• No results found

Effect of consumer reviews on consumers’ purchase intention and sales - the moderating role of consumer expertise

N/A
N/A
Protected

Academic year: 2021

Share "Effect of consumer reviews on consumers’ purchase intention and sales - the moderating role of consumer expertise"

Copied!
60
0
0

Bezig met laden.... (Bekijk nu de volledige tekst)

Hele tekst

(1)

Effect of consumer reviews on consumers’ purchase intention and sales -

the moderating role of consumer expertise

Groningen, July 2012

Master thesis

University of Groningen, the Netherlands Faculty of Economics and Business Master of Business Administration Marketing Management

Sebastiaan Willemse Taco Mesdagstraat 54

9718 KN Groningen, the Netherlands + 31(0)634126879

willemse.sb@gmail.com Student number: 1796577

Supervision

University of Groningen, Faculty of Economics and Business, Department of Marketing First supervisor: dr. S. (Sonja) Gensler

(2)

Management summary

Nowadays more and more consumers are using the Internet to search for information and to buy products and services. The most important difference between online and offline

shopping is that consumers cannot see, try and touch products or services. To overcome this limitation online retailers offer consumers the possibility to evaluate products and services through consumer reviews. These reviews are shown on the Internet and are for both consumers and online retailers important. Consumers use the reviews as source of

information helping them to make purchase decisions and online retailers use the reviews as a service to influence purchase intention and sales.

The influence of review quantity, variance and valence on sales is measured in this study. This first study used logistic regression models to see what influences these review

characteristics have on sales. Three mayor conclusions were drawn: (1) review quantity has a positive effect on sales, (2) review variance has a negative effect on sales and (3) review valence has a positive effect on sales.

Also the effect of review quality and review type on purchase intention is measured in this study. Besides, the effect of consumer expertise is measured. This second study used a 2 x 2 x 2 between-subjects design, review quality (high vs. low), review type (attribute-centric vs. benefit-centric) and consumer expertise (high vs. low). Independent samples t-test are used to measure what influence these different aspects have on purchase intention. In despite of favorable effects on the purchase intentions means no significant effects, between groups, were found for: (1) the effect of review quality on purchase intention, (2) the effect of

attribute-centric reviews on purchase intention (moderated by consumer expertise) and (3) the effect of benefit-centric reviews on purchase intention (moderated by consumer expertise). The findings of this study can be used by online retailers to manage consumer reviews on the retailers’ websites. Online retailers have to take the different characteristics of reviews into account when designing the review display on their website. Using the results of this study in a ‘online review strategy’ can certainly contribute in the way website sales and consumer’ purchase intention are influenced by reviews.

(3)

Preface

This master thesis is written to obtain the masters title in economics and business, marketing management, at the University Groningen. Months of intensive research have resulted in this thesis: ‘Effect of consumer reviews on consumers’ purchase intention and sales – the

moderating role of consumer expertise’.

To find a good topic for this thesis, I have gathered a lot of information of several marketing subjects. I have read a lot about interesting topics like, experience marketing, consumer behavior and viral marketing. Yet, these subjects are not specific enough for people who are not working daily on these subjects. Therefore I have chosen a subject in online marketing which is recognizable for many consumers: consumer reviews.

More and more consumers are using the Internet to purchase products or services. A way to gather information on products and services is to use consumer reviews. In this study the influence of these reviews on consumers’ purchase intention and sales is examined. Besides, consumer expertise as a moderator of the effect is examined.

Without help of other people it was not possible to write this thesis, therefore I would like to thank some persons. First of all I would like to thank my thesis supervisor, Sonja Gensler, for her advice and guidance during the writing process of my thesis. Besides, I would like to thank my second supervisor, Thorsten Wiesel, for reviewing my thesis. I would also like to thank Daan Lodewijks, from Vakantiepanel.nl for his advice and supply of hotel reviews data.

July 2012

(4)

Table of contents

Management summary...2

Preface...3

Table of contents...4

1 Introduction ...5

1.1 Problem statement & research question ...6

1.2 Managerial- and academic contribution ...7

1.3 Structure of the thesis ...9

2 Theoretical framework ...10

2.1 Consumer reviews as a form of electronic word of mouth ...10

2.2 Effect of review quality on purchase intention ...11

2.3 Effect of review quantity on sales ...12

2.4 Effect of review variance on sales...13

2.5 Effect of review valence on sales ...14

2.6 The moderating effect of consumer expertise ...14

2.7 Conceptual model...16

3 Research design and method ...17

3.1 Research design and method study I ...17

3.2 Research design and method study II...18

4 Results ...21

4.1 Results of study I...21

4.2 Results of study II...24

4.3 Summary of findings ...29

5 Conclusion...30

5.1 Main conclusion and recommendation...30

5.2 Limitations and future research...31

References...33

(5)

1

Introduction

“ Consumers love to review”

- J. Roos, Commercial manager Zoover, 2011-

With the increasing popularity of the Internet also electronic commerce is growing rapidly (e.g. McKinsey Global Institute, 2011).As a new marketing channel, online retail formats differ from the traditional brick and mortar retail formats. Where consumers in brick and mortar stores are able to touch, feel, see and ask employees about search products, online purchase judgments must be based on product information (e.g. Alba et al., 1997). To

overcome this limitation, online retailers offer consumers the opportunity to review a product and share their evaluations via their website. These product evaluations, consumer reviews, are helpful for consumers in making purchase decisions. Especially for experience products, consumer reviews are important. Experience products have attributes that cannot be known until the consumer has purchased and used the product. Reviews can help consumers to understand and imagine these attributes (e.g. Zeithaml, 1988). This research will mainly focus on the characteristics of reviews of experience products.

The quantity and quality of reviews are important characteristics which influence the way consumers process information.(e.g. Park, Lee & Han, 2007).The number of reviews (review quantity) are an indication of the popularity of the product, since it is reasonable to assume that the number of reviews is related to the number of consumers who have bought the product (e.g. Chen & Xie, 2004).A review of high quality is a review that contains logical and persuasive arguments and supports its evaluation with reasons based on facts about a product (Park, Lee & Han, 2007).Next to quantity and quality of reviews also variance and average score are important characteristics which affect consumer information processing. The variance in reviews can increase or decrease the purchase decision risk for consumers (e.g. Chevalier & Mayzlin, 2006).For example, one review has an average score of 3 out of 10 and an other review has an average score of 8 out of 10. The lack of consensus of reviews can create uncertainty for consumers such that consumers are faced with the task of judging contradictory reviews and product attribute information (e.g. Park & Park, 2011).

(6)

reviews. Hu, Liu & Zhang (2008) stated that consumers use the latest posted consumer reviews at first to ‘check’ the average score. The authors stated that matching scores and higher scores will increase purchase intention and low average grades and grades below the average score will decrease purchase intention. Average scores can influence consumers’ purchase intentions with the result that the amount of sales will be influenced (e.g. Armstrong, Morwitz and Kumar, 2000). Not only review characteristics can influence purchase intention and sales. Also consumer characteristics, like consumer expertise, can influence purchase intention and sales.

