to promotions for a certain brand or product at hand. Consequently, if consumers have higher promotion expectations this has a negative effect on the internal reference price (Kalwani, Rinne, & Sugita, 1990; Kalyanaram & Winer, 1995). In summary, prior prices are the strongest influencer of internal reference prices and more recent prices have a higher impact than less recent prices. Furthermore, prices observed at the point-of-sale are also considered by consumers.

2.1.6 O

PERATIONALIZATIONS OF

R

EFERENCE

P

RICE

Operationalizations of IRP are diverging between various studies. Researchers agree that it is an

internal, psychological price construct but the operationalizations differ to large degrees. Thaler (1985)

conceptualizes it as the consumer’s expected or just price. Hence, the fairness perspective plays a

14

major role in his operationalization. Others operationalized it as consumers’ expected average market prices (Urbany, Dickson & Wilkie,1989). Kalyanaram and Little (1994) operationalize it as a weighted average of recent purchase prices and Stoetzel (1970) as the lowest price that it is acceptable without doubting the quality of the respective product. Others ask for the the expected future price (Jacobson

& Obermiller, 1990). The variety of definitions and the resulting operationalizations lead to an increase in integrated, multidimensional measures of reference price. For example, studies used expected high, average and low prices together to measure IRP (Biswas & Blair, 1991).

The present study uses the operationalization formerly employed by Grewal et al. (1998). The construct is grounded in prior research by Lichtenstein and Bearden (1989) and Urbany, Bearden and Weilbaker (1988). The measures are based on consumers estimates of an average market price and a fair price for the respective product.

2.2 C ^{HARM AND} R ^OUND P ^RICING

However, not just the price level may influence consumers’ judgements and behaviors. The way price cues are presented may have effects on price judgements as well. There are generally two main forms of prices in retail: round prices and unrounded prices. In this study, rounded prices are defined as prices that end on the rightmost digit 0 (e.g. € 2.00 or 3.50) while unrounded prices are defined as prices that end on a rightmost digit different from 0 (e.g. € 2.42 or € 7.99). Furthermore, unrounded prices can be distinguished into charm prices and precise prices. In the present study, charm prices are defined as ‘prices that end on the rightmost digit 9 (e.g. € 1.99, € 79.99 or € 199). Thomas, Simon and Kadiyali (2010) describe precise prices as a price that has fewer ending zeroes compared to a round price (e.g. € 364,578). However, their research looked at prices that are untypically high in grocery retail. In a grocery retail setting, typical precise prices are prices like € 2.57 or € 68.73. However, in the case of precise and charm prices a clear distinction is hard to make. One can see the definitions charm and precise prices as the ends of a continuum. Some studies, like the present study, only consider prices that end on a 9 (e.g. Gedenk & Sattler, 1999; Schindler & Kibarian, 2001; Anderson &

Simester, 2003a; Anderson & Simester, 2003b; Liang & Kanetkar, 2006; Manning & Sprott, 2009) while others include prices that end with other digits that are just below a round number (i.e. end on 95, 96, 97 or 98) (e.g. Lambert, 1975; Stiving & Winer, 1997; Choi, Li, Rangan, Chatterjee & Singh, 2014).

A study investigated charm pricing using scanner panel data of a major supermarket. It found that

more than 80 percent of the prices ended on the digit 9. Furthermore, they showed that charm pricing

can indeed lead to a sales advantage. For example, a price reduction of a margarine brand from $ .89

to $ .71 lead to a price elasticity of 3.2 but a reduction of from $ .98 to $ .69 lead to an elasticity of

9.9. Thus, charm pricing may lead to ‘greater-than-expected’ elasticities (Blattberg & Wisniewski,

15

1983). Further, a field study Anderson & Simester (2003b) by found that raising the price for a dress in a catalogue from $ 34 to $ 39 has been able increase consumer demand. Thus, there exists some support on an aggregate level.

