Customer Engagement Differences in the effect of reviews and/or samples on product attitude per product type

(1)

Customer Engagement

Differences in the effect of reviews and/or samples on

product attitude per product type

Laura Sneeboer

Marketing Department University of Groningen

Master Thesis

23-06-2014

Oosterhamrikkade 96A

9714BJ Groningen

(06) 27274891

laurasneeboer@hotmail.com

Student number S2400316

Supervisor: dr. Jenny van Doorn

Faculty of Economics and Business

(2)

2

Abstract

This research investigates how samples can complement reviews in the online purchasing environment. It also looks at whether the effects of samples and/or reviews differ per product type which are categorized as search and experience goods. Since the effects of positive or negative reviews are already known, this study focuses on the differences of these known effects for search and experience goods. Moreover, whether the effect of a sample on product attitude is stronger for search or experience goods is also examined. It was found that samples have the ability to complement reviews and allow consumers to form a more positive attitude towards the product. It was expected that samples and reviews would be of more importance for experience goods because of the inability to evaluate its performance before purchase, however this was not supported. The contrary was found, samples and reviews are of more importance for search goods instead of experience goods.

(3)

3

1. Introduction

Consumers can nowadays easily exchange and access information and opinions about companies and products (Godes and Mayzlin, 2004). Before the introduction of the world wide web consumers could primarily gain their information through interpersonal word-of-mouth communication within limited social contact boundaries (Duan, Gu and Whinston, 2008) or through their own experience. Today, the world wide web is an important source of product information and online word-of-mouth has become more important in consumer purchase decisions (Duan, Gu and Whinston, 2008). Over the years it has become relatively easy for consumers to post their opinions online, and share them with others (Chen, Liu and Zhang, 2011). Through these reviews consumers can share their personal experiences with the product. Bart, Shankar, Sultan and Urban (2005) state that “customers increasingly rely on the internet for information”. This is supported by figures, as in the Netherlands 9,8 million people have ever shopped online (CBS, 2013). The amount of frequent online shoppers has increased from 3,9 million in 2005 to 7,1 million in 2012 (CBS, 2013). Which indicates an increase in the amount of people searching for information about the product beforehand by reading reviews for example. When purchasing a product through the internet there is no option of touching or experiencing the product before buying it. For that reason, third-party reviews have become an important source of information in assessing the quality of the product beforehand (Chen and Xie, 2005). However, not only reviews allow consumers to gain information about a product or service. Sampling is also a very popular tool to let consumers try a product before purchase (Heiman, McWilliams, Shen and Zilberman, 2001). Companies spend billions of dollars annually on product sampling (Wadhwa, Shiv, and Nowlis, 2008). Samples are found to have a greater effect on sales than advertising because they serve as a direct source of information (McGuinness, Dalton, Philip and Mathew, 1992). More importantly, since sampling allows the consumers to directly experience the product it can reduce the risk of product uncertainty (Heiman, McWilliams, Shen and Zilberman, 2001). In the online environment this direct experience is not possible for all products, but consumers can for example read the first couple of pages of a book to directly experience it. The increase in online purchases could be due to the fact that more tools (reviews and samples) are available for consumers to better evaluate the product before purchase to reduce their uncertainty.

(5)

5 and strong brands differ across maturing and emerging product categories. They found that customer reviews can increase the sales of weak brands and help them to become strong, although this effect does not occur to the same extent once the brand had become strong. In addition, reviews matter less for strong brands. Also, in mature categories it is possible for new or weak brands to use reviews to increase sales. Zhang, Ma and Cartwright (2013) looked at the impact of online user reviews on sales of search goods, specifically cameras. They found that there is a significant relationship between online user reviews and sales of search goods. Moreover, the influence of reviews on search good sales is different from that on experience goods. This indicates that reviews are important online attributes that consumers consult on a regular basis.

Several websites offer samples in addition to reviews. This combination can be found for a couple of products like, books, DVDs and CDs. A study by Biswas, Labrecque, Lehmann and Markos (2014) mentions that “the increased availability of sampling opportunities is largely driven by companies that find giving free samples are more powerful and cheaper than alternative traditional forms of advertising”. This can be translated to the online setting as well. Providing a sample of a book for example, is cheaper than having to advertise what the book is all about. Moreover, advertising what the book is all about can be a very difficult task as well. Traditionally, sampling has been described as an excellent way to introduce new and unusual products, change the image of a product and generate word-of-mouth (Marks & Kamins, 1988). By offering samples in addition to reviews consumers can evaluate the quality of a product better. A small preview of the book, song or album can help a consumer determine the quality more easily and reduce their uncertainty. A study by Smith (1993) found that attitudes which are formed through direct experience (product trial) are stronger, more confidently held, and better predictors of behavior than those formed through indirect experience such as advertising. This is another argument how samples can influence attitude towards the product positively. Marketers nowadays are confronted with the difficult task of deciding when to display reviews to improve consumer’s attitude/ increase purchase intention or when to use both reviews and samples. In addition, they have to decide upon for which type of products which tool works the best and has the best potential for sales increase. This leads to the following research question. Can samples complement reviews to influence consumer’s attitude towards the product? Does this differ for search and experience goods?

(6)

6 kind of sample that can be shown for a camera are pictures of how the camera looks. But does that really constitute a sample similar to that of a book or CD? Moreover, is the effectiveness of a sample different per product type? In the literature many different product type classifications can be found. This paper will use the classification of search and experience goods. Several definitions of search and experience goods are available but in this paper the definition by Nelson (1970) will be used who states that ‘the classification of search and experience goods is based on consumers’ ability to discover product quality before purchase’. Search and experience goods are used because according to Norton and Norton (1988) this classification relies primarily on the fundamental attributes of the product itself and not the buyer’s perceptions of it, like with the low-involvement/ high-involvement product categorization. A study by Huang, Lurie and Mitra (2009) examined shoes and home furniture as search goods and cameras and automotive parts as experience goods. However, in this study we will be looking at products that also have samples available online. Some examples are books, CDs and DVDs. These are all examples of experience goods based on Nelson’s definition since the quality of these products cannot be discovered before purchase without a sample.