Consumer expertise has a moderating effect on the information processing of consumers (e.g. Alba & Hutchinson, 1987). The elaboration likelihood model (ELM) suggests that the same information can be processed in different ways depending on the expertise of consumers. Consumers with high expertise use prior product experience and knowledge to evaluate the information in the reviews. Besides, consumers with low expertise may be hampered by the lack of knowledge and can experience difficulties in processing review information (e.g. Park & Kim, 2008).

To conclude this paragraph, could it be said that reviews are important factors in affecting purchase intention and sales? Reviews can recover the absence of ‘offline’ store attributes in online retail stores, but how these reviews influence purchase intension and sales is dependent on the expertise of the consumer. Consumers love to review, but will reviews really affect consumer purchase intention and sales?

1.1 Problem statement & research question

Literature showed that reviews have an effect on purchase intention and sales (e.g. Chevalier and Mayzlin, 2006; Duan, Gu & Whinston, 2005; Chen, Wu & Yoon, 2004). However

questionable is whether consumer expertise moderates the effect of review quality and review type on purchase intention and whether review quantity, variance and average score (review valence) affect sales.

Several authors researched the effect of review quality and review quantity on purchase intention (e.g Park, Lee & Han, 2007; Park & Kim, 2008). Park, Lee & Han (2007), found that the strength of this effect is dependent on consumer involvement. Park & Kim (2008), found that consumer expertise moderates the effect of review quality and reviews quantity on purchase intention. Both researches focused on search products. Interesting to research is what the effects are on experience products.

(7)

variance on sales (e.g Chevalier & Mayzlin, 2006; Chen, Wu & Yoon, 2004; Duan, Gu & Whinston, 2005; Clemons et al, 2006). Chevalier & Mayzlin (2006) found that if the average score of a review increase also the relative sales for that product increases. Chen, Wu & Yoon (2004) showed that the quantity of reviews have a positive effect on sales but that average score of reviews is not related to sales. In line with this research Duan, Gu & Whinston (2005) showed that sales are significantly influenced by the quantity of reviews. Clemons et al. (2006) found that both mean and variance of review ratings are positively related to product sales. Interesting to research is how all characteristics influence sales of one product. This is missing in previous researchers because all characteristics are tested on different products / datasets.

Although various results are shown about the effect of consumer reviews on purchase intention and sales, never the quality, quantity, variance and average score of reviews are combined in one research for the same experience product. In this research the effects of quality, quantity, variance and average score of reviews on both, purchase intention and sales, are tested. Besides, the effect of consumer expertise as moderator of the effect of review quality on purchase intention is tested.

For this research, the following research question is formulated:

What is the effect of consumer reviews as well as consumer expertise on consumers’ purchase intention and sales?

In order to come up with a comprehensive answer on this research question, attention will be devoted to the following sub questions:

• To what extent does the quantity, variance and average score affects sales? • To what extent does review quality influence purchase intention?

• What is the moderating effect of consumer knowledge on the effect of review type on purchase intention?

1.2 Managerial- and academic contribution

(8)

as result from reading a review and that purchase intention measures the likelihood that a consumer will buy a product after reading a consumer review. Purchase intention can been seen as the process of attitude forming towards the product, using reviews, which leads to a purchase probability.

Previous studies on reviews are limited to several characteristics of reviews; quality, quantity, variance and average score (e.g. Park, Lee & Han, 2007; Park & Kim, 2008). This research considers all main characteristics of reviews in one study to measure the effect of all characteristics on purchase intention and sales. Main advantage of combining the

characteristics is that it provides information to what extent purchase intention is changed and how sales is affected by which characteristic of the review. Next to these effects, this study also takes consumer expertise into account. Park & Kim (2008) researched the effect of consumer knowledge for review quality and review quantity. The authors defined knowledge as product knowledge and not knowledge as expertise with respect to the usage of consumer reviews in the buying process.

This study makes several managerial implications. The results of this study show the importance of managing consumer reviews. With the change of traditional word-of-mouth to electronic word-of-mouth also the way consumers share information and recommend

products change. For managers it is important to follow this change since eWOM, online consumer reviews, function as information source and recommender. Because these reviews can be used strategically and as communication channel it is important to know which characteristics of reviews influence purchase intention and which affects sales. With the results of this study managers can adjust reviews in that way that reviews have a positive effect on purchase intention and sales. For example, when managers know that the average score of reviews is very important in affecting sales, managers can adjust reviews in that way that the average score have a major position in the review.

(9)

1.3 Structure of the thesis

(10)

2

Theoretical framework

In this theoretical framework the focus is on the influence of consumer reviews on purchase intention and sales. Prior research on the influence of consumer reviews in general (2.1), review quality (2.2) and quantity (2.3), review variance (2.4) and valance (2.5) on purchase intention and sales will be discussed. Prior research on the moderating effect of consumer expertise will be described in paragraph 2.6 followed by two conceptual frameworks in paragraph 2.7.

2.1 Consumer reviews as a form of electronic word of mouth

The advancements in Internet technologies enabled new forms of communication platforms in which consumers can share information and opinions. In the context of this research we consider the concept of eWOM.Westbrook (1987) stated that electronic word of mouth (eWOM) can be defined as all informal communications through the Internet which is directed at consumers. These communications are related to product or service usage and characteristics. According to Lee (2009) electronic word of mouth differs in three ways from traditional word of mouth. First difference is that almost all eWOM communication is between anonymous consumers in which sender and receiver never have met. Therefore, the receiver of the information cannot look for similar sources and information of experts. It is therefore impossible for the receiver to check whether the information is credible. Second difference is that consumers experience no limits on space and time when generating eWOM. Consumers who use the Internet can share in eWOM by writing or reading reviews. Third and last difference is that eWOM is relatively longer available than WOM because

consumers who use the Internet can trace the eWOM history simply by scrolling down on website history. In the context of the definition, eWOM includes communication between companies and consumers as well as those between consumers themselves.

One form of eWOM are consumer reviews. Reviews can be defined as, online information, experience and opinion sharing between consumers concerning a product or service (e.g. Zhu & Zhang, 2010). Consumers search for all available product information and recommendations when the have the intention to purchase a product or using a service. Reviews have the capability to influence purchase intention because reviews have an

(11)

offer more information which is consumer-oriented, whereas online retailers offer more information which is product-oriented (e.g. product ratings, attributes and benefits). In reviews the attributes of the product are described as experiences and the performance of the product is measured from the perspective of the user (e.g. Bickart & Schindler, 2001). Besides, additional information which is not mentioned by online retailers is provided in reviews. Previous studies suggest that that customer-created information is more credible than seller-created information viewed from the perspective of trustworthiness (e.g. Dellarocas, 2003). Because reviews are based on consumers’ honest evaluations of the strengths and weaknesses of a product and seller information is only based on good aspects, reviews are likely to be more trustworthy than seller-created information (e.g. Bickart & Schindler, 2001). Next to the informer function, reviews also provide recommendations (recommender function).This is similar as in traditional WOM in which a consumer recommends a product to an other consumer (e.g. Chatterjee, 2001). However, source, volume, reachability and measurability are distinctive characteristics of reviews. Reviews are posted by, anonymous, individual consumers. Due to easiness of posting online reviews, reviews have a far greater abundance online that traditional WOM in the offline world. With current technologies it is easy to observe and count the amount of reviews. The dual role of reviews cause that reviews are an easy and effective tool to provide information and

recommendations on products.