The explanatory approaches for the impact of charm pricing have been largely tested on an individual level to uncover the responsible psychological mechanisms. The effects can be broadly categorized into level effects and image effects. First, level effects say that consumers systematically perceive charm prices (e.g. € 2.99) to be lower than the rounded neighboring price (e.g. € 3) (Stiving & Winer, 1997). Consequently, marketers expect higher price-elasticities from charm-prices in comparison with the surrounding price interval. For example, there is a higher expected sales increase from a price cut from € 1 to 99 cents that from decreasing 99 to 98 cents or € 1.01 to 1 (Wedel & Leeflang, 1998).

Second, image effects influence the image perception of a firm. For example, consumers may connect charm prices with a specific firm or product attribute (i.e. quality image or price image) (Stiving &

Winer, 1997).

2.2.1 L

EVEL

E

FFECTS

There are various explanatory approaches on level effects. First, the rounding down-effect hypothesizes that consumers have the tendency to round prices down (i.e. drop-off the rightmost digits). For example, consumers may perceive € 1.99 as € 1 and some change and not as about € 2 (Stiving & Winer, 1997). However, Stiving & Winer (1997) criticize that there is a lack of evidence for the rounding down effect. Indeed, early studies had very mixed and often contradictory results (e.g.

Lambert, 1975; Alpert, McGrath & Alpert, 1984). However, recent research finds some support for the rounding-down effect. Bizer & Schindler (2005) provide direct evidence for a rounding down-effect of charm pricing. They found that respondents estimated their spending power to be significantly higher when receiving charm prices as stimuli compared to a group that received the rounded counterparts.

An analysis of the errors respondents made additionally supported the drop-off hypothesis.

Second, left-to-right comparison assumes that consumers directly compare prices with another price.

However, they tend to compare prices from the left digit to the right digit and pay no or less attention to the right digit(s). Stiving and Winer (1997) illustrate the mechanism with the following example:

When people see following pairs of prices ($ .89, $. 75) and ($ .93, $ .79) they tend to perceive that $ .79 is the better deal although the price difference is identical ($ .14). Furthermore, the comparison can either happen on-line or off-line. An on-line comparison occurs when a consumer compares a price cue with another price cue in-store (e.g. another price in the shelf). For example, if a consumer compares the prices of two items in store s/he will pay no attention to the rightmost digits if the leftmost digits are not identical (Coutler, 2001) Finally, an off-line comparison takes place when a consumer compares a price with a reference price held in memory (Stiving & Winer, 1997).

Consumers learn over time that the more valuable parts of a price are usually the leftmost digits and

16

thus use a left-right encoding strategy (Coutler, 2001). Coutler (2001) found support for the left-to- right comparison in a lab-setting. The study found that left digit recall was higher than right digit recall when price-digits were exposed simultaneously but not when exposed from right to left.

Furthermore, the demand for brands with a charm-price was higher than the demand for brands with rounded prices when respondents had been exposed to the digits simultaneously or from left to right but not when exposed from right to left. Stiving and Winer (1997) investigated the effects of charm prices using scanner panel data. They built three logit models estimating consumers’ indirect utility function. The first model was built under the assumption that consumers process the price holistically (i.e. compare prices based on all digits). The second model assumed that consumers weight rightmost digits differently. Finally, the third model assumed that consumers selectively ignore rightmost digits.

The third model was built to model the left-to-right processing effect. The last model was the best fitting model, which implies that consumers indeed ignore the rightmost digit. The authors affirm that this behavior is certainly not irrational. Consumers trade of the likelihood of making a mistake against the mental costs of processing the rightmost digits. Thus, they may not make the best decision in every case but good decisions with the least amount of mental costs (Stiving & Winer, 1997)

Third, another explanation frequently used is explained by the limited memory capacity of humans (Brenner & Brenner, 1982; Stiving & Winer, 1997; Coutler, 2001; Wedel & Leeflang, 1998).