However, given that information search costs differ across channels, a search good through one channel may be an experience good through another channel (Weathers, Sharma & Wood, 2007). Consider buying a book in a brick and mortar store. Theoretically, you can read the entire book there. Read the first couple of pages and the last for example to get a good feeling what it is about and shape your own opinion about the product. This way you can easily determine the quality before purchasing it, which means it is a search good according to Nelson (1970). However, in the online environment you are not able to do so which makes a book an experience good in the online channel. Since this paper is only discussing the effect of reviews and samples in the online environment it will therefore focus on the online channel which makes a book an experience good. An example of a search good in the online environment could for instance be a phone application (app). Before purchasing an app you can see some technical product specifications such as the category it belongs to, version, size, compatible devices and an extensive description of what it is for and how it can be used. Reviews are also available before purchasing an app with a star ranking possibility. Samples are also included by showing some screenshots of the app. All these features allow the consumer to evaluate the quality of the app before purchase, which is especially important when one has to pay for the app. This leads to the following research questions:

Does the effect of samples on attitude towards the product depend upon the type of product? Is there a difference in the effectiveness of a sample for search and experience goods?

(7)

7

1.1 Contribution

A lot of research investigated the relationship between online consumer generated reviews on purchase intention and on conversion rates (Ludwig, de Ruyter, Friedman, Brüggen, Wetzels and Pfann, 2013). To the best of my knowledge, no research has looked at the effect samples, in addition to consumer generated reviews, have on consumer’s product attitude. However the study by Ludwig, de Ruyter, Friedman, Brüggen, Wetzels and Pfann (2013) did use samples in their investigation but they measured conversion rates instead of attitude towards the product. When the effect of a review is not strong enough to persuade consumers to buy the product, a product sample might be. In addition, this study will investigate if there is a difference in the effectiveness of using samples for search and experience goods. Therefore this study primarily contributes to existing literature by adding the possibility of having a sample of that product available online and how much it will influence a consumer. This study will examine the effects negative or positive reviews with a sample have on attitude towards the product. For practitioners the insights this study may give can be useful for them in order to decide when it pays off to include samples as well. Or when only reviews are enough, this allows them to not waste any money on unnecessary investments for samples. It can help to specifically attribute samples only to those products where it will help to persuade consumers to actually buy the product. Perhaps only having reviews available and product detailed specifications is already enough information consumers want for a search good. Hopefully this study can provide relevant insights into this.

(8)

8

2. Literature review

2.1 Consumer generated reviews

According to Godes and Mayzlin (2004) one of the most frequently accessed online information sources are customer reviews. Online consumer-generated reviews can help consumers by providing them with additional information about usage, reliability and overall satisfaction of the product. Chevalier and Mayzlin (2006) investigated the effect of book reviews on sales for both Amazon.com and bn. com. They found that “customer word of mouth affects consumer purchasing behavior for both retail sites”. To add, they found that favorable reviews can increase the sales of a book and that consumers actually read and respond to reviews and do not only look at the average star-ranking. More benefits of reviews are explained by Holmes and Lett (1977) who said: “Product-related information transmitted by satisfied customers provides the double benefit of accelerating the brand’s acceptance and reducing the firm's promotional expenditures”.

An important component of reviews is the valence of it, how positive or negative a review is (Basuroy, Chatterjee and Ravid, 2003). Studies by Basuroy, Chatterjee and Ravid (2003) and Tsang and Prendergast (2009) found that negative reviews have a bigger negative effect on purchase intention than positive reviews have a positive effect on purchase intention. Stated by Tsang and Prendergast (2009) “the harmful impact of negative information is much greater than the beneficial impact of positive information”. This implies that negative online reviews can hurt sales more than positive reviews can increase sales. Chen, Liu and Zhang (2011) also found that the review valence is important as they looked at the impact of third- party reviews on firm value and found that reviews can exert a big impact on stock returns. This impact is caused by the valence of a review relative to that of another, previously published review instead of the absolute valence of the review itself. Negative reviews can hurt, cause bad publicity and eventually reduce purchase intention (Huang and Chen 2006, Tybout, Calder and Sternthal, 1981). While at the same time negative publicity can sometimes have positive effects (Berger, Sorensen and Rasmussen, 2010).

(9)

9 Table 1: Previous studies about reviews

2.2 Samples and direct experiences

According to Freedman (1986) sampling is a tool used to introduce consumers to a new product. Bettinger, Dawson and Wales (1979) mention it is also used to introduce an existing product to a new consumer group. In the current literature a lot of research has looked at traditional sampling, for example in a supermarket, instead of online sampling. Hu, Liu, Bose and Shen (2009) looked at the effects of CD sampling in the online environment. One of their findings is that when uncertainty from product reviews is high, sampling plays a more important role as it can reduce this uncertainty. Sampling can provide consumers with the experience of a product before purchasing it. Wang and Zhang (2009) state that “consumers need real experience with a product before they can reasonably evaluate its value. Advertising is less effective in generating this real experience”. Overby and Jap

Study Findings Goods Type

Basuroy, Chatterjee and Ravid (2003)

Negative reviews have a bigger negative effect on purchase intention than positive reviews have a positive effect on purchase intention.

Movie tickets (box office performance)

Not specified

Berger, Sorensen and Rasmussen (2010)

Negative reviews can sometimes have positive effects. Negative publicity, through negative reviews, can increase purchase likelihood and sales by increasing product awareness.

Books Experience

Chen, Liu and Zhang (2011)

Reviews can exert a big impact on stock returns. This impact is caused by the valence of a review relative to that of another, previously published review instead of the absolute valence of the review itself.

Movies Experience

Chevalier and Mayzlin (2006)

Customer word of mouth affects consumer purchasing behavior. Negative review more powerful in decreasing sales than positive review is in increasing sales.

Books Experience

Duan, Gu and Whinston (2006)

The rating of a review has no significant impact whereas the volume or amount of online reviews does have an effect.

Movies (box performance)

Experience

Mudambi and Schuff (2010)

Reviews with extreme ratings were less helpful for experience goods, the depth of a review had a positive effect on the helpfulness of the review for both product types and the length of a review in general increases its helpfulness, especially for search goods. MP3 player, music CD, PC video game cell phone, digital camera and laser printer.

