Intentions to use an online restaurant review Web site and purchase behavior after reading reviews
First Monday

Intentions to use an online restaurant review Web site and purchase behavior after reading reviews by Joshua Fogel and Mohit Kumar



Abstract
The purpose of this paper is to study associations of variables with intentions to use an online restaurant review Web site and variables associated with purchase behavior at a restaurant after reading restaurant reviews. Participants were 613 college students who answered questions about demographics, trust, knowledge, and Internet experience. Hispanic and Asian/Asian American race/ethnicity, increased trust, and Internet experience variables were significantly associated with intentions. Intentions, knowledge about review fraud, and Internet experience variables were significantly associated with visiting and purchasing at restaurants. Restaurant managers should consider expanding their online presence by responding professionally to reviews.

Contents

Introduction
Literature review
Method
Results
Discussion
Conclusion

 


 

Introduction

The Internet has created new venues for consumers to share opinions and assessments of products and services (Zhang, et al., 2010). Online reviews have an important role in consumer decisions for purchasing products with one survey reporting that 90 percent of individuals indicated that positive reviews affected their purchase decisions and 80 percent indicated that negative reviews affected their purchase decisions (Dimensional Research, 2013). Restaurant reviews are the most commonly searched topic in online reviews with 67 percent of consumers searching for reviews about restaurants (BrightLocal, 2013) and 15 percent indicating that they use online review Web sites every time to search for restaurant reviews (Ghiselli and Ma, 2015).

Although online restaurant reviews are important to consumers, there appears to be limited research on the topic of how restaurant reviews influence consumers to choose to purchase at restaurants. One study of an anonymous Chinese restaurant review Web site found that positive reviews and greater number of reviews were associated with increased restaurant sales while negative reviews were associated with decreased restaurant sales (Lu, et al., 2013). However, this study was done on the aggregate level and did not provide any information about individual consumer demographics. A major research gap worth understanding is the consumer demographics that are associated with consumer restaurant purchases after reading online restaurant reviews. In addition, there is a study about characteristics of online restaurant reviews such as rating type, number of reviews, and restaurant chain affiliation and the association of these characteristics with review fraud (Luca and Zervas, 2016). However, there is no research about either trust in online restaurant reviews or knowledge about review fraud for online restaurant reviews with consumer intention to use an online restaurant review Web site or purchase behavior at restaurants after reading online restaurant reviews.

This paper has two purposes. One aim is to study intentions to use an online restaurant review Web site. Another aim is to study consumer use of a service or purchase of a product after reading reviews on an online restaurant review Web site. The methodology used is to consider relevant variables from the areas of demographics, trust, knowledge about online review fraud, Internet experience, and the theory of planned behavior as predictor variables for analyses. A multivariate framework is used to consider the impact of these variables when included in the same analytical model.

This paper provides several contributions. First, many research papers often examine intentions but do not study behavior. However, intentions do not always translate into behavior. This paper helps understand not only consumer intentions for using an online restaurant review Web site but also consumer purchase behavior after reading restaurant reviews. Second, this paper provides guidance for restaurant managers on how online consumer reviews are perceived and reacted upon by potential consumers. Third, this paper provides applied recommendations for restaurant managers on how best to interact with online restaurant review Web sites. Fourth, this paper provides internationally applied recommendations for restaurant managers.

 

++++++++++

Literature review

Demographic factors are known to affect online purchases and consumer decisions for purchases after examining an online review Web site. Younger age groups are more likely to shop online than older age groups (Lian and Yen, 2014). Younger adults such as college students often read and are influenced by content from online review Web sites (Mangold and Smith, 2012). Online reviews alter purchase intention to a greater degree in women than men (Bae and Lee, 2011). Koreans reported a higher level than those from the United States that the perceived usefulness of online reviews leads to reviews influencing purchases (Park and Lee, 2009). We are not aware of any literature on whether demographic factors affect choosing to visit an online restaurant review Web site or how online restaurant reviews are perceived by consumers and whether consumers will choose to purchase from a restaurant after reading a restaurant review.

We hypothesize:

Hypothesis 1a: Demographic factors of younger age, female sex, Asian/Asian American race/ethnicity, and those not born in the United States are positively associated with intention to use an online restaurant review Web site.

Hypothesis 1b: Demographic factors of younger age, female sex, Asian/Asian American race/ethnicity, and those not born in the United States are positively associated with the behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site.