Consumers are continuously bombarded with information but have a limited memory capacity for storing directly accessible information. The left digits (e.g. Euros or Dollars) have a higher financial impact than the rightmost digits (e.g. cents). Thus, they tend to focus on only storing the most important part of a price. For example, if consumers see a price of € 9,99 they may only remember € 9 (Coutler, 2001; Stiving & Winer, 1997). Rounding up would mean an additional decision from the consumer compared to just storing the rightmost part of the price (Wedel & Leeflang, 1998). There is some support for this theory on possible level effects. For example, research found that consumers have a poorer memory for charm prices than for rounded prices (Schindler, 1984). In another study participants have been exposed to a set of prices and were asked to recall those prices two days later. It was found that round prices were more likely to recall than charm prices. Furthermore, charm pricing increased the likelihood that it will be underestimated when recalled later (Schindler & Wiman, 1989).

Coutler (2001) suggests that limited memory capacity theory and the left-to-right comparison theory are interrelated because consumers learn that the left digits are of higher magnitude and are therefore are more likely to be kept in memory.

2.2.2 I

MAGE

E

FFECTS

Image effects are using a different argument approach of explaining the impact of charm pricing on

consumer behaviour. Furthermore, literature subcategorizes image effects into price-image effects and

quality image effects (Stiving & Winer, 1997). Price image effects result from assumptions consumers

17

may make based on whether a product has a round price or a charm price. Schindler and Kibarian (2001) propose that charm prices can convey a low-price image. For example, subjects in a lab-study were more likely to judge an item as “probably on sale” if it had a charm-price compared to a rounded price (Quigley & Notarantonio, 1992; Schindler & Kibarian, 2001). Another piece of research suggests that people are less likely to believe that a price has been recently raised (Schindler, 1984).

Schindler and Kibarian (2001) propose that such a low-price image is generally favourable for firms as it conveys that the price is relatively low. Anderson and Simester (2003b) found in a series of field experiments that prices ending on a 9 have been able to increase consumer demand. Thus, charm prices may function in a similar way as other sales cues.

However, research suggests that a favourable price image (i.e. cheaper) may come at the expense of the quality image. Round prices may indicate superior quality while charm prices may indicate inferior quality (Coulter, 2001). Schindler and Kibarian (2001) found that charm prices can convey an unfavourable quality image. Ads with charm prices by higher quality retailers had a negative impact on subjects’ quality evaluations. The perceived lower quality affected both the perceived item quality and the overall merchandise quality. Stiving (2000) provides an alternative theoretical explanation for why it seems like firms use round prices to signal quality. He proposes that price endings themselves are not a signal of quality. However, firms use the price level as a signal of quality and high-quality firms are more likely to use round prices. Round prices are artefacts of a firm’s initial choice to signal the use of price as a quality signal. It is concluded that it may seem that round prices are the quality signal even if the actual quality signal is the firm’s high price level. Nevertheless, consumers may observe this relationship and learn that there is a correlation between round prices and quality (Stiving, 2000).

2.2.3 C

HARM

P

RICING IN

D

UTCH

R

ETAIL

Data gathered from the websites of four leading Dutch grocery chains

²

confirms that charm pricing is still the rule and not the exception. 68 percent of the prices in the online-store of the Netherland’s leading grocer, Albert Heijn, end on the digit 9. Directly followed by the ‘still charming’ digit 5 (14%). Round priced items count for 5 percent of the items while the remaining digits account for 1 to 2 percent each. The runner-up Jumbo is using less charm pricing than its competitor. 37 percent of the items have prices ending on a 9 and, again, 14 percent end on the number 5. A result from that is that the remaining ending digits all have percentage ranges between 5 to 8 percent. However, charm prices still account for the clear majority of the prices. The supermarkets Plus and Hoogvliet lie between Albert Heijn and Jumbo when it comes to the use of this pricing tactic. Another interesting observation one can make is that the use of charm prices is generally increasing with the price. For example,

2 The data has been provided by EFMI Business School located in Baarn (NL). The institute specializes on education and research for the (food) retail sector.