2.2 Effect of review quality on purchase intention

Information quality can be described in terms of relevance, understandability, sufficiency and objectivity (e.g Park, Lee & Han, 2007). In their research the authors stated that the more extensive and better the information is, the greater the consumer is satisfied with the gained information. In addition, as increasing satisfaction of the consumer will also increase purchase intention (e.g. Bailey & Pearson 1983, Negash et.al. 2003). According to Douglas and Wind (1971), purchase intention is affiliated purchase behavior. Therefore there can be stated that if consumers’ purchasing intention increases purchase behavior will be positively affected. In addition, Su et al. (2008) found that information quality can have a positive effect on purchase behavior in that way it increases purchase intention.

(12)

There is no standard information format for consumers posting reviews, and as a result, each review is different from others. There are generally two types of reviews. Some reviews are subjective, emotional, and do not make reasoned arguments: “This hotel was so good that I go again next week” or “ I can’t believe I stayed at this hotel, it was so cool”. Other reviews are specific, clear, and back up with claims with reasons: “This hotel was really good: the rooms were clean, personal was friendly and the location was perfect!” (e.g. Forman, Ghose & Wiesenfeld, 2008).

In this study, review quality is defined as the quality of a review’s content from the perspective of information characteristics (understandability, relevance, objectivity and sufficiency) (e.g. Park, Lee & Han, 2007). Using this definition of review quality, the last review example is a high-quality review because it is logical and persuasive and gives reasons based on specific facts about the hotel. In contrast, the earlier review examples are low-quality reviews because they are emotional, subjective, and vacuous, offer no factual information, and simply make a recommendation. Consumers may treat reviews as a source of supplementary information of word of mouth. Since reviews are posted by consumers, who have booked the hotel in question, even subjective and emotional reviews (defined as low-quality reviews in this research) provide important and useful information when they are positive. If a review contains objective and understandable comments with sufficient reasons of recommendation, it is more persuasive than a comment that expresses feelings and

recommendations without specific reasons. Since previous consumers are anonymous on the Internet, consumers generally will not easily accept or believe a review posted on a website if it does not provide enough information (e.g. Ratchford et al. 2001). Therefore we can state the following hypothesis:

H1: Review quality affects purchase intention positively.

2.3 Effect of review quantity on sales

(13)

of message quantity, the more messages are processed, the more favorable the attitude towards the message. This favorable attitude towards the message will reduce or eliminate the uncomfortable feeling of risk exposure (e.g. Petty & Cacioppo, 1984). As a result, the number of reviews will have a positive effect on purchase intention. Interesting to research is the fact of there is a linear relationship between review quantity and sales. This non-linear relationship can be substantiated by the information overload theory. Information overload is defined as a state induced by a level of information exceeding the ability of an individual to assimilate or process during a given unit of time (e.g. Jacoby, Speller & Kohn, 1974). Besides, the information overload theory, research showed frequently that the quantity of reviews positively affects sales. We expect that the information overload theory is less relevant for reviews because more reviews cause less purchasing risk. Therefore, the following hypothesis is proposed:

H2: The quantity of reviews positively affects sales.

2.4 Effect of review variance on sales

Variance can be defined as the difference or discrepancy between reviews for a product. Sometimes reviews contradict each other to some or greater extent, which affect purchase intention and sales. Suppose, for example, consumers who consider to book a certain hotel read reviews from past bookers prior to making their final purchase decision. If most reviewers liked the hotel, it is highly likely that consumers would not hesitate to book the hotel. As is often the case, however, consumers observe that summary statistics for the hotel indicate similar mean ratings, yet the variance of those ratings differs across reviewers.

A hotel can be rated with a mean grade of 7 out of 10. Besides, individual reviews for the hotel represents grades of 10 out of 10 but also grades of 3 out of 10. Here, the lack of consensus in the reviews for the hotel rather increases decision risks, thereby creating an important problem for consumers such that consumers are faced with the task if integrating contradictory reviews and product attribute information into their own preferential judgment.

(14)

positive as well extreme negative reviews. Forman, Ghose & Wiesenfeld (2008) showed in their research that consumers find moderate reviews less helpful in decision-making than extreme reviews. We expect that consumers want to reduce or eliminate the uncomfortable feeling of risk exposure when purchase a product. Therefore we expect that variance in reviews will negatively effect consumer purchase intention and therefore lower the sales. Therefore the following hypothesis is proposed:

H3: The higher the variance of reviews the lower the sales

2.5 Effect of review valence on sales

The average ratings (valence) of a review represent the summed grades of each review divided by the total number of reviews. Chevalier and Mayzlin (2006) found that there is a significant relationship between review valence and sales. On the other hand, the lower the valence the stronger sales will be negatively effected. Besides, the authors’ results show that low (negative) average scores have a stronger effect than high (positive) scores grades. The impact of valence of reviews is mixed. For example, Liu (2006) and Duan et.al (2005) found that the valence of movie reviews on Yahoo! Movies do not have significant impact on weekly box office revenues (sales). Besides, Zhang & Dellarocas (2006) found a positive significant relationship between the valence of reviews and the weekly box office revenues. The authors reported that a 1-point increase in rating of use reviews on Yahoo! Movies is associated with an increase in box office revenues between 4% and 10%. We expect that how higher the average rating of a review (valence) the higher the sales will be. Therefore the following hypothesis is proposed

H4: High valence of a review has a positive effect on sales

2.6 The moderating effect of consumer expertise

(15)

merits of a stimulus. If a consumer is highly motivated and have the ability to process information the consumer will process the information through the central route. However, when the consumer in not motivated, have no ability and is under time pressure he/she will process information through the peripheral route and will rely on heuristics. (e.g. Alba & Hutchinson, 1987). Therefore, a review with many arguments can be accepted if a consumer thinks that ‘more is better’, without the need to carefully evaluate those arguments (e.g. Petty & Cacioppo, 1984).

In the ELM, expertise is associated with the ability to process information. Consumers with high expertise have knowledge and can rely on previous experiences to judge and evaluate the information. These high expertise consumers process information through the central route. Consumers with low expertise have less knowledge, are not motivated an have no ability to process the information in detail. These consumers trust on basic heuristics and process information via the peripheral route (e.g. Petty & Cacioppo, 1986). There can be two types of reviews distinguished: attribute- centric reviews and benefit-centric reviews (e.g. Park & Kim, 2008). Attribute-centric reviews contains information about ‘technical’ aspects of an hotel (how many rooms, facilities and personal). Benefit-centric reviews contain information about the benefits of staying at the hotel (feeling good, meeting people). Consumers with high expertise need reviews framed as attribute-centric to meet their

information-processing strategy. On the other hand, reviews framed as benefit-centric do not meet their information processing needs because benefit-centric reviews do not give that kind of information high expertise consumers need. On the other hand, consumers with low expertise lack the ability to understand and assess information from attribute-centric reviews, so they prefer the reviews framed as benefit-centric. Therefore, attribute-centric reviews have a stronger effect on the purchase intention of consumers with high expertise than consumers with low expertise. On the other hand, attribute-centric reviews have a stronger effect on purchase intention of consumers with low expertise than consumers with high expertise. Therefore the following hypotheses are proposed:

H5: Attribute-centric reviews have a stronger positive effect on the purchase intention of consumers with high expertise than consumers with low expertise.