Experience

(10)

10 (2009) also looked at the quality uncertainty of products in the online channel. They found that transactions done through the online channel are primarily for products with low uncertainty while high uncertainty products are more often bought through physical store. With this they indirectly imply that the online channel might not be suitable for products with high uncertainty. However, according to Roselius (1971) this could be overcome with the availability of samples. As he identified a couple of risk relievers and one of them is receiving free samples. Indicating that samples could reduce uncertainty and perceived risk. Biswas, Labrecque, Lehmann and Markos (2014) had a look at whether the order in which consumers sample the products can influence their choices. They mention this as relevant as consumers have the possibility of sampling multiple products before purchase. They found that when sampling products with similar sensory cues there seems to be a higher preference for the first sampled product. And when sampling products with dissimilar sensory cues a greater preference can be found for the product sampled last. Another study by Biswas, Grewal and Roggeveen (2010) also looked at the impact of sample sequence but then for experience goods. For example, when being in a supermarket you are able to sample and taste different products, and online you can see a few trailers of different movies before you make your decision. In their study they found serial position effects. For experience goods consumers were shown to have a better product recall for the later sampled products, recency effect. This indicates that when consumers sample two products they prefer the second sampled product. These results show how critical the order is in which consumers sample multiple products. However, marketers cannot instruct consumers to sample their product first in the online environment.

(11)

11 experiencing the product is important to consumers whereas with search goods an ad can already provide them with enough information.

The information integration theory and the integrated information response model help to explore how consumers combine information from advertising and trial (Smith, 1993). One of Smith’s findings is that “the ability of ad attitudes to influence brand cognitions and brand attitudes is significantly reduced after trial”. In other words, once a consumer has tried the product, an ad experiences more difficulties trying to influence that consumer’s attitude towards the brand. Kempf and Laczniak (2001) studied positive ad effects on product trial through an experiment. One of their main findings was that a trial was perceived more diagnostic by consumers when exposure to an ad occurred before trial. Therefore they state that “consumers who receive an ad exposure prior to product trial will form more positive purchase intentions”. In addition they show that consumers who have tried the product have a higher expected value for the product and also a higher purchase intention than consumers who have not used the product in advance.

(12)

12 Table 2: Previous studies about samples

Study Findings Goods Type

Biswas, Labrecque, Lehmann and Markos (2014)

Sampling sequence is relevant as it can influence consumer’s choices.

Beverages and music

Experience

Hu, Liu, Bose and Shen (2009)

When uncertainty from product reviews is high, sampling plays a more important role as it can reduce this uncertainty.

Digital music Experience

Kempf and Laczniak (2001)

Consumers who have tried the product have a higher expected value for the product and also a higher purchase intention than consumers who have not used the product in advance.

Soft drink Experience

Scott (1976) Providing a sample does not always lead to an increase in sales.

Newspaper Not specified Smith (1993) Once a consumer has tried the

product, it becomes more difficult to influence the consumer with an ad.

Soft drink Experience

Smith and Swinyard (1982)

Attitudes which are based upon product trial can predict purchase very well.

Cheese-filled pretzel

Experience

Wright and Lynch (1995)

Direct experience is superior in communicating experience attributes whereas advertising is superior in communicating search attributes.

Chocolate candy bar Pencils Stationary exercise bicycle Chair-bed All 4 products examined on both search and experience attributes

2.3 Search and experience goods

(13)

13 communication practices frequently used by online retailers, such as the use of vivid pictures, offering information from third-party sources and allowing consumers to control information presentation. They found that differences exist between search and experience goods as to which communication practice is more helpful in reducing uncertainty. With an experience good, retailers should primarily focus on improving the vividness of the information given, by pictures for example. Although this proved to be difficult, especially for experience goods.

Huang, Lurie and Mitra (2009) discovered that online purchasing behavior is different for search and experience goods. For example, consumers searching online for experience goods view fewer pages online but spend more time per page. However, the total amount of time consumers are searching online for product information is almost equal for search and experience goods. Moreover they found that buyers of experience goods are more likely to buy from the website where they obtained information than buyers of search goods. Buyers of search goods are more likely to engage in free-riding: purchasing a product from a vendor that is not the primary source of information (Huang, Lurie and Mitra, 2009).

The fact that online purchasing behavior differs per product type has also been confirmed by Chiang and Dholakia (2003) found that the product type has an influence on consumers to shop online. They found that for search goods, the intention to shop online is higher than for experience goods. They argue that because the search costs are reduced, search goods have a greater chance of success in the electronic environment. For experience goods is has been found that their dominant product attributes cannot be obtained prior to purchase in the online environment. This is true for quite some experience goods but not for all. Previous studies have shown that online samples of books, movies and CDs (Chevalier and Mayzlin 2006; Tu and Lu 2006; Hu, Liu, Bose and Shen 2009) can increase purchase intention as it offers consumers a chance to obtain information before purchase.

(14)

14 through the online channel or having a sample available might increase the appropriability of the online channel for experience goods. Moreover, the internet is able to offer perceptual experiences in the form of visuals that can give more and additional information besides verbal descriptions (Peterson, Balasubramanian and Bronnenberg, 1997) which might be enough information for frequently purchased experience goods. The type of product is found to be an important determinant as to whether products can and should be sold through the internet.

(15)

15

3. Conceptual model

It has already been established in the literature that positive reviews have a positive effect and negative reviews have a negative effect. This model has been chosen because so far no other study has investigated the effect of both samples and reviews on product attitude for two product types. It is expected that samples have a positive effect on product attitude because samples have the possibility to increase the conversion rate for high uncertainty products (Hu, Liu, Bose and Shen, 2010). However, this effect can be influenced by the type of product. It is expected that the effect will be stronger for experience than for search goods, since experience goods bear more uncertainty (Huang, Lurie and Mitra, 2009). The effect reviews have on product attitude is expected to be stronger for experience goods as well, compared to search goods. This is anticipated since according to Zhang, Ma and Cartwright (2013) customers do not require interaction when evaluating a search good. While evaluating the existing literature is was found that no study has looked at the effect of both reviews and samples on purchase intention and if differences exist between search and experience goods.

4. Hypotheses

Berger and Calabrese (1975) have developed the uncertainty reduction theory. This theory consists of several axioms but the third one seems applicable here. The third axiom states: “high levels of uncertainty causes increases in information seeking behavior. As uncertainty levels decline, information seeking behavior decreases”. The authors found that information seeking behavior and amount of communication are inversely related. In other words, more information seeking causes a reduction in uncertainty and less information seeking causes an increase in uncertainty.

(16)

16 received a free sample (Bawa and Shoemaker, 2004). When a consumer experiences a sample, they can easily determine whether they like it and if it will fit their needs by seeing, touching or hearing it. Traditionally buyers search for information prior to their purchase to reduce their uncertainty to a more tolerable level (Cox, 1967). Several studies show that uncertainty is related to search (Bennett and Mandell 1969; Bucklin 1966; Moore and Lehmann 1980; Reilly and Conover 1983). When searching for a specific product online it is more difficult to evaluate it compared to when looking for it in a store due to the lack of sensory elements. This is confirmed by Eggert (2006) who says that “product intangibility increases consumers’ perception of risk”. Information search functions as a risk reducer as it makes it easier for a consumer to evaluate the product. Information search can also diminish intangibility by helping the customer to grasp and visualize the product mentally. Perceived risk in the online environment is higher due to the lack of physical contact with the product and lack of face-to-face contact with a salesperson (Koernig, 2003). Therefore it is expected that due to the fact that samples can provide more product information in the online environment, it will have a positive effect on attitude towards the product.