Trust can play a large role in the valuation of online review Web sites and online reviews. It is well known that increased trust is associated with increased intentions to purchase products online (Ponte, et al., 2015). Increased trust in company product recommendations for both calculator and music products is associated with increased intentions to purchase (Benlian, et al., 2012). A greater number of people endorsing and trusting a reviewer is associated with greater trust perception (Xu, 2014). However, online reviewer credibility for positive reviews is not associated with purchase intentions for cameras (Jensen, et al., 2013).

Positive or negative reviews may affect how a review is trusted. A study of ratings of both DVD and book reviews on Amazon.com found that negative reviews had a greater impact than positive reviews for receiving more votes and for being rated as more helpful (Kuan, et al., 2015). The greater amount of content in a review containing both positive and negative statements is viewed as more trustworthy than reviews with a lesser amount of content containing both positive and negative statements (Jensen, et al., 2013). Multiple reviews of a product with both positive and negative reviews is associated with increased purchase intentions; also women have greater purchase intentions than men when reading such reviews (Zhang, et al., 2014). A sequence along with a positive or negative review approach affects perceived usefulness. The sequence of a negative review, positive review, and then negative review is perceived as most useful as compared to other sequences of positive/negative/positive and positive/negative but not negative/positive (Purnawirawanv, 2012). With regard to restaurants we are aware of only one study that found that positive reviews and greater number of reviews were associated with increased restaurant sales while negative reviews were associated with decreased restaurant sales (Lu, et al., 2013).

In general by online reviews, trust that a brand can satisfy consumer needs affects the relationship of reading online reviews with willingness to buy a given product (Chang, et al., 2013). In addition, it is possible that trust in online review content used to increase search engine optimization is associated with willingness to buy a product. There is limited research on trust of online reviews and intention to purchase and we are not aware of any research on trust of online reviews and purchase of a product online. Furthermore, we are not aware of any research on trust of online reviews for restaurants and either intention to use an online restaurant review Web site or purchase of a product based upon reading an online restaurant review. We also are not aware of any research on the impact of either positive or negative reviews on the trust of online reviews for restaurants and either intention to use an online restaurant review Web site or purchase of a product based upon reading an online restaurant review.

We hypothesize:

Hypothesis 2a: Brand trust is positively associated with intention to use an online restaurant review Web site.

Hypothesis 2b: Brand trust is positively associated with the behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site.

Hypothesis 3a: Search engine optimization trust is positively associated with intention to use an online restaurant review Web site.

Hypothesis 3b: Search engine optimization trust is positively associated with the behavior of use of a service or purchase of a product after reading reviews from an online restaurant review Web site.

Hypothesis 4a: Positive wording in a review is positively associated with intention to use an online restaurant review Web site.

Hypothesis 4b: Positive wording in a review is positively associated with the behavior of use of a service or purchase of a product after reading reviews from an online restaurant review Web site.

Hypothesis 5a: Negative wording in a review is negatively associated with intention to use an online restaurant review Web site.

Hypothesis 5b: Negative wording in a review is negatively associated with the behavior of use of a service or purchase of a product after reading reviews from an online restaurant review Web site.

Consumer awareness of online review fraud may affect consumer purchase intention and purchase behavior. Consumers believe that newly advertised books, DVDs, and videos on Amazon.com contain manipulated reviews and focus more on price as a quality indicator rather than the review (Hu, et al., 2011). Also, when consumers are unsure if a book review on Amazon.com is a manipulated review, consumers ignore these reviews when deciding which product to purchase (Hu, et al., 2012). In a study about the impact of online restaurant reviews, low ratings were associated with increased positive review fraud, a greater number of reviews were associated with decreased positive review fraud while a lesser number of reviews were associated with increased positive review fraud, and chain restaurants were associated with decreased positive review fraud (Luca and Zervas, 2016). We are not aware of any literature for restaurants on knowledge of online review fraud and its association with either intention to use online review Web sites or purchases after reading online review Web sites.

We hypothesize:

Hypothesis 6a: Knowledge of online review Web site fraud patterns is positively associated with intention to use an online restaurant review Web site.

Hypothesis 6b: Knowledge of online review Web site fraud patterns is positively associated with the behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site.