(16)

2.7 Conceptual model

After reviewing a variety of literature, a conceptual framework is displayed. The conceptual framework exists of two conceptual models, figure 1 and figure 2. The first model is

constructed by combining hypotheses H2 review quantity, H3 review variance and H4 review

valence. The effects of these three components of reviews on sales will be tested using a hotel

dataset with hotel review information.

The second model is constructed by combining hypotheses H1 review quality, H5

Attribute-centric reviews and H6 Benefit-centric reviews. This second model also integrates consumer expertise as a moderator of the effect of reviews on purchase intention.

Two models are constructed because two different datasets are used in this research, both testing different independent and dependent variables. The first dataset ‘hotel dataset’ is used to test the relations stated in conceptual model 1. The second dataset, collected from a survey, is used to test the relations stated in conceptual model 2. For this research the following conceptual models are proposed:

Figure 1: Conceptual model I

Figure 2: Conceptual model II

Review Quantity

Review Valence

Review Variance Sales

(17)

3

Research design and method

This chapter contains a detailed description of the research design and –method of study I and II. The Netherlands was selected as an appropriate place for investigation for both studies, because in recent years more and more consumers are using reviews when they search online information of products and services (ABN AMRO & CBW-MITEX, 2011). Paragraph 3.1 start with a description of the research design and method of study I, followed by a

description of the research design and method of study II in paragraph 3.2.

3.1 Research design and method study I

In this first study the hypotheses of conceptual model I will be tested. A dataset from Vakantiepanel.nl will be used to test how review quantity (H2), review variance (H3) and review valence (H4) will affect sales (i.e bookings last 30 days on Vakantiepanel.nl).

Vakantiepanel.nl is one of the biggest holiday review websites in The Netherlands offering reviews of more than 65.000 hotels. Vakantiepanel.nl allows consumers to review hotels on several aspects: general impression, location, facilities, food & drinks, rooms, hygiene and price/quality. Component scores are added and divided to an average score. All average scores of reviews are calculated to an overall average score. In our dataset we randomly selected 40 hotels in 4 cities around the world; London, Paris, New York and Istanbul. To control for city preferences of reviewers we selected 10 hotels in each city. Variables. The data from Vakantiepanel.nl consists of data on product characteristics and reviews (e.g. table 1; appendix IV).

Subject Variable Type Measurement

Hotel Name Control Nominal

City Control Nominal

Star rating Control Ordinal

Average price Control Scale

Hotel

Bookings last 30 days on Vakantiepanel.nl

Dependent Scale Total average score Independent Scale Number of reviews Independent Scale Review

Variance average score Independent Scale

Table 1: Data summary study I

(18)

retailers know how these review characteristics influence sales, retailers are able to customize their strategies regarding the display of reviews in such a way that it increase their sales.

3.2 Research design and method study II

In this second study the hypotheses of conceptual model II will be tested. Using data from an survey the effect of review quality (H1) and review type (H5, H6) on consumer purchase intention will be tested. Besides, the moderating effect of consumer expertise will be tested.

The hypotheses will be tested with a 2 x 2 x 2 between-subjects design. The two independent variables are review quality (high vs. low), review type (attribute-centric vs. benefit-centric).Consumer expertise (high vs. low) is used to investigate a moderating effect. The hypotheses will tested empirically with the help of a survey among Dutch consumers.

Experimental product. A hotel in New York. There are two reasons for using a hotel as product: (1) hotels are frequently booked at online travel agencies, (2) consumers tend to rely on reviews of consumers because a hotel stay is an experience product. Because almost every consumer know the city New York (United States), New York is opted as the location of the hotel.Travel agency information and reviewer information will not be presented in order to prevent any effects on the interpretation of the review.

Experimental procedure. Two surveys will be prepared and randomly supplied by email. The surveys are, except the reviews, the same for all respondents.The scale items are developed based on prior literature. Except for the demographic questions and the questions regarding purchase intention all questions are 7-point Likert scales with a range of 1: strongly disagree and 7: strongly agree. The first survey shows an high quality review and an

attribute-centric review, the second survey shows a low quality review and a benefit-centric review. A focus group interview (10 subjects) was used to decide the review quality and whether the reviews are attribute- or benefit-centric. The focus group had to judge 12 selected reviews. With this information four reviews are composed and used in the surveys. To

(19)

another review will be shown (attribute-centric- or benefit-centric review).The survey ends with questions on purchase intention and demographics. Purchase intention is measured on a 10-point consumer buying intention scale (e.g. Juster, 1966). These measurements ranged from 10, representing certain, practically certain, to 0, no change, almost no change. The surveys can be found in appendix I and II.

Variables. In table 2 the variables, scales, items and references are described.

Type of variables Scale Items Reference

Independent variable Review Quality 1) This hotel review has sufficient reasons supporting the given grade.

2) This review is objective. 3) This review is

understandable.

4) This review is credible. 5) This review is relevant. 6) In general, the quality of this review is high.

Park, Lee & Han (2007)

Dependent variable Purchase Intention 1) If the price of this hotel is reasonable, I definitely will book this hotel. 2) After reading this hotel review, what is the probability that you will book this hotel?

Juster, F.T. (1966)

Moderating variable Consumer expertise 1) I always use hotel reviews, written by other consumers, during my search for a hotel. 2) When planning a city trip, I know where I can book a hotel online. 3) I keep current on new hotel offers.

4) I have a good overview of the latest information on hotel offers.

5) I consider myself as experienced in using hotel reviews, when I book a hotel online.

Kerstetter & Cho (2004); Alba & Hutchinson (1987)

Table 2: Overview of different scales to measure variables

Control variables. Survey results could be affected by stimuli like price and the brand name of products. This is particularly the case when testing consumer reviews (e.g. Hong, Thong & Tam, 2004). Besides, the product stimulus characteristics of the reviewer can affect the stimulus (e.g. Hennig-Thurau & Walsh, 2004).

(20)

names of the hotels and reviewer characteristics are deleted from the review. To control for individual differences and personal web experiences respondents will be randomly assigned to the survey.

In this study it is also necessary to control for other variables. It is important that all respondents know what hotel reviews are in order to secure validity of answers. Therefore questions on familiarity with hotel reviews are asked.