18

Albert Heijn uses the digit 9 for 60 percent of their items between € 1 and € 2 but for 80 percent of the prices between € 9 and € 10. Presumably, this is done to decrease the perceived price of more expensive items.

Nevertheless, more recently some retailers go back to more rounded figures (e.g. € 2.50 instead of €

2.49). Examples are the Dutch grocery chain Deen (see image 1) or the Dutch variety store Hema (see

image 2). The Dutch grocery chain Hoogvliet actively advertises that it uses round prices for its

special offers (see Image 3). However, it is not clear if this strategy is successful so far.

19

IMAGE 1

Rounded Prices at Dutch Grocery Retailer Deen (Source: Prof Dr. Laurens Sloot)

IMAGE 2

Rounded Prices at Dutch Variety Store Hema (Source: Hema.nl)

20

IMAGE 3

Advertising for Round-Priced Special Offers at Hoogvliet (Source: Hoogvliet.com)

21

3. C ONCEPTUAL F RAMEWORK AND D EVELOPMENT OF

H ^YPOTHESES

FIGURE 1 Conceptual Model

In the present chapter, a conceptual model is proposed that integrates the concepts of price promotions, reference prices and charm pricing into a comprehensive framework of relationships. The aim is to explain the impact of the mentioned concepts on consumers’ perception and their behavioral intention.

A way to draw attention to price promotion in form of temporary price reductions (TPR), is presenting it alongside with an ARP. Hence, consumers are confronted with two ERP. First, an ERP that informs consumers about the current price of the product. Second, an ARP contrasting the current advertised sales price with another price. Very often this is the former, more expensive, sales price to highlight the attractiveness of the current deal or the advised selling price from the brand manufacturer. This is done to highlight a discrepancy between the two prices. Consumers’ evaluations are not purely based on those two price cues but are influenced by contextual cues (e.g. the focal brand), situational influences (e.g. the store) and their IRP (Grewal et al., 1998; Rajendran & Tellis, 1994).

The proposed conceptual model (see figure 1) extends a model proposed by Grewal et al. (1998). They

applied a comparable model in a durable consumer goods setting (i.e. bikes). In this study, the focus

lies on FMCG products. Their model incorporated the exogenous constructs ARP, a lower advertised

selling price and the consumer’s perceived quality. Both external prices had two levels (i.e. high and

22

low). However, the current study proposes a model that does not only discriminate the advertised sales price based on the advertised price level (APL) but also presents the prices as a charm or round prices.

Hence, it discriminates between APL and advertised price type (APT).

Consequently, the model has four exogenous constructs (perceived quality, APL, APT and ARP.

Furthermore, the model has four endogenous constructs (IRP, perceived transaction value, perceived acquisition value, WTB). In the following, the theoretical background, the proposed relationships and the respective hypotheses are proposed.

3. 1 P ERCEIVED Q UALITY

The present paper employs Zeithaml’s (1988) definition of perceived quality. The quality concept is about a product’s excellence or superiority compared to its competition. However, perceived quality differs from actual product quality. Actual quality is more mechanistic in its nature. Hence, it

“involves an objective aspect or feature of a thing or event” (Holbrook & Corfman, 1985: 33).

Perceived quality on the other hand is more humanistic and “involves the subjective response of people to objects and is therefore a highly relativistic phenomenon that differs between judges”

(Holbrook & Corfman, 1985: 33). Furthermore, perceived quality is a global assessment, which is related to an overall attitude towards a product. Additionally, it is rather abstract and usually not directed towards a specific feature (Zeithaml, 1988). Hence, perceived quality is defined as a consumer’s “estimate of a product’s cumulative excellence” (Grewal et al., 1998: 47).

The price level can increase consumers’ perceptions of quality (Dodds, Monroe & Grewal, 1991).