H1: The availability of samples has a positive effect on attitude towards the product.

(17)

17 this theory it is expected that because of the inability to evaluate the quality of experience goods before purchase, these consumers posses a higher motivation to learn about the product to reduce the risk perception. Especially the ambiguity of the online information environment can prove the effectiveness of samples. Hoch and Deighton (1989) mention that many information environments provide little opportunity for learning from experience. This is particularly true in the online environment where products cannot be touched or experienced beforehand. Samples allow consumers to experience the product beforehand and allows them to learn about the product and its attributes. Moreover, when the product reviews show primarily mixed opinions, a combination of some positive and some negative points, the information becomes very ambiguous. When this is the case, samples can provide the consumer with additional information which allows them to consider whether they agree or disagree with the opinions given in the reviews. Thus, it is expected that samples have a greater effect on attitude towards the product for experience than for search goods.

H2: The effect of samples on attitude towards the product is stronger for experience goods than for search goods.

Besides learning from experience, consumers can also learn more about the product through consumer feedback. By reading product reviews consumers can learn through the experience of others, which is an important predictor of product adoption according to Chevalier and Mayzlin (2006). Since the internet can be an ambiguous environment with so many retailers offering the same or similar products, reviews can help reduce this ambiguity and uncertainty concerning purchasing in the online environment. Reviews itself can also be ambiguous, as some users describe their product experience as positive where others experience its performance as negative and some even have mixed opinions about the product. Products bought online bring more uncertainty to the table for consumers (Thompson, 2006). For consumers to ensure that the product performs accordingly they tend to spend more time and effort examining the product, if possible, before purchase (Landon and Smith 1997). Because with search goods consumers can easily asses product performance through the product information given they rely less on reviews compared to experience goods. As with experience goods they need confirmation from others that the product performs accordingly and is worth the investment. Therefore it is expected that reviews have a greater effect on attitude towards the product for experience than for search goods.

(18)

18

5. Methodology

5.1 Data collection method

This study is a between- subjects design to test for H1, H2 and H3. The subjects were presented with different scenarios each containing reviews and some a sample in addition. The data for this research was collected through the use of a questionnaire, distributed primarily via social media and email. In order to test the hypotheses, twelve different scenarios were used. A scenario either contained positive, negative, or mixed reviews. Where some had a sample in addition to the reviews and two different products were used, search and experience. (visual of data collection set-up; figure 1.) The experience good used in this study was an audio book and the search good was a - more sophisticated- calculator application for a smart phone instead of the simplified calculator already installed on a smart phone upon purchase. A pre-test among 20 people indicated that respondents significantly perceived an audio book as an experience good and a calculator app as a search good. For that reason, these two products were used in the actual questionnaire. Each of the twelve scenarios was randomly assigned to each participant. An example of the pre-test can be found in appendix 3.

Figure 1: Research set-up

(19)

19

5.2 Pre-test

In order to make sure that respondents understood that they were looking at either a search or an experience good, the pre-test used a scale from a study by Weathers, Sharma and Woods (2007). In this study participants were presented with five items on a seven-point likert scale for each good, three to assess experience qualities (It’s important for me to (1) see/ (2)experience/ (3) use this product to evaluate how well it will perform). Two to assess search qualities (I can adequately evaluate this product using only information provided by the retailer or manufacturer about the product’s attributes and features; I can evaluate the quality of this product simply by reading information about the product).

5.3 Design and procedure

This research is a 3 (valence: positive, negative, mixed) x 2 (sample: yes/no) x 2 (search/experience good) between subjects design. The reviews were written with the help of the technology acceptance model of Davis (1989). This study found two variables that are fundamental determinants of user acceptance; perceived usefulness and perceived ease of use. The author mentions that people tend to use or not use an application to the extent they believe it will help them perform their job better which they refer to as perceived usefulness. In addition, when potential users believe the application is useful they might find it is too hard to use. For that reason the author finds that usage is also influenced by perceived ease of use. Each variable is made up from 6 factors, see table 3, and these were used to construct the reviews. The reviews indicate positively or negatively how easy the product is in usage, how effective it is, how understandable etc.

Table 3: Perceived usefulness and ease of use factors

For the audio book reviews a scale by Hinterleitner, Neitzel, Möller and Norrenbrock (2011) was used which mentions a couple of factors that can help to evaluate audio text/ speech. The audio book reviews shown in this research used a couple of those factors when evaluating the audio book, namely: emotion, intonation, voice pleasantness, listening effort and overall impression. The setting of the reviews was emulated to resemble an online shopping environment, and the style of the reviews was adapted to be informal and to look like legitimate user generated reviews. The reviews

Ease of use 1 Ease of learning 2 Controllable .

3 Clear and understandable 4 Flexible

5 Easy to become skillful 6 Easy to use

Perceived usefulness 1 Work more quickly 2 Job performance 3 Increase productivity 4 Effectiveness

(20)

20 for each of the three conditions (positive, negative, mixed) were essentially identical except for the adjectives used such as great opposed to awful, effective opposed to ineffective and so on. Each scenario contained four reviews in total, 4 positive ones, 4 negative ones or 2 positive and 2 negative reviews. Moreover, neither of the reviews included information that the book was a thriller for example. This to make sure people did not responded to the question more positively (negatively) because they like (dislike) thrillers. An example of a couple of reviews can be found below.

Positive and negative reviews search good

If I had to be concise, I would just say: great app. It is an effective and easy to use tool, which made my life easier. I do more in less time. One of the things that fascinated me the most, however, was the intuitive ease of learning. I literally started to achieve results in less than an hour.

If I had to be concise, I would just say: awful app. It is not an effective and easy to use tool, which can make my life easier. I do less in more time. One of the things that “fascinated” me the most, was the lack of the intuitive ease of learning. I couldn’t get started to achieve results in less than an hour.

Positive and negative reviews experience good

Interesting book. Good voice. It is pleasant to hear and really easy to understand not even giving the full attention to. Probably the intonations of the actor make it work that good. Sometimes she gets a little too emotional for my taste, but i think it is just me being too callous. So I think it is surely worth trying.