Research method. Independent samples t-tests are used to show how ‘purchase intention’ is influenced by ‘review quality’ , ‘attribute-centric reviews’, ‘benefit-centric reviews’ and

(21)

4

Results

The results of the studies are divided into 2 parts. The first part shows the regression models, were the influence of ‘review quantity’, ‘review variance’ and ‘total average score’ on ‘sales

last 30 days’ is estimated (paragraph 4.1). With the results of these models it is possible to

answer hypotheses 2, 3 and 4. The second part shows the results from the independent samples t-test were the influence of ‘review quality’, ‘attribute-centric reviews’ and

‘benefit-centric reviews’ on ‘purchase intention’ is measured (paragraph 4.2). Besides, using a

median-split and independent samples t-test, the results of the moderating effect of ‘consumer

expertise’ are discussed.

4.1 Results of study I

First general results will be presented. Table 3 shows the descriptive statistics, table 4 a correlation analysis of the independent and dependent variables and table 5 a regression analysis of the independent- dependent and control variables.

N Minimum Maximum Mean Std. Deviation

Average price 40 60.00 494.46 163.30 104.11

Sales last 30 days 40 73.00 770.00 368.28 145.44

Total average score Range of grades: 1-10

40 5.30 8.60 7.09 0.89

Review quantity 40 9.00 446.00 57.93 83.23

Variance average score 40 .14 3.17 1.26 0.87

Table 3: Descriptive statistics

Pearson Correlation Sales last 30 days

Sales last 30 days 1.00

Review quantity .53

Review variance -.42

Total average score .27

Sig. (1-tailed)

Sales last 30 days .

Review quantity .00

Review variance .01

Total average score .04

(22)

R R Square Adjusted R Square

Std. Error of the Estimate

.73 .53 .44 108.61

Table 5A : Model Summary; average price, city, star rating, review quantity, review variance, total average score. Sum of squares df Mean Square F Sig. Regression 435724.45 6 72620,74 6.16 0.00 Residual 389259,52 33 11795.74 Total 824983,98 39 Table 5B : ANOVA

Table 5A shows that 53% of the variability in sales can by accounted by the independent variables. The ANOVA results in table 5B shows that the correlation results of table 5A are statistically significant with a p-value of 0.00.

Review quantity. We propose that ‘review quantity’ positively affects sales. To discover whether there is enough support for this hypothesis a simple regression analysis is performed.

‘Sales last 30 days’ was the dependent variable in this model, and ‘review quantity’ the

independent variable. The following model illustrates the first main relation of the conceptual model:

Model 1: Y = α + ß1*X1 + εi

Y Sales last 30 days α Constant

ß1 Regression coefficient for review quantity

X1 Review quantity variable

εi Residual term

To make sure that the location of the hotel (city), star rating of the hotel and the average price do not explain the effect of the quantity of reviews on sales these variables are also added in the model. Pearson Correlation (table 4) showed that review quantity and sales are positive correlated (.53).

(23)

Addition of a quadratic term for review quantity allows to look for non-linear effects. Results show an R Square value of 0.27 and a p-value of 0.20 (t=-1.30) for the quadratic term of review quantity. Because this value is not significant we can conclude that there exists a linear relation between review quantity and sales. (see appendix V for SPSS data).

Review variance. In chapter two, there is hypothesized that ‘review variance’ negatively affects sales. To discover whether there is enough support for this hypothesis a simple regression analysis is performed. ‘Sales last 30 days’ was the dependent variable in this model, and ‘review variance’ the independent variable. The following model illustrates the first main relation of the conceptual model:

Model 2: Y = α + ß1*X1 + εi

Y Sales last 30 days α Constant

ß1 Regression coefficient for review variance

X1 Review variance variable

εi Residual term

To make sure that the location of the hotel (city), the star rating of the hotel and the average price do not explain the effect of the variance of reviews on sales these variables are also added in the model. Pearson Correlation (table 4) showed that review variance and sales are negative correlated (-.42).

The results of the simple regression analysis show an R Square value of 0.25 which entails that 25% of the variation sales last 30 days is explained by the variance of reviews. Furthermore, because of the p-value of 0.03 (t=-3.29) we can conclude that there is a

significant correlation between both variables. The constant show 620,020 sales over the last 30 days when there is no variance in reviews. Unstandardized coefficients showed a B-value of -76.75 for review variance. An increase of review variance with 1, results in an decrease in sales of 77. To conclude this section, it could be said that hypothesis 3 is accepted: the variance of reviews have a negative effect on sales (see appendix V for SPSS data).

Review valence. In chapter two, there is hypothesized that ‘review valence’ positively affects sales. To discover whether there is enough support for this hypothesis a simple regression analysis is performed. ‘Sales last 30 days’ was the dependent variable in this model, and

‘Total average score’ the independent variable. The following model illustrates the first

(24)

Model 3: Y = α + ß1*X1 + εi

Y Sales last 30 days α Constant

ß1 Regression coefficient for total average score

X1 Total average score variable

εi Residual term

To make sure that the location of the hotel (city), the star rating of the hotel and the average price do not explain the effect of the total average score of reviews on sales these variables are also added in the model. Pearson Correlation (table 4) showed that total average score and sales are positive correlated (.27).

The results of the simple regression analysis show an R Square value of 0.16 which entails that 16% of the variation sales last 30 days is explained by the total average scores of reviews (review valence). Furthermore, because of the p-value of 0.04 (t=-2.38) we can conclude that there is a significant correlation between both variables. The constant show 60.45 sales over the last 30 days when the total average score of reviews is zero.

Unstandardized coefficients showed a B-value of 66.45 for total average score. An increase of reviews’ total average score with 1, results in an increase in sales of 66. To conclude this section, it could be said that hypothesis 4 is accepted: the valence of reviews has a positive effect on sales (see appendix V for SPSS data).

4.2 Results of study II

176 respondents participated voluntarily. Their average age was 45 and the distribution of men and woman was 38,6% and 61,4% (see appendix III for sample statistics). Primarily, question 1 (familiarity with hotel reviews) is used to control for respondents who are familiar with hotel reviews. 176 respondents answered that they were familiar with hotel reviews and 13 respondents answered that they were not. Because 13 respondents were not familiar, these respondents are deleted from the dataset.

Familiar with hotel reviews? Frequency Percent

Yes 176 93.10

No 13 6.90

Total 189 100.00

Table 6: Frequency of familiarity with hotel reviews (question 1)

(25)

of internal consistency. A Cronbach’s alpha shows whether all questions measure the same construct; in this case it concerns the quality (six items) and expertise (five items) constructs. Literature tells that the questions will be internally consistent if the Cronbach’s alpha shows a value of above 0.6 (e.g. Malhotra, 2007). In table 8, an overview of all constructs with the corresponding Cronbach’s alpha is given.

Construct Cronbach’s alpha N of items

Review quality 0.83 6

Consumer expertise 0.75 5

Table 7: Reliability statistics of the two constructs

It appears that both constructs have a Cronbach’s alpha of above 0.6, which indicate a strong internal consistency. Therefore, all single items are used to measure the dimensions.