Furthermore, research on the image effects of charm pricing have shown that charm pricing can have a negative effect on perceived quality. However, Grewal et al. (1998) state that the influence of price level on perceived quality is only expected to be increasing when there are few other cues available or if the product is rather ambiguous (e.g. wine) (Almenberg & Dreber, 2011). However, in this study we will use an established product of a well-known national brand. Hence, it is assumed that price level and price type will not affect perceived quality. However, similarly to Grewal et al. (1998) the relationship between APL will be tested for. Additionally, the study tests for a possible relationship between APT and perceived quality caused by a possible negative image effect of charm prices.

However, no relationships are expected and hypothesized.

3.2 I NTERNAL R EFERENCE P RICE

Reference prices are “price[s] or price scale[s]) in buyers’ memories that serves as a basis for judging or comparing actual prices” (Grewal et al., 1998: 47). People may enter the store with an IRP in mind.

Thus, they may already have an idea what a product from a certain brand or category costs. However,

prior studies have shown that not only past shopping occasions have an impact on internal consumer

23

reference prices. Consumers may use external cues that they observe in the current shopping environment.

It is assumed that consumers’ perceptions of quality can impact their IRP. Consumers may have an initial level of perceived quality that is activated by external cues. In the current study, it is likely that the focal brand (i.e. Ariel) or the focal product (i.e. Ariel color detergent) elicits some knowledge structures in the brains of the respondents which is based on marketing activities of the brand or previous usage of the brand or product (Keller, 1993). Subsequently, their initial level of perceived quality is used to develop their IRP (Grewal et al., 1998; Monroe, 1973). Hence, it is assumed:

H1: The relationship between perceived quality and IRP is positive.

It has been shown that such ERP and other external cues can have an impact on consumers’ internal reference price. The current study focuses on a possible integration of advertised sales prices and a (higher) ARP as external reference prices. The respective sales price can be higher (i.e. lower discount) or lower (i.e. higher discount) and can be framed as a round or a charm price (e.g. € 4.00 or

€ 3.99). Thus, people may have an IRP for a certain type of product in their mind. However, it is assumed that consumers may use the provided external reference prices as anchors and adjust their IRP accordingly (Tversky & Kahneman, 1974; Einhorn & Hogarth, 1986; Wansink et al., 1998).

Hence, a higher (lower) APL and a higher (lower) ARP in a store may increase (decrease) the IRP.

H2: APL has a positive effect on IRP.

H3: ARP has a positive effect on IRP.

As mentioned before, not only the APL may have an impact on peoples’ perceptions. An additional factor that may have an impact is whether a price is shown as a round or a charm price. Showing a charm price (i.e. ending with the digit 9) can lead to level effects which may have an impact on the relationship between price level and the IRP. First, due to the rounding down-effect consumers may have the tendency to round prices down (i.e. drop-off the rightmost digits) when looking at prices.

Second, there may be mental processing errors due to left-to-right comparison. This assumes, that

consumers directly compare prices with another external reference price but tend to compare prices

from the left digit to the right digit and pay no or less attention to the right digit(s) (Stiving & Winer,

1997; Coutler, 2001). Third, due to the limited memory capacity of humans, people tend to only

process the most important part of a price. Consumers have learned that the left digits have a higher

financial impact than the rightmost digits. Thus, they are more likely to store the left digits of a price

(Brenner & Brenner, 1982; Stiving & Winer, 1997; Coutler, 2001; Wedel & Leeflang, 1998).

24

Finally, the discussed mechanism all may influence the proposed relationship between price-level and consumers’ internal reference prices. Consumers might 1) drop-off the last digit(s), which in turn leads to perceived lower price level; 2) consumers may form reference prices based on the presented shelf- prices but consider only the first digits when making inferences; 3) due to the limited capacity of memory consumers may only process the first digits of the prices. Therefore, it is hypothesized:

H4: As the APT changes from charm to rounded, ceteris paribus, the effect of APT on IRP will be positive.