Not an interesting book. Bad voice. It is not really pleasant to hear and sometimes difficult to understand without giving the full attention to. Probably the intonations of the actor make it work like that. Sometimes she gets a little too emotional for my taste, and i am not being callous at all. So I think it is hardly worth trying.

* Examples of all the reviews used in the questionnaire can be found in appendix 1.

Survey set-up

(21)

21

5.4 Measures

In order to measure if the respondents perceived the reviews they were exposed to as positive, negative or mixed they were asked to indicate how positive or negative they found the reviews they just read. This was done on a scale from 0 (negative) to 100 (positive) which was also used in the study of Chen, Liu and Zhang (2011). This scale is originally from Metacritic, which summarizes each review by giving a so called “metascore”. A score from 61-100 indicated a positive review, a score from 40-60 indicates a mixed or neutral review and a score from 0-39 indicates a negative review. As said by Chen, Liu and Zhang (2011) “these scores provide researchers with an independently, professionally assessed summary index or review valence”.

Subjects also had to indicate their attitude towards the product after having read the reviews or seen the sample. Their attitude was measured on a five-item, seven-point semantic differential scale; unappealing- appealing, bad- good, unpleasant- pleasant, unfavorable- favorable, and unlikeable-likeable (Spears and Singh, 2004).

5.5 Plan of analysis

The first step in the analysis would be to analyze the correlations and the Cronbach alpha for each multiple item questions. Followed by dummy coding of the variables, which will be necessary for the regression analysis. Scenarios with samples were dummy coded (1= sample yes, 0= sample no), valence of the reviews (Positive= 1, mixed and negative= 0, or mixed= 1, positive and negative=0) and type of product used in scenario (1=search 0= experience). A regression will be performed to discover any main effect of the valence of reviews and the sample on purchase intention and attitude, but also to analyze the moderation effect of type of good. The demographical variables can also be analyzed with the regression analysis to see whether gender or age, for example influences product attitude.

(22)

22

6. Results

This chapter describes the results of the analysis. First, he descriptive information of the respondents is described followed by the manipulations which analyzes is the respondents understood the manipulations in the questionnaire. The main and interaction effects that were found will be explained in subsection 6.4 and 6.5. Finally, a visual representation of the results will be shown.

6.1 Descriptive statistics

A total of 490 responses were collected with the questionnaire, however 251 responses were completed and could be used for the analysis. The average age of the respondents is 29 years and 44.6% were males and 55.4% females. Most of the respondents came from Europe (82.3%) and the average level of education is a Bachelor’s degree. The time people took to read the reviews and look at/hear the sample was timed in the questionnaire. The average time people read the reviews was 59 seconds and for the sample this was 58 seconds. Respondents were mainly collected through snowball- sampling through family and friends via email and social media. All the results are summarized in table 4.

Table 4: Descriptive statistics per scenario

Gender

Scenario N Mean

age

(23)

23

6.2 Scale items

Different scale- items were used to measure a range of independent variables. The reliability of these scales were tested with cronbach’s alpha and showed that all scales could be used effectively to measure the variable intended. The average response to each item and the average total response can be found in table 5 together with the exact reliability statistics. The correlations between each of the items has been tested before the cronbach’s alpha and can be found in appendix 2.

Table 5: Multi- item scales

Scales Source Mean SD

Cronbach’s α

Attitude towards the product Spears and Singh, 2004 0.935

1. unappealing- appealing 4.19 1.744 2. bad- good 4.39 1.607 3. unpleasant- pleasant 4.33 1.510 4. unfavorable- favorable 4.24 1.608 5. uninteresting- interesting 4.29 1.669 Total 4.29 1.471 Purchase intention

Li, Daugherty, Biocca,

2002 0.929

1. unlikely- likely 3.63 1.936

2. improbable- probable 3.64 1.894

3. uncertain- certain 3.63 1.827

4. not definitely- definitely 3.58 1.786

Total 3.62 1.699

Impression sample Kempf and Smith, 1998 0.891

1. dislike- like 5.28 1.528 2. unfavorable- favorable 5.05 1.544 3. bad- good 5.16 1.573 Total 5.16 1.415 Attitude sample

Bello, Pitts and Etzel,

1983 0.882*

1. exciting- boring 3.65 1.649 * after reverse

2. attention getting- dull 3.54 1.701 coding

3. memorable- easy forgettable 3.65 1.797

4. interesting- uninteresting 3.68 1.871

Total 3.63 1.507

Helpfulness

Li, Huang, Tan and Wei,

2013 0.807

(24)

24

product

2. helped me evaluate the product 5.03 1.392

3. helped me understand the performance of

the 4.99 1.510

product

Total 4.97 1.245

Experience attributes Weathers, Sharma and 0.883

Wood, 2007

1. it’s important for me to see this product to 3.85 1.565

evaluate how it will perform

2. it’s important for me to experience this

product to 4.05 2.012

evaluate how it will perform

3. it’s important for me to use this product 4.10 1.410

product to evaluate how it will perform

Total 4.00 1.514

Search attributes Weathers, Sharma and 0.920

Wood, 2007

1. I can adequately evaluate this product using

only 4.35 1.755

information provided by the retailer or

manufacturer

about the product’s attributes and features.

2. I can evaluate the quality of this product 4.50 1.878

simply by reading information about the

product.

(25)

25

6.2 Manipulation checks

Two manipulations were performed in the questionnaire. First the valence of the reviews was manipulated and second the type of product (search or experience) was manipulated. In order to check if the respondents understood that they read a positive, negative or mixed review and if they recognized it was about a search or experience good, an one-way ANOVA test was performed. As mentioned in the methodology section, a review with a score between the 61-100 is positive; 40-60 is mixed; and 0-39 is negative. The average score respondents gave was 77 in the positive review scenario, 56 in the mixed, and 25 in the negative scenario. The differences between them were found to be significant (p=.000). The results are summarized in table 6.

Respondents either read reviews about a search good (calculator app) or about an experience good (audio book). A pre-test that was performed among 20 respondents had to indicate whether people would perceive the audio book as an experience good and the calculator app as a search good. On a scale from 1-7 they had to indicate whether they agreed to certain statements which contained search and experience attributes. From the analysis it was found that for search products respondents scored low on the experience attributes (M =2.7) and respondents who had an experience good in their scenario scored high on the experience attributes (M= 5.3). This effect was found to be significant (p=.000). The results are summarized in table 7.