Review quality. We proposed that that ‘high quality reviews’ have a stronger influence on

‘purchase intention’ than ‘low quality reviews’. To discover whether there is enough support

for this hypothesis an independent samples t-test is performed. ‘purchase intention’ was the dependent variable in this model, and the ‘high quality review group’ and ‘low quality review

group’ the independent variable.

The mean of the items was used to check whether the quality of reviews were manipulated as was intended. The six items of the high quality review showed an average mean of 5.79 and the six items of the low quality review showed an average mean of 5.05. Besides, equal variances were assumed with a p-value of 0.00. Therefore, there can be stated that the reviews were manipulated the right way.

The purchase intention mean show that the purchase intention of respondents who read a high quality review is stronger influenced than the purchase intention of a respondent who read a low quality review.

N Mean – Purchase intention Std. Deviation Std. Error Mean Purchase intention – High quality 81 4.78 2.78 2.76 Purchase intention – Low quality 95 4.23 2.37 2.37

(26)

High quality respondents show a purchase intention mean of 4.78 and low quality

respondents 4.23. An independent samples t-test can show if the purchase intention of both groups is different influenced regarding to the quality of reviews.

To test whether high quality reviews have a stronger effect on purchase intention than low quality reviews the two experimental groups are compared.

Test for equality variances

F Sig. T Sig. (2-tailed) Mean

difference Equal variances assumed 3.41 .07 -1.41 .16 -.55 Equal variances not assumed -1.39 .17 -.55

Table 9: T-test for equality variances

An independent samples t-test can show if the purchase intention of both groups is different influenced regarding to the quality of the review. The independent samples T-test showed with Levene’s test for equality of variances a p-value of 0.07. When we look at a sig. (2-tailed) of 0.16 there can be concluded that there are no differences between the two groups when we use a significance level of 0.05. To conclude this section, it could be said that hypothesis 1 is not accepted: high quality reviews have not a stronger effect on purchase intention than low quality reviews.

Consumer expertise. In chapter two, there is hypothesized that ‘attribute-centric reviews’ have a stronger influence on ‘purchase intention’ of consumers with ‘high expertise’ than consumers with ‘low expertise’. To discover whether there is enough support for this hypothesis an independent samples t-test is performed. ‘Purchase intention’, in response to the attribute-centric review, was the dependent variable in this model. The ‘expert group’ (51 respondents) and ‘non-expert group’ (44 respondents) were the independent variables / grouping variables.

(27)

and 7.00 (experts).

The purchase intention mean show that the purchase intention of non-expert respondents is stronger influenced by attribute-centric reviews than purchase intention of experts. N Mean – Purchase intention Std. Deviation Std. Error Mean Purchase intention -Expert 51 3.82 2.66 .37 Purchase intention -Non expert 44 4.05 2.38 .36

Table 10: Mean purchase intention expert vs. non-expert (attribute-centric reviews)

Non-experts show a purchase intention mean of 3.82 and experts 4.05. An independent samples t-test can show if the purchase intention of both groups is different influenced regarding to the degree of expertise.

Test for equality variances F Sig. T Sig. (2-tailed) Mean difference Equal variances assumed 1.81 .18 -.43 ,.67 -.22 Equal variances not assumed -.43 .67 -.22

Table 11: T-test for equality variances

The independent samples T-test showed with Levene’s test for equality of variances a p-value of 0.18. When we look at a sig. (2-tailed) of 0.67 there can be concluded that there are no differences between the two groups when we use a significance level of 0.05. To conclude this section, it could be said that hypothesis 5 is not accepted: consumer expertise does not influence the effect of attribute-centric reviews on purchase intention.

In chapter two, there is hypothesized that ‘benefit-centric reviews’ have a stronger influence on ‘purchase intention’ of consumers with ‘low expertise’ than consumers with ‘high

expertise’. To discover whether there is enough support for this hypothesis an independent

samples t-test is performed. ‘Purchase intention’, in response to the benefit-centric review, was the dependent variable in this model. The ‘expert group’ (42 respondents) and

‘non-expert group’ (39 respondents) were the independent variables / grouping variables.

(28)

N Mean – Purchase intention Std. Deviation Std. Error Mean Purchase intention - Expert 42 3.90 3.04 .47 Purchase intention -Non expert 39 3.15 2.03 .33

Table 12: Mean purchase intention expert vs. non-expert (benefit-centric reviews)

Non-experts show a mean of 3.15 and experts 3.90. An independent samples t-test can show if the purchase intention of both groups is different influenced regarding to the degree of expertise.

Test for equality variances F Sig. T Sig. (2-tailed) Mean difference Equal variances assumed 10.01 .00 1.30 .20 .75 Equal variances not assumed 1.32 .19 .75

Table 13: T-test for equality variances

(29)

4.3 Summary of findings

Based on the results of the tests in paragraph 4.1 and 4.2 the hypotheses are empirically tested. The following table gives an overview: empirically tested. The following table gives an overview:

Table 14: Summary findings


 
 
 
 
 
 
 


Hypotheses Supported / not supported

H1: Review quality affects purchase intention positively. Not supported

H2: The quantity of reviews positively affects sales. Supported

H3: The higher the variance of reviews the lower the sales Supported

H4: High valence of a review has a positive effect on sales Supported

H5: Attribute-centric reviews have a stronger effect on the purchase

intention of consumers with high expertise than consumers with low expertise.

Not supported

H6: Benefit-centric reviews have a stronger effect on purchase

intention of consumers with low expertise than consumers with high expertise.

(30)

5

Conclusion

This final chapter describes the way online retailers should manage their online reviews in order to influence sales and purchase intention. In paragraph 5.1 a conclusion regarding the results is displayed followed by limitations and possibilities for future research in paragraph 5.2.

5.1 Main conclusion and recommendation

This study investigated the effect of consumer reviews (review characteristics) on purchase intention and sales. The results of this study show that it is important for online retailers to show consumer reviews on their website and allow consumers to write reviews on their products. Results of this study help online retailers to manage their online strategy with respect to consumer reviews.

Previous studies stated that the quality of reviews have an effect on purchase intention in such a way that high quality reviews influence purchasing intention stronger than low quality reviews (e.g. Su et al. , 2008; Forman, Ghose & Wiesenfeld, 2008). Purchase intention means showed higher means for respondents who saw a high quality review than consumers who saw a low quality review. In despite of these positive means, other tests doesn’t show significant results for the effect of review quality on purchase intention. For managers it is important to notice that, as long as reviews are positive, it doesn’t matter for consumers or the quality of a reviews is high or low.

Consumers use word-of-mouth as a risk-reduction strategy to prevent negative results of a future purchase. The more reference possibilities the consumer have, the more

consumers can reduce their uncomfortable feeling of risk exposure. Regarding the online setting of consumer reviews as reference, there can be stated that the more consumer reviews are available the more the uncomfortable feeling of risk exposure can be reduced. Previous literature showed that less risk of purchase will lead to a positive effect on sales (e.g. Petty & Cacioppo, 1984). This study also showed that the quantity of reviews have a positive effect on sales. For managers it is therefore important to ask consumers to write a product review on their purchase so the quantity of reviews will increase, which will positively affect sales.