3.3 P ERCEIVED V ALUE

The present paper uses a distinction between perceived acquisition value (PAV) and perceived transaction value (PTV).

Perceived acquisition value (PAV) is defined as the “perceived net gains associated with the products or services acquired” (Grewal et al., 1998: 48). A former study using the PAV approach by Dodds et al. (1991) states that the perception of value is a result of a cognitive trade-off between perceived quality and the monetary sacrifice that must be made (i.e. the price). Hence, PAV is positively influenced by the benefit the consumers expect from an acquisition and usage. The perceived product quality and price perceptions are major antecedents. Furthermore, the PAV is negatively influenced by the price that must be paid. Hence, raising (decreasing) the price decreases (increases) PAV while deceasing (raising) the perceptions of quality lowers (raises) PAV. The PAV is highly related to the concept of satisfaction in the post-purchase literature. Satisfaction implies that the sacrifices made are adequately rewarded (Grewal, 1995; Howard & Sheth, 1969). Perceived acquisition value can be seen as the counterpart in a pre-purchase environment. Hence, following hypotheses are made:

H5: The relationship between perceived quality and PAV is positive.

H6: The effect of APL on PAV is negative.

Perceived transaction value (PTV) is defined as the “perception of psychological satisfaction or pleasure obtained from taking advantage of the financial terms of the price deal” (Grewal et al., 1998:

48). The concept is highly related to price fairness constructs from equity theory. Equity theory states

that individuals evaluate the gains of investments in comparison to the investments they make

(Martins & Monroe, 1994). PTV is associated with the pleasure derived from getting a fair price. The

PTV thus implicates that consumers confronted with a price-comparison may perceive additional

value beyond the pure transaction value. The concept shares perceived quality as an antecedent with

PAV. Hence, if the perceptions of quality are higher it is assumed that a respective offer becomes

more attractive. Furthermore, it is proposed that they judge the respective price comparison with their

25

internal reference price. Hence, a high internal reference price makes the advertised reference price more attractive (Dodds et al., 1991). Respectively, it is proposed that:

H7: The effect of APL on PTV is negative.

H8: The effect of IRP on the PTV is positive.

H9: The relationship between perceived quality and PTV is positive.

Furthermore, the researcher proposes a relationship between APT and PTV. This proposition rests on the assumed existence of price-image effects of charm pricing. As mentioned earlier, studies suggest that charm prices can convey a low-price image. For example, in a study consumers were more likely to judge items as “probably on sale” if they had a charm-price (Quigley & Notarantonio, 1992;

Schindler & Kibarian, 2001). Hence, this research hypothesizes that the “sales image” may increase the psychological satisfaction of the offering. Furthermore, Schindler (1984) found that people are less likely to believe that a price has been recently raised when framed as a charm-price. This may add to the fairness notion of PTV concept. Hence, people may feel that this is a fair deal because prices have not been increased and are on discount. Hence, it is hypothesized that:

H10: As the APT changes from charm to rounded, ceteris paribus, the effect of APT on PTV will be negative.

Prior studies assumed that PTV and PAV are two independent constructs (Thaler, 1985; Monroe &

Chapman, 1987). Dodds et al.(1991) as well assumed that both constructs are independent from each other. In their study, a Principal-Component-Analysis (PCA) demonstrated that PAV and PTV are indeed separate constructs (i.e. the constructs’ items clearly loaded on two separate factors). However, they found that PAV and PTV are not fully independent. They propose that PTV is rather increasing PAV and is not directly influencing consumers’ willingness-to-buy. Hence, “getting a good deal” can increase the overall perceived net gains associated with the products or services acquired and is not directly affecting consumers’ willingness-to-buy. Thus, the present paper proposes that:

H11: The effect of PTV on PAV will be positive.