Table 6: Manipulation check valence

Scenario N Mean valence Std. Deviation Positive 72 77.61 16.270 Mixed 72 56.47 18.174 Negative 71 25.04 23.188 Total 215 53.17 28.990

Table 7: Manipulation check product type

Scenario N Mean Std. Deviation Search 10 2.7000 0.90880 Experience 10 5.3000 0.50796 Total 20 4.0000 1.51407

(26)

26 reviews (M=77.61) than for negative reviews (M=25.04). It also showed that when there was a sample available, the valence perception was higher (M=56.44) than when there was no sample (M=49.94).

6.3 The main effects

Before analyzing the results of this research, the regression equation will be shown.

=

+

= Attitude towards the product

x1= Gender x2= Age x3= Nationality x4= Education

z1= Positive review valence z2= Negative review valence z3= Mixed review valence z4= Search/ experience good z5= Sample yes/no

= Constant = Error term i = Respondent

This equation helps to clarify how and if attitude towards the product is explained by demographics, review valence, type of product and presence of samples. With demographics the age, gender, nationality and education of the respondents is meant, and it can be discovered whether any of these variables affect attitude towards the product. Moreover, the equation mentions three valence types to see what the effect of positive, negative or mixed reviews are on attitude towards the product. Finally, the regression also helps to discover if the effects are different for search and experience goods and if it makes a difference when there was a sample involved.

(27)

27 Additionally, valence of the reviews and type of product were added to measure their main and interaction effect on attitude towards the product. Although the results of the interaction effects model are more fine-grained, the effect of a sample on product attitude will be described by using the main effects model. The main effects model proved to be significant (F=17.279, p=.000) and about 40% of the variance was explained by the model (R²=.402). The Beta’s show that the presence of a sample has a significant, positive effect on attitude towards the product (B= 0.755 p=.000).

Moreover, the interaction model proved to be significant as well (F=23.216; p=.000) and explained about 45% of the variance (R²=.449). From this model it can be seen that compared to negative reviews, positive reviews have a positive effect on attitude (B= 1.467, p= .000). This is also true for the mixed reviews, those also have a positive effect on attitude towards the product compared to negative reviews (B= 1.171, p=.000). The results are summarized in table 8. These results support H1, that samples have a positive effect on attitude towards the product.

6.4 The moderating role of search/ experience goods

The moderation effect of search vs. experience goods was also tested with the regression. It was found that there is a significant moderating effect of the type of product on the relation of samples on attitude towards the product (B= 1.137, p=.000). In other words, the effect of the presence of samples on attitude towards the product is strengthened when there is a search good involved rather than an experience good. Furthermore, the main relationship of samples on attitude towards the product was also significantly strengthened when there was a positive review about a search good, compared to a negative review about a search good (B=.730, p=.056). However, a mixed review for a search good had no significant effect on the relation of samples on attitude (p=.387). All the results are summarized in table 8.

(28)

28 Table 8: Regression results on attitude towards the product

Main effect Interaction effect

B Std. Error t P B Std. Error t P (Constant) 2.467 0.448 5.501 .000 2.906 0.456 6.375 .000 Gender 0.200 0.163 1.228 .221 0.220 0.158 1.395 .165 Age 0.007 0.008 0.904 .367 0.004 0.008 0.467 .641 Nationality -0.043 0.078 -0.542 .588 -0.019 0.077 -0.248 .805 Education -0.122 0.088 -1.386 .167 -0.102 0.085 -1.201 .231 Sample yes 0.755 0.160 4.718 .000*** 0.184 0.218 0.843 .400 Positive review 1.838 0.196 9.387 .000*** 1.467 0.266 5.519 .000*** Mixed review 1.331 0.196 6.804 .000*** 1.171 0.265 4.416 .000*** Search good 0.627 0.158 3.954 .000*** -0.292 0.307 -0.951 .343 Sample x search 1.137 0.310 3.674 .000*** Positive x search 0.730 0.380 1.922 .056* Mixed x search 0.325 0.375 0.866 .387 R² .402 .449

*Significant at the 10% level. **Significant at the 5% level. ***Significant at the 1% level.

Figure 2: Results search goods

(29)

29 Figure 3: Results experience goods

(30)

30

7. Discussion and recommendations

This chapter will explain the findings in more detail and relate them to the research questions proposed in the first chapter. Managerial implications will be given from the results found and the limitations of this research will be explained, followed by future research directions.

7.1 Discussion

This research tried to discover the effects of samples in the online environment and whether these effects would be different for search or experience goods. One of the first research questions was if the addition of a sample to a review has a stronger effect on attitude towards the product than reviews alone. This was found to be true, as people tend to have a more favorable attitude towards the product when there is a sample than when there is only a review. This is related to the findings of Hu, Liu, Bose and Shen (2009) who state that when there is high uncertainty from product reviews, sampling plays a more important role in reducing this uncertainty. Since reviews are written by other product users and for a particular product, one can find both positive and negative reviews which make it difficult to decide which opinion to trust. Samples allow the consumers to experience a part of the product themselves and allow them to form their own opinions.

Furthermore, according to the results of this research online sampling has a positive influence on attitude towards the sampled product, irrespective of the type of product. This finding is in line with previous findings concerning sampling as Smith and Swinyard (1982) found that direct experiences lead to more strongly held attitudes. Thus, having the ability to experience the product before purchase could lead to more positively held attitudes. In addition, Roselius (1971) mentioned that products with high uncertainty can benefit strongly from samples. However, this author did not make any distinctions between search and experience goods in his research, it can be assumed that because the internet is still perceived as a purchasing channel with higher levels of uncertainty (Eggert, 2006) this holds for any product in general.

(31)

31 brand (MacKenzie, Lutz and Belch 1986; Homer 1990; Moore and Lutz 2000). In this research, attitude towards the sample influences attitude towards the product. The audio book sample which represented an experience good was not comparable to the calculation app sample which represented a search good. Because the sample of the calculator app was more entertaining, interesting and attention getting than the audio book sample it could have influenced the attitudes of the respondents stronger. It is thought that respondents were more involved and therefore more motivated to process the information given by the calculator sample than by the audio book sample. Perhaps because the calculator sample was visual and the audio book was not. Lang, Potter and Bolls (1999) confirm that visual encoding is easier than audio encoding of media content. A study by Christie and Collyer (2008) looked at video and audio clips and found that the visual components in a video clip were more interesting than audio only clips. They also found evidence that the visual components can increase a person’s confidence in what they remember. The authors mention that the combination of words and pictures can be more effective for learning purposes than when only using words. Another study by Faraday and Sutcliffe (1997) showed that a multimedia presentation would be more effective in promoting learning than that same presentation with only text and speech. Therefore it could be argued that due to the fact that visual media content is easier to encode and found to be more effective and interesting, respondents were more motivated to watch the sample of the calculator app and formed a more favorable attitude as opposed to the audio content.