Prior online review Web site experience can affect purchase intention and purchase behavior. Prior online shopping experience is associated with increased purchase intentions (Ling, et al., 2010; Tong, 2010). Acceptance that online reviews, blog posts, and other forms of electronic word of mouth are credible is associated with increased purchase intention (Fan and Miao, 2012). Reading online reviews about automobiles is associated with increased purchase intentions for cars (Jalilvand and Samiei, 2012a). We are not aware of any literature for restaurants on previous review Web site experience and its association with either intention to use online review Web sites or purchases after reading content on online review Web sites.

We hypothesize:

Hypothesis 7a: Prior online restaurant review Web site experience is positively associated with intention to use an online restaurant review Web site.

Hypothesis 7b: Prior online restaurant review Web site experience is positively associated with the behavior of use of a service or purchase of a product after reading reviews from an online restaurant review Web site.

The theoretical framework used in this study is the theory of planned behavior. This theory suggests that subjective social norms, attitudes, and perceived behavioral control lead to intentions which then leads to behavior (Askew, et al., 2014). The theory of planned behavior has been successfully used as a conceptual framework to relate electronic word of mouth, a construct that includes reviews on online review Web sites, to intentions to travel (Jalilvand and Samiei, 2012b). Also, this theory has been successfully used for food topics for understanding intentions to purchase halal food (Alam and Sayuti, 2011).

We hypothesize:

Hypothesis 8a: The theory of planned behavior is useful for understanding intention to use an online restaurant review Web site with attitudes, social norms, and behavioral control positively associated with intentions.

Hypothesis 8b: The theory of planned behavior is useful for understanding the behavior of use of a service or purchase of a product after reading reviews an online restaurant review Web site with attitudes, social norms, and behavioral control positively associated with behavior.

 

++++++++++

Method

Participants

The survey was conducted at a commuter inner city public college located in New York City. There were 745 students approached and 60 students declined to participate. Also, there were 23 surveys collected that were invalid and an additional seven students did not respond to the outcome questions. The response rate of 87.9 percent was calculated from the 655 completed surveys [(655/745) * 100%]. Also, 42 people above the age of 36 were excluded from the sample in order to maintain a sample of young adults. The data were analyzed from 613 surveys. Data collection occurred in December 2013 and January 2014. The study was declared except from formal full-board Institutional Review Board review.

After providing informed consent, participants completed an anonymous survey either before, during, or after class. The survey began with a description of the purpose of three consumer review Web sites of Yelp (many types of services), TripAdvisor (hotel and travel services), and UrbanSpoon (restaurant services). Participants were then asked to answer a series of questions in regard to these Web sites. The analyses in this paper focus on UrbanSpoon, a review Web site that focuses on eating establishments (e.g., restaurants, cafes). At the time of the survey and data collection, UrbanSpoon was one of the major online restaurant review Web sites in the United States. UrbanSpoon had both a Web page and mobile application and had a total of 255 million visits in 2011 (Perez, 2012). UrbanSpoon allowed for general searches and for specific searches such as those focusing on neighborhood, type of cuisine, and special services provided by the restaurant (e.g., free wifi) (UrbanSpoon, 2015). In 2015, UrbanSpoon was purchased by the international online restaurant review Web site of Zomato. In June 2015, Zomato chose to rename UrbanSpoon content under the name of Zomato and made some changes for how consumers review restaurants such as now using a five-point scale rather than a like/dislike dichotomous response (Gupta, 2015).

Measures

       Theory of planned behavior

The theory of planned behavior variables were questions on intentions, attitudes, social norms, and behavioral control. The manual for creating a survey using the theory of planned behavior were used to create these questions (Francis, et al., 2004).

       Intentions

The intentions scale had three items. A Likert-style scale was used to measure the items with a range from 1=strongly disagree to 7=strongly agree. A sample item is, “I intend to read reviews about a merchant or vendor on UrbanSpoon to influence my decision about whether to use a service or purchase a product.” The total score was calculated by adding the three items. Greater scores indicate greater intentions. Cronbach alpha in this sample was 0.96.

       Attitudes

The attitudes scale had four items. A Likert-style scale was used to measure all the items with a range from 1 to 7 with negative and positive attitudes as the endpoints. Two items were reverse coded to be in the direction of 7=positive. All items were prefixed by the phrase of “Reading reviews about a merchant or vendor on online review-based Web sites such as Yelp, TripAdvisor, or UrbanSpoon is:” A sample positive endpoint had worthless=1 and useful=7. The total score was calculated by adding the four items. Greater scores indicate greater attitudes. Cronbach alpha in this sample was 0.75.