(31)

review structures ensures that consumers have no possibility to come up with self-invented review categories. If every consumer review the product on the same categories the degree of variation will decrease which will have a positive effect on sales.

Previous reviews showed that the impact of review valence on sales is mixed (e.g Chevalier & Mayzlin, 2006; Liu, 2006, Duan et. al, 2005; Zhang & Dellarocas, 2006). This study showed that review valence has a positive effect on sales. For managers it is therefore important to show the average score of all reviews for that product. The higher the average score of a product the stronger sales will be affected.

This study also implies to investigate the effect of review type; attribute-centric and benefit-centric reviews on purchase intention. Besides, also the moderating effect of

consumer expertise was investigated. Expected was that purchase intention for consumers with high expertise is stronger influenced by attribute-centric reviews than for consumers with low expertise. Also was expected that purchase intention for consumers with low expertise is stronger influenced by benefit-centric reviews than for consumers with high expertise. Previous studies on Elaboration Likelihood and the effects of review type supported this expectation (e.g. Alba & Hutchinson, 1987; Petty & Cacioppo, 1984). However, this study does not find any significant relations between the effect of the two review types on purchase intention, which is moderated by consumer expertise. For managers this is an important finding. Integrating a system in which reviews are sorted based on the expertise of consumers can be a costly investment, which not will contribute in influencing purchase intention. Because an investment of a expertise based review system will not achieve the desired results it is not wise to implement such a system.


5.2 Limitations and future research

With respect to this study there are some limitations. First, we limit our investigation to one product, hotel stays. Because only one product is used it is not possible to generalize the results for different product categories.

Second, in study II only positive reviews were used. The reason only positive reviews were used was to control for side effects and influences on attitude and expertise. In general, consumers read both positive and negative reviews and therefore this is something to take into account in future research.

(32)

Fourth, all results of the survey were not significant. In future research we will recommend to assign more respondents so there can be looked at more extreme answers to get significant results.

Fifth, despite our attempt to look for non-linear effects of review quantity we did not find significant results. Because it is interesting to know if there is a non-linear relationship between review quantity and sales we would like to advice to use this in future research.

In order to close this final chapter, there can be stated that consumer reviews have an effect on purchase intention and sales. This study showed that some characteristics of reviews affect purchase intention and sales stronger than other effects. In spite of the non-significant results of the moderating effect of consumer expertise is it advisable to research this effect in other studies, with other products and with other experimental designs.

(33)

References

ABN AMRO (2009). Hoe reizen we verder – ABN AMRO. http://twinklemagazine.nl/nieuws-/2009/03/ABN_Amro_Een_op_vijf_reisbureaus_verdwijnt/Travel_rappport_internet.pdf ABN AMRO & CBW-MITEX, (2011) De consument van 2015 – Cross channel retail.

http://www.cbwmitex.nl/websites/cbwmitex/files-/Actueel/3129_ABN_CrossChannelRetail_v3.pdf

Alba, J.W., & Hutchinson, J.W. (1987). Dimensions of consumer expertise. Journal of

Consumer Research, 13, 411-454.

Alba, J., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, A., & Wood, S. (1997). Interactive home shopping: Consumer, retailer and manufacturer incentives to participate in electronic marketplaces. Journal of Marketing. 61, 38-53.

Bailey, J.E., & Pearson, S.W. (1983). Development of a tool for measuring and analyzing computer user satisfaction. Management Science. 29 (5), 530-545.

Bickart, B., & Schindler, R.M. (2001). Internet forums as influential sources of consumer information. Journal of Interactive Marketing. 15 (3), 31-40.

Buttle, F.A. (1998) Word of mouth: understanding and managing referral marketing. Journal

of Strategic Marketing. 6 (3), 241 – 254.

Chatterjee, P. (2001). Online review: do consumers use them? Advances in Consumer

Research. 28, 129-133.

Chavalier, J., & Mayzlin, D. (2006). The effect of word of mouth on sales: online book reviews. Journal of Marketing Research. Working Paper, 1-43.

Chen, Y. & Xie, J. (2005). Third-party product review and firm marketing strategy.

Marketing Science. 24 (2), 218-240.

Chen, Y. & Xie, J. (2004). Online consumer review: Word of mouth as a new element of marketing communication mix. Management Science. 54 (3), 477-491.

(34)

Clemons, E.K., Gao, G.G., Hitt, L.M., When online reviews meet hyper differentiation: a study of the craft beer industry. Journal of Management Information Systems. 23 (2), 149-171.

Dellarocas, C., (2003). The digitization of word of mouth: promise and challenges of online feedback mechanisms. Management Science, 49 (10), 1401-1424.

Duan, W., Gu, B., & Whinston, A. (2005). Do online reviews matter? An empirical investigation of panel data. Decision Support Systems. 45 (4), 1007-1016.

Douglas, S.P. & Wind, Y. (1971). Intentions to buy as predictions of buying behavior.

Proceedings of the second annual conference of the association for consumer research.

331-343.

Ellsberg, D. (1961). Risk, ambihuity, and the savage axioms. Quaterly Journal of Economics. 75, 643-69.

Fiske, S.T. (1980). Attention and weight in person perception: the impact of negative and extreme behavior. Journal of Personality and Social Psychology, 38 (6), 889-906.

Forman, C., Ghose, A., & Wiesenfeld, B. (2008). Examining the relationship between reviews and sales: the role of reviewer identity disclosure in electronic markets. Information

Systems Research, 19 (3), 291-313.

Gretzel, H., & Yoo, K.H. (2008). Use and impact of online travel reviews. Information and

communication technologies in tourism. 2, 35-46.

Hennig-Thurau, T., & Walsh, G. (2004). Electronic word of mouth: Motives for and consequences of reading customer articulations on the internet. International Journal of

Electronic Commerce, 8 (2), 51-74.

Hong, W., Thong, J.Y.L, & Tam, K.Y. (2004). The effects of information format and shopping task on consumers’ online shopping behavior: A cognitive fit perspective. Journal

of Management Information Systems, 21, 149-184.

Hu, N., Liu, L., & Zhang, J. (2008). Do online reviews affect product sales? The role of reviewer characteristics and temporal effects. Information Technology and Management. 9 (3), 1-33.

(35)

Juster, F.T. (1966). Consumer buying intentions and purchase probability: an experiment survey design. Journal of the American Statistican Association. 61 (315), 658-696. Kempf, D.S, Palan, K.M. (2006). The effects of gender and argument strenght on the

processing of word-of-mouth communication. Academy of Marketing Studies Journal. 10 (1), 11-16.

Kerstetter, D., Cho, M.H. (2004). Prior knowledge, credibility and information search.

Annals of Tourism Research. 31 (4), 961-985.

Lee, S.H. (2009). How do online reviews affect purchase intention? Journal of Business

Management. 3 (10), 576-581.

Lee, J., Park, D.H., & Han, I. (2007). The effect of negative online consumer reviews on product attitude: an information processing view. Electronic Commerce Research and

Applications. 7, 341-352.

Liu, Y. (2006). Word of mouth for movies: Its dynamics and impact on box office revenue.