(32)

32 will use the information from the reviews more extensively than when consumers are asked to state their attitude towards the product. The latter entails no risk as it concerns no purchasing decisions and maybe therefore makes consumers less attentive to the reviews.

The fact that no significant effects were found that experience goods benefit more from samples and reviews than search goods could also be due to the characteristics of an experience good. Because evaluation prior to purchase is important for an experience good to be bought, the internet might not be the appropriate channel to sell those goods (Peterson, Balasubramanian and Bronnenberg, 1997). Especially since products sold online can never be touched or seen in real life before purchase, experience goods might rather be bought in a store by consumers. In addition, if respondents were to use their smart phones daily, a calculator app was found to be more relevant than an audio book, which may be used less frequently. This was then reflected in their attitude towards the product, due to usage frequency.

Even though the effect of samples and reviews on product attitude was not stronger for experience goods, they do prove to be valuable for search goods. The reason for this, could also be due to the characteristics of both products, the type of sample (visual vs. audio) or simply because of their usage frequency. All in all, samples do have an effect on consumer’s product attitude, is found to be stronger for search goods, and reviews affect attitude towards a search good.

7.2 Recommendations

(33)

33 Limitations

Since this study is concerned about certain effects in the online environment, the finding of a suitable search good for this study posed to be a challenge. Most of the distinctions in current literature based on search or experience attributes concern products that do not have the ability to sample the product online. Moreover, only limited products have the ability to show or experience samples online. Another limitation is the type of products that were chosen. The audio book is not a typically bought product compared to a regular book. Instead of using an audio book as a search good, a normal book might have given different responses in the questionnaire, where respondents were then allowed to read a couple of pages of the book. This is more interactive and therefore maybe more attention getting than having to listen to audio playing. Furthermore, the questionnaire was distributed online and respondents were asked to read reviews and some also had to view or listen to a sample. However, there was no possibility to control for this, making sure respondents actually read all the reviews. Finally, the questionnaire distributed was in English and since it was distributed mainly to Dutch friends and relatives, some might not have understood the questions entirely since their level of English is not sufficient.

The conclusions derived from this research are based upon data collected through the use of two product samples, an audio fragment of an audio book and a video of a calculator app. As explained before, watching a video is less energy consuming for respondents than having to listen. Therefore it cannot be determined whether the differences in results are due to the fact that some respondents listened to an audio and some watched a video or whether these differences are truly because of the differences between search and experience goods.

Future research directions

(34)

34

8. References

Bart, Y., Shankar, V., Sultan, F., & Urban, G. L. 2005. Are the drivers and role of online trust the same for all web sites and consumers? A large-scale exploratory empirical study. Journal of Marketing, 69 (4): 133- 152.

Basuroy, S., Chatterjee, S., & Ravid, S. A. 2003. How critical are critical reviews? The box office effects of film critics, star power, and budgets. Journal of Marketing, 67 (4): 103-117.

Bawa, K., & Shoemaker, R. 2004. The effects of free sample promotions on incremental brand sales.

Marketing Science, 23 (3): 345-636.

Bearden, W. O, Lichtenstein, D. R., & Teel, J. E. 1984. Comparison price, coupon, and brand effects on consumer reactions to retail newspaper advertisements. Journal of Retailing, 60 (2): 11-34.

Bennett, P. D., & Mandell, R. M. 1969. Information seeking behavior of new car purchasers- The learning hypothesis. Journal of Marketing Research, 6: 430-433.

Berger, C. R., & Calabrese, R. J. 1975. Some explorations in initial interaction and beyond: Toward a developmental theory of interpersonal communication. Human Communication Research, 1 (2): 99-112.

Berger, J., Sorensen, A. T., & Rasmussen, S. J. 2010. Positive effects of negative publicity: When negative reviews increase sales. Marketing Science, 29 (5): 815-827.

Bettinger, C. O., Dawson, L. E., & Wales, H. G. 1979. The impact of free-sample advertising. Journal of

Advertising Research, 19 (3): 35-39.

Biswas, D., Grewal, D., & Roggeveen, A. 2010. How the order of sampled experiential products affects choice. Journal of Marketing Research, 47 (3): 508-519.

Biswas, D., Labrecque, L. I., Lehman, D. R., & Markos, E. 2014. Making choices while smelling, tasting and listening: The role of sensory (dis)similarity when sequentially sampling products. Journal of

(35)

35 Bucklin, L. P. 1966. Testing propensities to shop. Journal of Marketing, 30: 22-27.

Centraal bureau voor de statistiek. Steeds meer mensen kopen online. Retrieved from

http://www.cbs.nl/nl-NL/menu/themas/vrije-tijd-cultuur/publicaties/artikelen/archief/2013/2013-3850-wm.htm. Accessed on the 19th of February 2014.

Chen, Y., Liu, Y., & Zhang, J. 2011. When do third-party product reviews affect firm value and what can firms do? The case of media critics and professional movie reviews. Journal of Marketing, 75: 116-134.

Chen, Y., & Xie, J. 2005. Third-party product review and firm marketing strategy. Marketing Science, 24 (2): 218-240.

Chen, Y., & Xie, J. 2008. Online consumer review: Word-of-mouth as a new element of marketing communication mix. Management Science, 54 (3): 477-491.

Chevalier, J. A., & Mayzlin, D. 2006. The effect of word of mouth on sales: Online book reviews.

Journal of Marketing Research, 43: 345- 354.

Chiang, K., & Dholakia, R. R. 2003. Factors driving consumer intention to shop online: An empirical investigation. Journal of Consumer Psychology, 13: 177-183.

Christie, B., & Collyer, J. 2008. Do video clips add more value than audio clips? Presenting industrial research and development results using multimedia. Behavior & Information Technology, 27 (5): 395-405.

Cox, D.F. 1967. Risk taking and information handling in consumer behavior. Boston: Harvard University Press.

Crowley, A. E., & Hoyer, W. D. 1994. An integrative framework for understanding two-sided persuasion. Journal of Consumer Research, 20 (4): 561-574.

(36)

36 Dellarocas, C. 2006. Strategic Manipulation of internet opinion forums: Implications for consumers and firms. Management Science, 52 (10): 577-1593.