       Social norms

The social norms scale had three items. A Likert-style scale was used to measure all the items with a range from 1=strongly disagree to 7=strongly agree. A sample item is, “I feel under social pressure to read reviews about a merchant or vendor on online review-based Web sites such as Yelp, TripAdvisor, or UrbanSpoon.” The total score was calculated by adding the three items. Greater scores indicate greater social norms. Cronbach alpha in this sample was 0.64.

       Behavioral control

The behavior control topic had four items. A Likert-style scale was used to measure all the items with a range from 1=strongly disagree to 7=strongly agree. A sample item is, “Whether I make a decision about a merchant or vendor after reading reviews about a merchant or vendor on online review-based Web sites such as Yelp, TripAdvisor, or UrbanSpoon is entirely up to me.” Two items were reverse coded. Due to poor Cronbach alpha for all four items, a total score could not be used and each item was analyzed separately.

       Behavior

Behavior was measured with choices of “no” or “yes” to the item of, “After reading reviews about a merchant or vendor on UrbanSpoon, these reviews resulted in my use of a service or purchase of a product.”

       Demographics

Demographic variables were age (years), sex (man, woman), race/ethnicity (White, African American, Hispanic American, Asian/Asian American, South Asian, or other), and born in the United States (no/yes).

       Brand trust

Brand trust had four items. The items were based upon topics used in a previous scale (Valta, 2013) and modified for the current topic. A Likert-style scale was used to measure all the items with a range from 1=strongly disagree to 5=strongly agree. A sample item is, “I trust online reviews that I read on online review-base Web sites such as Yelp, TripAdvisor, or UrbanSpoon about merchants or vendors.” The total score was calculated by adding the four items. Greater scores indicate greater brand trust. Cronbach alpha in this sample was 0.78.

       Search Engine Optimization (SEO) trust

SEO trust had four items. The items were based upon topics studied in previous research (Anderson and Magruder, 2012; Short, 2013). A Likert-style scale was used to measure all the items with a range from 1=strongly disagree to 5=strongly agree. As example item is, “The number of recent reviews within the past month for a vendor or merchant influences me to trust the online reviews for a vendor or merchant.” The total score was calculated by adding the four items. Greater scores indicate greater SEO trust. Cronbach alpha in this sample was 0.76.

       Word use trust

Two items measured use of either positive or negative words in reviews for determining trust in that review. The items were based upon topics studied in previous research (Ludwig, et al., 2013). A Likert-style scale was used to measure all the items with a range from 1=strongly disagree to 5=strongly agree. These items were, “Use of positive words such as love, nice, or sweet makes an online review more trustworthy” and “Use of negative words such as ugly, dumb, or hate makes an online review more trustworthy.”

       Knowledge

There were five questions used to measure knowledge. Questions were based upon topics studied in previous research with Yelp (Luca and Zervas, 2016). Choice responses were “yes” or “no.” These items were: 1) A vendor or merchant that has a large number of low star or negative ratings on an online review-based Web site has increased risk for fraud with positive reviews; 2) A vendor or merchant with very few reviews on an online review-based Web site has increased risk for fraud with positive reviews; 3) A vendor or merchant with many reviews on an online review-based Web site has decreased risk for fraud with positive reviews; 4) A vendor or merchant that is part of a branded chain (and is not an independent establishment) has decreased risk for fraud with positive reviews on an online review-based Web site; and, 5) A vendor or merchant with a claimed page on an online review-based Web site where the vendor or merchant can respond to consumer comments, add photos, and post information about the service establishment has increased risk for fraud with positive reviews. Correct knowledge for each question was an answer of yes.

       Internet experience

There were five different topics on previous online review-based Web site experience. One item was measured on a Likert-style with a range from 1=strongly disagree to 5=strongly agree. This item was, “I typically read online review-based Web sites such as Yelp, TripAdvisor, or UrbanSpoon, before visiting a new vendor or merchant.” Two questions were about number of reviews read and percentage of reviews read that were believed to be real. These questions were: “How many online reviews do you read from review-based Web sites such as Yelp, Trip Advisor, or Urban Spoon, before visiting a new vendor or merchant?,” and “What percentage of online reviews from review-based Web sites such as Yelp, TripAdvisor, or Urban Spoon do you believe are those that were really composed by consumers?” These questions were based upon topics studied in a previous article (Short, 2013). Two other items were, “I previously wrote a review on UrbanSpoon” and “I never read a review on UrbanSpoon.” Choice responses were either “no” or “yes.” There was also an additional sixth question on general Internet use of, “Approximately how many hours do you use the Internet each week?”