Journal of Marketing, 70 (3), 74-89.

McKinsey Global Institute (2011). The great transformer: The impact of the Internet on

economic growth and prosperity. http://www.mckinsey.com/~/media/McKinsey/dotcom

/Insights%20and%20pubs/MGI/Research/Technology%20and%20Innovation/The%20great %20transformer/MGI_Impact_of_Internet_on_economic_growth.ashx.

Nielsen Company (2010). Global Trends in online shopping. A Nielsen global consumer

report. http://hk.nielsen.com/documents/Q12010OnlineShoppingTrendsReport.pdf.

Malhotra, N.K., Birks, D.F. (2007). Marketing Research. An Applied Approach. Pearson/Prentice Hall.

Meyer, R.J. (1981). A model of multiattribute judgments under attribute uncertainty and informational constraint. Journal of Consumer Research, 8, 428-41.

Mudambi, S., & Schuff, D. (2010). What makes a helpful online review? A study of customer reviews on Amazon.com. MIS Quaterly, 34 (1), 185-200.

(36)

Park, D., H., & Kim, S. (2008). The effects of consumer knowledge on message processing of electronic word-of-mouth via online consumer reviews. Electronic Commerce Research

and Applications. 7, 399-410.

Park, D., H., Lee, J., & Han, I. (2007). The effect of on-line consumer reviews in consumer purchasing intention: the moderating role of involvement. International Journal of Electronic

Commerce. 4, 125-148.

Park, D.H., & Park, S.B. (2011). How consumers incorporate variance in conflicting online consumer reviews into judgments: the role of causal attributions.

http://ybri.yonsei-.ac.kr/app/tools/download.html


Petty, R. E., & Cacioppo, J. T. (1981). Attitudes and persuasion: Classic and contemporary approaches. Dubuque, IA: William C. Brown.

Petty, R.E., & Cacioppo, J.T. (1984). The effects of involvement on response to argument quantity and quality: central and peripheral routes to persuasion. Journal of Personality and

Social Psychology. 46 (1), 69-81.

Petty, R.E., & Cacioppo, J.T. (1986). Communication and persuasion: central and peripheral routes to persuasion. Springer-Verlag, New York.

Petty, R.E., Cacioppo, J.T. & Schumann, D. (1983). Central and peripheral routes to advertising effectiveness: the moderating role of involvement. Journal of Consumer

Research. 10 (2), 135-146).

Ratchford, B.T., Talukdar, D., & Lee, M.S (2001). A model of consumer choice of the

Internet as an information source. International Journal of Electronic Commerce, 5 (3), 7–22. Shen, W. (2008). Essays on online reviews: the relationships between reviewers, reviews, and product sales, and the temporal patterns of online reviews. Krannert School of Management,

Phd Report.

Su, Q., Li, Z.,& Ting, C. (2008). Conceptualizing consumer’s perceptions of e-commerce quality. International Journal of Retail and Distribution Management. 36 (5), 360-374. Westbrook, R.A. (1987). Product / consumption-based affective responses and post purchase processes. Journal of Marketing Research, 24-3, 258-270.

(37)

Zhang, X., Dellarocas, C. & Awad, N.F. (2004). Estimating word-of-mouth for movies: the impact of online movie reviews on box office performance. Workshop on Information

Systems and Economics, College Park MD.

Zhu, F., & Zhang, M. (2010). Impact of online consumer reviews on sales: The moderating role of product and consumer characteristics. Journal of Marketing, 74 (2), 133-148.

(38)

Appendices

Appendix I Survey I Appendix II Survey II

Appendix III Respondents sample statistics Appendix IV Overview hotel dataset Appendix V Results study I

Appendix VI Results of study II

(39)

Appendix I

Survey I

This survey is designed to test the effects of high quality and attribute-centric reviews on purchase intention.

Fill in this survey and win a BOL.COM gift voucher of 25 euro!

Thank you for your willingness to fill in this survey. By answering the questions you contribute to academic research towards consumer reviews. The results of this study will be used to write my master thesis on consumer reviews at the University of Groningen. It will only take 5 minutes to answer all questions. I would really appreciate if you dedicate your time and answer the following questions precisely. All the answers to the questions are anonymous and confidential.

If you have comments, questions about this research, or want to participate in the lottery for the BOL.COM gift voucher please leave your email address in the last comment box. Thank you very much in advance

---

The following questions are about hotel reviews. Hotel reviews are evaluations from anonymous consumers, who stayed at a hotel. In these evaluations consumers describe what they perceived as positive and negative during their stay at the hotel. Next to the short messages consumers can evaluate the hotel with a grade.

(1) Are you familiar with hotel reviews written by other consumers? * Yes

* No

Please indicate how much you agree with the following statements about hotel reviews. 1 means that you strongly disagree, while 7 means that you strongly agree.

(2) When I book a hotel through the website of an online travel agency, I always use reviews that are presented on the website of that agency.

Strongly disagree * * * * * * * Strongly agree (3) When I book a hotel through the website of an online travel agency, the reviews presented on the website of that agency are helpful for my decision-making.

Strongly disagree * * * * * * * Strongly agree

(4) When I book a hotel through the website of an online travel agency, hotel reviews on the website of that agency make me confident in booking a hotel.

Strongly disagree * * * * * * * Strongly agree

(40)

book a hotel online, I worry about my decision.

Strongly disagree * * * * * * * Strongly agree

(6) When I book a hotel through the website of an online travel agency, reading the reviews presented on the website of that agency impose a burden to me.

Strongly disagree * * * * * * * Strongly agree

(7) When I book a hotel through the website of an online travel agency, reading the reviews presented on the website of that agency impose a burden to me.

Strongly disagree * * * * * * * Strongly agree ---

Before completing the next questions I would like to ask you to read the following text and hotel review.

Suppose that you are planning to go on a city trip to New York. You already have booked your flight but you are searching for a hotel in New York. You searched across the Internet and found a hotel, which seems to fit your needs. Next to hotel information, prices and

availability – the booking website offer hotel reviews from consumers who stayed at the hotel. Before you decide to book the hotel you read the following hotel review:

Referenties

GERELATEERDE DOCUMENTEN

After formulating the diffusion systems by regarding the noisy observation data near the systems boundary region as the system’s boundary inputs, we derive the Kalman filter and

e moderator effect of ADHD was only signi�cant for the teacher data; children with higher levels of ADHD showed a weaker relationship between peer problems and prosocial

autumn auctions in October and November.. To compute average prices, the total sum of all auction records within a specific half-year period before or after an

One possible explanation for this is that conscientiousness positively predicted motivation to learn, which in turn was related to learning outcomes (Colquitt et

[r]

The main objective of this research was to understand the relationship between CSR and sales and the moderating effects of brand equity and GDP on the relation between CSR and

As the results of most of the prior studies that were discussed showed that online reviews have a positive effect on sales or willingness to pay (e.g. Wu et al., 2013; Kostyra et

III) the combination of online and offline advertisement on sales. H3b: The age of different consumer groups has a moderating effect on H3c: Income differences have a