3.4 W ^ILLINGNESS - ^TO -B ^UY

It is ideal to investigate the impact of the employed marketing tactics on consumer behavior. In

survey-research it is not possible to observe actual consumer behavior. However, based on Ajzen’s

(1991) theory of planned behavior one can argue that the behavioral intention to perform certain

behavior may help predicting actual behavior. Ajzen’s (1991: 181) theory states that, “the stronger the

intention to engage in a behavior, the more likely should be its performance”. Hence, it is desirable to

26

measure the behavioral intention to engage in a purchase based on the offer that is proposed to the participants. Thus, instead of purchases the participants’ willingness-to-buy (WTB) is measured.

WTB is defined as “the likelihood that the buyer intends to purchase the product” (Grewal et al., 1998:

48). Similar concepts exist under various names. For example, purchase intention, which is “the subjective probability of buying a certain product or brand by the consumer” (Khan & Azam, 2016:

23). The perception of value is said to directly influence the WTB (Dodds et al., 1991). However, this study distinguishes between PAV and PTV. It is proposed that PTV directly influences the consumers’

WTB.

H12: The effect of PAV on WTB is positive.

27

4. M ^ETHODOLOGY

In this chapter, the researcher describes the research plan, proposes some control measurements and describes the population and sample of the current study. Before continuing to the following results chapter, it proposes a plan of analysis.

4.1 R ESEARCH P LAN

The present study used a 2x2x2 between-subjects-design. Hence, two APL (low, high), two APT (charm, round) and two ARP (low, high). Respondents have been randomly assigned to one of the eight manipulation conditions. Random allocation has the advantage that causal relationships can be determined with a higher certainty and it gives some certainty that personality traits and moods do not affect the results (Aronson et al., 1998).

In a self-administrated digital survey, respondents have been asked to imagine that they are on a usual shopping trip and realize or noted down on their shopping list that they are in need for a color- detergent. Promotional displays; which resemble displays that can be found in supermarkets, have been designed to highlight the price cues (see image 4). In a short introduction text the basic features of the product (i.e. color-detergent for 18 loads) have been presented together with a picture of the product.

IMAGE 4

Promotional Displays Showing Either a Charm or a Rounded Price

4.1.1 P

RODUCT

C

HOICE AND

C

ATEGORY

C

HOICE

The current study uses Ariel liquid color detergent for 18 wash-loads, a product from the detergent

category. The product category has been chosen because it is expected that respondents have a high

level of familiarity with the category. Washing detergent is a necessity and it is expected that detergent

is purchased and/or used by a clear majority of the respondents. The product has been chosen because

the product brand is widely known in the Netherlands and Germany. It is manufactured by Procter and

28

Gamble, the largest FMCG company in world (Gensler, 2016). It was expected that most respondents are familiar with the brand and the product’s features and have some expectations and knowledge about the products quality. Furthermore, the author assumed that it is not a so-called ‘signpost’ item.

Signpost items are items where respondents are very certain about the actual price level. For such items, respondents easily could distinguish expensive and inexpensive price levels without any further cues, which could make them ineffective as the focal products for the study of external price cues on consumer behavior (Anderson & Simester, 2003a).

4.1.2 P

RICE

M

ANIPULATIONS

Actual prices for the product used varied between retailers. The Dutch chains Albert Heijn and Plus both asked for € 6.79. However, at the time of the study market leader Albert Heijn offered a ‘2 for € 8’ special offer. Hence, taking advantage of the offer would lead to an effective sales price of € 4 at that time. Competitor Jumbo offered the product slightly cheaper regular sales price for a price of € 6.49. However, Jumbo had a year-round 3-for-2 offer which could potentially lead to an effective price of € 4.33 per item. Dutch drug store Blokker sold the detergent for € 5 and used the manufacturer’s suggested retail price of € 7.49 as an advertised reference price.

The ARP used in the study have been € 6.49 and € 7.49. The lower price level is close to normal prices (i.e. non-sale prices) at Dutch supermarkets. The higher price level resembles the manufacturer’s suggested retail price of € 7.49 that e.g. Blokker used as a price comparison.