Statistical analysis

Descriptive statistics of mean and standard deviation were used for the continuous variables and percentage and frequency for the categorical variables. Linear regression analysis studied the outcome of intentions to use an online restaurant review Web site. Predictors included the theory of planned behavior variables (attitudes, social norms, and behavioral control), demographic variables, trust variables, knowledge variables, and Internet experience variables. Logistic regression analysis studied the behavior outcome of reading reviews on an online restaurant review Web site resulting in use of a service or purchase of a product. Predictors included all the above variables used for linear regression and also included intentions. For both the linear and logistic regression analyses, univariate analysis were initially conducted. Only those variables statistically significant in the univariate analysis were then simultaneously included in the multivariate analysis. The variable of percentage of reviews really composed by consumers had a skewed distribution. As there were responses of zero precluding a logarithmic transformation, the value of 1 was added to all participants and then the variable was logarithmic transformed. All analyses used IBM SPSS, version 23 (IBM, 2013). All p-values were two-tailed.

 

++++++++++

Results

Table 1 describes the variables. With regard to the demographic variables, the mean age was nearly 23 years old and slightly more than half were women. More than two-thirds were non-White with almost a quarter Asian/Asian American, and both Hispanic Americans and African Americans were each at over 10 percent. Slightly more than half were born in the United States. With regard to the theory of planned behavior variables, attitudes were above the midpoint toward strongly agree. Social norms were below the midpoint towards strongly disagree. All behavior control variables were above the midpoint toward strongly agree. With regard to the trust variables, brand trust, SEO trust, and positive word trust were all above the midpoint toward strongly agree. Negative word trust was at the midpoint between strongly disagree and strongly agree. With regard to the knowledge variables, knowledge ranged from as high as 57.3 percent to as low as 34.4 percent. With regard to the Internet experience variables, typically read reviews before visiting was above the midpoint towards strongly agree. Slightly less than one-tenth had written an UrbanSpoon online restaurant review while slightly more than half never read an UrbanSpoon online restaurant review. The mean number of reviews read before trying a new vendor was almost seven and the mean percentage of reviews that were thought to be real was more than 55 percent. Mean Internet weekly use was 27 hours. With regard to the outcome variables, intentions was at the midpoint between strongly disagree and strongly agree. Also, slightly more than one-third engaged in the behavior of use of a service or purchase of a product after reading a review on the online restaurant review Web site of UrbanSpoon.

 

Sample characteristics of college students
Sample characteristics of college students
Sample characteristics of college students
Larger version of complete table available here.

 

Table 2 shows linear regression analysis for intentions to use an online restaurant review Web site. In the univariate analysis, the demographic variables of Asian/Asian American and Other were both statistically significantly associated with increased intentions. With regard to the theory of planned behavior variables, increased attitudes, increased social norms, increased confidence in decision-making, and increased decision-making entirely up to me were each statistically significantly associated with increased intentions. Increased control of decision making was statistically significantly associated with decreased intentions. With regard to the trust variables, increased brand trust, increased SEO trust, increased positive word trust, and increased negative word trust were each statistically significantly associated with increased intentions. With regard to the knowledge variables, knowledge of review fraud for vendor part of branded chain and knowledge of review fraud by vendors who can respond to consumer comments were each statistically significantly associated with increased intentions. With regard to the Internet experience variables, increased typically read review-based Web sites before visiting a new vendor or merchant and wrote an UrbanSpoon online restaurant review were each statistically significantly associated with increased intentions. Never wrote an UrbanSpoon online restaurant review was statistically significantly associated with decreased intentions. Increased number of reviews read and increased percentage of reviews believed to be real were each statistically significantly associated with increased intentions. Internet use was not significantly associated with intentions.

Table 2: Linear Regression Analyses for Intention to Use An Online Restaurant Review Web site.
Note: SD=standard deviation, SEO=search engine optimization.