Duan, W., Gu, B., & Whinston, A. B. 2008. Do online reviews matter? An empirical investigation of panel data. Decision Support Systems, 45: 1007-1016.

Duan, W., Gu, B., & Whinston, A. B. 2008. The dynamics of online word-of-mouth and product sales- An empirical investigation of the movie industry. Journal of Retailing, 84: 233-242.

Eggert, A. 2006. Intangibility and perceived risk in online environments. Journal of Marketing

Management, 22: 553-572.

Faraday, P., & Sutcliffe, A. 1997. Designing effective multimedia presentations. CHI 97: 272-278.

Freedman, A. M. 1986. Use of free product samples wins new favor as sales tool. Wall Street Journal, p. 19.

Godes, D., & Mayzlin, D. 2004. Using online conversations to study word-of-mouth communication.

Marketing Science, 23 (4): 545-560.

Heiman, A., McWilliams, B., Shen, Z., & Zilberman, D. 2001. Learning and forgetting: Modeling optimal product sampling over time. Management Science, 47 (4): 532-546.

Hinterleitner, F., Neitzel, G., Möller, S., & Norrenbrock, C. 2011, An evaluation protocol for the subjective assessment of text-to-speech in audio book reading tasks. Proc. Blizzard Challenge.

Hoch, S. J., & Deighton, J. 1989. Managing what consumers learn from experience. Journal of

Marketing, 53: 1-20.

(37)

37 Holmes, J. H., & Lett, J. D. 1977. Product sampling and word of mouth. Journal of Advertising

Research, 17 (5): 35-40.

Homer, P. M. 1990. The mediating role of attitude toward the ad: Some additional evidence. Journal

of Marketing Research, 27 (1): 78-86.

Hu, N., Liu, L., Bose, I., & Shen, J. 2009. Does sampling influence customers in online retailing of digital music? Information Systems and E-Business Management, 8 (4): 357-377.

Huang, J. H., & Chen, Y. F. 2006. Herding in online product choice. Psychology and Marketing, 23: 413- 428.

Huang, P., Lurie, N. H., & Mitra, S. 2009. Searching for experience on the web: An empirical examination of consumer behavior for search and experience goods. Journal of Marketing, 73: 55-69.

Karni, E., & Schwarz, A. 1977. Search theory: The case of search with certain recall. Journal of

Economic Theory, 16: 38-52.

Kempf, DA. S., & Laczniak, R. N. 2001. Advertising’s influence on subsequent product trial processing.

Journal of Advertising, 30 (3): 27-38.

Klein, L. R. 1998. Evaluating the potential of interactive media through a new lens: Search versus experience goods. Journal of Business Research, 41: 195-203.

Koernig, S. K. 2003. The electronic environment and service intangibility. Psychology and Marketing, 20 (2): 151-159.

Landon, S., & Smith, C. E. 1997. The use of quality and reputation indicators by consumers: The case of Bordeaux wine. Journal of Consumer Policy, 20: 289-323.

Lang, A., Potter, R. F., & Bolls, P. D. 1999. Something for nothing: Is visual encoding automatic?

(38)

38 Ludwig, S., de Ruyter, K., Friedman, M., Brüggen, E. C., Wetzels, I., & Pfann, G. 2013. More than words: The influence of affective content and linguistic style matches in online reviews on conversion rates. Journal of Marketing, 77: 87-103.

MacKenzie, S. B., Lutz, R. J., & Belch, G. E. 1986. The role of attitude toward the ad as a mediator of advertising effectiveness: A test of competing explanations. Journal of Marketing Research, 23 (2): 130-143.

Marks, L. J., & Kamins, M. A. 1988. The use of product sampling and advertising: Effects of sequence of exposure and degree of advertising claim exaggeration on consumers’ belief strength, belief confidence and attitudes. Journal of Marketing Research, 25: 266-281.

McGuinness, D., Philip, G., & Mathew, S. 1992. The effect of product sampling on product trial, purchases and conversion. International Journal of Advertising, 11: 83-92.

Moore, W. L., & Lehman, D. R. 1980. Individual differences in search behavior for a nondurable.

Journal of Consumer Research, 7: 296-307.

Moore, E. S., & Lutz, R. J. 2000. Children, advertising, and product experiences: A multimethod inquiry. Journal of Consumer Research, 27: 31-48.

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

Nelson, P. 1970. Information and consumer behavior. Journal of Political Economy, 78 (2): 311-329.

Norton, S. W., & Norton, W. 1988. An economic perspective on the information content of magazine advertisements. International Journal of Advertising, 7 (2): 138-148.

(39)

39 Park, D. H., Lee, J., & Han, I. 2007. The effect of on-line consumer reviews on consumer purchase intention: The moderating role of involvement. International Journal of Electronic Commerce, 11 (4): 125-148.

Peterson, R. A., Balasubramanian, S., & Bronnenberg, B. J. 1997. Exploring the implications of the internet for consumers marketing. Journal of the Academy of Marketing Science, 24 (4): 329-346.

Reilly, M. D., & Conover, J. N. 1983. Meta-analysis: Integrating results from consumer research studies. Advances in Consumer Research, 10: 509-513.

Roselius, T. 1971. Consumer rankings of risk-reduction methods. Journal of Marketing, 35 (1): 56-61.

Scott, C. A. 1976. The effects of trial and incentives on repeat purchase behavior. Journal of

Marketing Research, 13: 263-269.

Smith, R. E. 1993. Integrating information from advertising and trial: Processes and effects on consumer response to product information. Journal of Marketing Research, 30: 204-219.

Smith, R. E., & Swinyard, W. R. 1983. Attitude-behavior consistency: The impact of product trial versus advertising. Journal of Marketing Research, 20 (3): 257-267.

Spears, N., & Singh, S. N. 2004. Measuring attitude towards the brand and purchase intentions.

Journal of Current Issues and Research in Advertising, 26: 53-66.

Thompson, S. H. T. 2006. To buy or not to buy online: adopters and non-adopters of online shopping in Singapore. Behavior and Information Technology, 25 (6): 497-509.

Tsang, A. S. L., & Prendergast, G. 2009. Is a ‘star’ worth a thousand words? The interplay between product-reviews texts and rating valences. European Journal of Marketing, 43: 1269- 1280.

Tu, Y., & Lu, M. 2006. An experimental analytical study of on-line digital music sampling strategies.

Customer Engagement Differences in the effect of reviews and/or samples on product attitude per product type