The APL used in this piece of research has been € 3.99 (€4 in the rounded condition respectively) in the low-condition and € 5.99 (€ 6) in the high condition. First, the low condition resembles a price level close to special offers at e.g. Albert Heijn or Jumbo. Second, the high condition price was just below the regular sales prices at Albert Heijn and Jumbo.

Van Heerde, Leeflang and Wittink (2001) state that there are threshold and saturation levels in the domain of price discounts, which vary between categories and items. In the beverage category, own- item threshold levels have been at around 5 percent and own-item saturation effects have been between 15 percent and 35 percent (varying between items). In the tuna category, own-item threshold levels have been at 10 percent and saturation levels have been between 40 and 45 percent (varying per item). However, the present study focuses on the detergent category. The discount levels in the study are at 7.6 (i.e. 7.7 percent in the charm condition due to the 1 cent price difference), 19.9 percent (i.e.

20.0 percent), 38.4 percent (i.e. 38.5 percent) and 46.6 percent (i.e. 46.7 percent). Hence, 7.6 (i.e. 7.7)

percent is presumably right at the threshold level and resembles a rather unattractive price discount,

while 38.4 (i.e. 38.5) and 46.6 (i.e. 46.7) percent might be just below and slightly above the saturation

level and resembles extreme price discounts. However, the price discount level of 19.9 (i.e. 20.0)

percent resembles a quite realistic and common discount level (Sloot, 2011).

29

TABLE 1

Conditions, Manipulations and Means

Condition 1 2 3 4 5 6 7 8

Frequency (n)

33 36 41 44 44 34 27 42

Frequency (%)

11.0 12.0 13.6 14.6 14.6 11.3 9.0 14.0

Manipulation

APL Low Low High High Low Low High High

APT Charm Round Charm Round Charm Round Charm Round

ARP Low Low Low Low High High High High

Prices Sales Price (€)

3.99 4.00 5.99 6.00 3.99 4.00 5.99 6.00

ARP (€) 6.49 6.49 6.49 6.49 7.49 7.49 7.49 7.49

Means Perceived Quality

3.23 3.41 3.31 3.16 3.39 3.28 3.44 3.30

IRP 3.89 4.06 4.19 4.10 4.36 4.12 4.93 4.69

PTV 4.46 4.46 4.13 3.89 4.50 4.50 4.78 4.22

PAV 5.08 5.30 4.21 3.96 5.50 5.24 4.97 4.65

WTB 4.63 4.62 3.75 3.67 5.18 4.78 4.77 4.23

4.1.3 M

EASUREMENT AND

S

CALING

The scales used to measure the four endogenous constructs (IRP, PAV, PTV, WTB), the exogenous construct perceived quality and the control variables (in-store deal proneness, brand equity) and the respective items are summarized in a table can be found in Appendix 1. The constructs of perceived quality, PAV, PTV, willingness-to-buy, in-store deal proneness, brand equity have been measured using a 7-point Likert scale. The respondents’ internal reference prices were measured using free estimates in Euro and Eurocents.

The measure for willingness-to-buy was measured using a set of three items that asked for the respondents purchase probability. The items originate from Dodds et al. (1991) and have been used in various studies (e.g. Grewal, Krishnan, Baker & Borin, 1998; Grewal et al., 1998). The IRP measure consists of consumers’ estimate for a normal price (i.e. an average market price) and their fair price perception as their basis. Both estimates are free estimates by participants in Euro and Eurocents. The measure has been formerly employed by Grewal et al. (1998). The scales are grounded in prior research by Lichtenstein and Bearden (1989) and Urbany et al. (1988). The measure for perceived quality has been initially used by Dodds et al. (1991) and employed in a variation by Grewal et al.

(1998). The measures for PAV and PTV both come from Grewal et al. (1998). However, items and in

the present study have been adjusted to fit the product at hand (i.e. detergent) and the FMCG-setting.