Variable

Univariate

B

 

SE

 

p-value

Multivariate

B

 

SE

 

p-value

Demographics

 

 

 

 

 

 

Age (years)

0.03

0.06

0.54

---

---

---

Sex (women)

0.30

0.38

0.44

---

---

---

Race/Ethnicity

   White

   African American

   Hispanic American

   Asian/Asian American

   Southeast Asian

   Other

 

Reference

0.27

1.06

1.73

1.48

2.05

 

 

0.64

0.60

0.50

0.78

0.71

 

 

0.67

0.08

0.001

0.06

0.004

 

Reference

0.78

1.13

1.76

0.59

2.71

 

 

0.63

0.58

0.49

0.74

0.68

 

 

0.21

0.048

<0.001

0.43

<0.001

Born in United States

0.02

0.38

0.96

---

---

---

Theory of Planned Behavior

 

 

 

 

 

 

Attitudes

0.25

0.04

<0.001

0.08

0.05

0.08

Social Norms

0.38

0.05

<0.001

0.06

0.06

0.32

Confidence in decision making based on reviews and review websites

1.05

0.13

<0.001

0.42

0.16

0.01

Easiness of decision making after reading reviews

-0.22

0.14

0.10

---

---

---

Control over decision-making after reading reviews

-0.68

0.12

<0.001

-0.42

0.13

0.001

Decision-making after reading reviews entirely up to me

0.35

0.12

0.004

0.05

0.13

0.69

Trust

 

 

 

 

 

 

Brand Trust

0.55

0.07

<0.001

0.07

0.10

0.46

SEO Trust

0.59

0.07

<0.001

0.28

0.09

0.002

Use of positive words makes an online review more trustworthy

0.96

0.18

<0.001

0.50

0.22

0.02

Use of negative words makes an online review more trustworthy

0.44

0.17

0.01

-0.25

0.20

0.22

Knowledge

 

 

 

 

 

 

Knowledge of review fraud based on number of negative reviews (yes)

0.61

0.39

0.12

---

---

---

Knowledge of review fraud based on few reviews (yes)

0.46

0.38

0.23

---

---

---

Knowledge of review fraud based on many reviews (yes)

0.49

0.38

0.20

---

---

---

Knowledge of review fraud for vendor part of branded chain (yes)

1.20

0.38

0.002

0.21

0.38

0.57

Knowledge of review fraud by vendors who can respond to consumer comments (yes)

0.85

0.40

0.03

0.30

0.39

0.44

Internet Experience

 

 

 

 

 

 

Typically read online review-based websites before visiting a new vendor or merchant.

1.15

0.17

<0.001

0.64

0.20

0.001

Previously wrote UrbanSpoon reviews

2.29

0.70

0.001

1.91

0.72

0.01

Never read review on UrbanSpoon

-1.67

0.37

<0.001

-1.58

0.37

<0.001

Number of reviews read before trying a new vendor

1.97

0.48

<0.001

-0.10

0.50

0.83

Percentage of reviews believed to be real

0.02

0.01

0.01

0.004

0.01

0.58

Internet hours (weekly)

-0.34

0.51

0.50

---

---

---

Intercept

---

---

---

0.94

1.57

0.55

In the multivariate analysis for intentions to use an online restaurant review Web site, the demographic variables of Hispanic American, Asian/Asian American and Other were statistically significantly associated with increased intentions. The theory of planned behavior variable of increased confidence in decision making was statistically significantly associated with increased intentions while increased control of decision-making was statistically significantly associated with decreased intentions. The trust variables of increased SEO trust and increased positive word trust were each statistically significantly associated with increased intentions. The Internet experience variables of increased typically read review-based Web sites before visiting a new vendor or merchant and wrote an UrbanSpoon online restaurant review were each statistically significantly associated with increased intentions. Never wrote an UrbanSpoon online restaurant review was statistically significantly associated with decreased intentions. None of the knowledge variables were significantly associated with intentions.

Table 3 shows logistic regression analysis for behavior of use of a service or purchase of a product after reading an online restaurant review. In the univariate analysis, with regard to the demographic variables, only Other race/ethnicity was statistically significantly associated with increased odds for behavior. With regard to the theory of planned behavior variables, increased intentions, social norms, and confidence in decision-making were each statistically significantly associated with increased odds for behavior. Control over decision-making was statistically significantly associated with decreased odds for behavior. With regard to the trust variables, increased brand trust, SEO trust, positive word trust, and negative word were each statistically significantly associated with increased odds for behavior. With regard to the knowledge variables, knowledge of review fraud based on few reviews, knowledge of review fraud for vendor part of branded chain, and knowledge of review fraud by vendors who can respond to consumer comments were each statistically significantly associated with increased odds for behavior. With regard to the Internet experience variables, increased typically read review-based Web sites before visiting a new vendor or merchant and wrote an UrbanSpoon online restaurant review were each statistically significantly associated with increased odds for behavior. Never read an online restaurant review on UrbanSpoon was statistically significantly associated with decreased odds for behavior. Also, increased number of reviews read was statistically significantly associated with increased odds for behavior.

 

Logistic regression analyses for behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site
Logistic regression analyses for behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site
Logistic regression analyses for behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site
Table 3: Logistic regression analyses for behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site.
Larger version of complete table available here.

 

In the multivariate analyses, only the theory of planned behavior variable of increased intentions was statistically significantly associated with increased odds for behavior. The knowledge variables of knowledge of review fraud based on few reviews and knowledge of review fraud by vendors who can respond to consumer comments were each statistically significantly associated with increased odds for behavior. The Internet experience variable of typically read review-based Web sites before visiting a new vendor or merchant was statistically significantly associated with increased odds for behavior while never read an online restaurant review on UrbanSpoon was statistically significantly associated with decreased odds for behavior. None of the demographic variables and trust variables were statistically significantly associated with odds for behavior.

 

++++++++++

Discussion

We found that the demographics of Hispanic American, Asian/Asian American and Other were significantly associated with increased intentions to use an online restaurant review Web site. None of the other demographic variables of age and sex had any significant association with intentions and also none of the demographic variables had any significant association with behavior of use of a service or purchase of a product after reading reviews on an online restaurant review Web site. We found that SEO trust and positive word trust were significantly associated with increased intentions. None of the other trust variables of brand trust and negative word trust had any significant association with intentions and also none of the trust variables had any significant association with behavior. None of the knowledge items had any association with intentions. Two of the knowledge items of knowledge of review fraud based on few reviews and knowledge of review fraud by vendors who can respond to consumer comments were each significantly associated with increased behavior. The Internet experience variables of typically read reviews before using a new vendor or merchant and those who wrote an UrbanSpoon online restaurant review were each associated with significantly increased intentions while never read an UrbanSpoon online restaurant review was significantly associated with decreased intentions. A similar significance pattern only occurred for typically read reviews with increased behavior and never read an UrbanSpoon online restaurant review with decreased behavior. The theoretical framework only showed an association of increased intentions with increased behavior.

We found partial support for hypothesis 1a where the race/ethnicity groups of Hispanic America, Asian/Asian American, and Other had a statistically significant association with increased intentions to use an online restaurant review Web site while the other demographic variables of age and sex did not have any significant association with intentions. Previous research with adolescents reports that Hispanics use the Internet less than whites as a reference source for clothing purchases (Seock and Hathcote, 2010). Our research with young adults differs and shows that Hispanics have greater intentions than whites to use an online restaurant review Web site. Also, previous research on reading online hotel reviews and digital camera reviews found that Chinese consumers have greater purchase intentions than those from the United Kingdom (Christodoulides, et al., 2012). This is similar to our findings of increased intentions for Asian/Asian American to use an online restaurant review Web site. We did not find any support for hypothesis 1b, as no demographic variable was significantly associated with behavior. Apparently, intentions to use an online restaurant review Web site do not necessarily translate into behavior of visiting restaurants reviewed even for Hispanic Americans and Asian/Asian Americans who have the increased intentions to use an online restaurant review Web site.

We found support for hypotheses 3a and 4a where both search engine optimization trust and positive word trust had a statistically significant association with increased intention to use an online restaurant review Web site. We found no support for hypotheses 2a and 5a concerning the association of brand trust and negative word trust with intentions. No support was found for hypotheses 2b, 3b, 4b, and 5b as no trust variables had a significant association with behavior. Previous research with reading online hotel reviews found a statistically significant association for both positive and negative online reviews where positive reviews are associated with greater purchase intentions than negative reviews (Mauri and Minazzi, 2013). There are a couple of possibilities why our finding differs from this research. First, our study separately compared positive and negative reviews to intentions while the hotel review study directly compared positive and negative reviews to each other. Second, it is possible that each industry differs where negative reviews may not have as much of an impact on purchase intention for restaurants. Our lack of any association for any trust variable with restaurant purchase behavior may have occurred because trust is not as relevant for a one-time relatively inexpensive purchase such as a restaurant meal, especially since people are willing to experiment with and try out the food at a restaurant.

Hypothesis 6a was not supported as none of the knowledge variables had a statistically significant association with intentions. However, hypothesis 6b was partially supported. Two of the five knowledge variables had a statistically significant association with behavior. These two variables were knowledge of review fraud based on few reviews and knowledge of review fraud by vendors who can respond to consumer comments. Previous research found that consumers are aware that newly advertised products on Amazon.com contain manipulated reviews (Hu, et al., 2011). Our finding with online restaurant reviews indicates that such knowledge of that there can be review fraud when there are only a few reviews is associated with increased purchase behavior. This apparent contradiction of awareness that the reviews are not necessarily factual yet there is increased purchase behavior may occur since consumers are curious about newly opened restaurants with few reviews and regardless of the truthfulness of the reviews may purchase food at this restaurant. This reasoning can also explain the lack of any statistically significant association of knowledge of review fraud based on few reviews with intentions to use an online restaurant review Web site, as from a consumer perspective the reviews are not critical towards intentions to purchase at the restaurant. Our finding that knowledge of review fraud by vendors who can respond to consumer comments is associated with increased purchase behavior can occur because consumers recognize that even if some reviews are manipulated by the vendor, the vendor is monitoring the online review Web site and is concerned about the online review content, and thus will make an effort at the restaurant to ensure appropriate service and food quality.

Hypothesis 7a was supported as typically reading online review based Web sites and previously writing a review on the online restaurant review Web site of UrbanSpoon was associated with greater intention to use an online restaurant review Web site. Previous research found that acceptance that online reviews, blog posts, and other forms of electronic word of mouth are credible is associated with increased purchase intentions (Fan and Miao, 2012). Our research with an online restaurant review Web site is similar to this approach. Hypothesis 7b was only partially supported as although the variable of typically reading online review based Web sites was associated with purchase behavior but the variable of previously wrote a review on the online restaurant review Web site of UrbanSpoon was not associated with purchase behavior. A possible suggestion for the lack of association for previously wrote a review on UrbanSpoon and purchase behavior is because those who write reviews are those with very specific and particular opinions about the restaurant food experience. They may have high opinions of their own reviews and not too high opinions about reviews from others. Thus, they are less likely to be influenced by the opinion of other food reviews that they read on UrbanSpoon. Future research should study this potentially concerning topic of the lack of interest in visiting restaurants from those who are the reviewers on UrbanSpoon.

Hypothesis 8a was minimally supported as we found almost no evidence that the theory of planned behavior was useful for understanding intentions to use an online restaurant review Web site. We found that attitudes and social norms did not have a statistically significant association with intentions. Although two variables of behavior control did have a statistically significant association with intentions, they were in contradiction with each other with one associated with increased intentions and the other with decreased intentions. Hypothesis 8b was partially supported as intentions had a statistically significant association with increased behavior. However, the antecedents of intentions of attitudes, social norms, and behavior control were not statistically significantly associated with purchase behavior. The theory of planned behavior is a very popular theory and previous e-commerce research with intentions to travel (Jalilvand and Samiei, 2012b) and food research of intentions to purchase halal food (religiously mandated food for Muslims) (Alam and Sayuti, 2011) both report that attitudes, social norms, and perceived behavior control were all statistically significantly associated with intentions. It is quite surprising to see only limited support for use of the theory of planned behavior for understanding intentions and purchases regarding the online restaurant review Web site of UrbanSpoon. It is possible that for non-religiously mandated food choices that are a one-time relatively inexpensive purchase such as a restaurant meal, people are willing to experiment with and try out the food at a restaurant without considering attitudes, social norms, and behavior control.

There are several study limitations. First, our study included those from one college and respondents from other colleges may have a different pattern of responses. Second, we did not ask questions about the typical length of restaurant reviews read. Length of review may influence trust in a review and also may influence intentions and behavior regarding a restaurant. Future research should include a question about the typical length of restaurant reviews read. Third, a number of variables deemed important for the theory of planned behavior were not associated with our outcomes. Future research can consider including other theories that may help understand the intentions and behavior of online restaurant review readers and their attendance at restaurants.

 

++++++++++

Conclusion

In conclusion, we found that the demographic variables of race/ethnicity of Hispanic American, Asian/Asian American and Other were significantly associated with increased intentions to use an online restaurant review Web site. Increased trust was associated with increased intentions to use an online restaurant review Web site. However, both race/ethnicity and trust were not associated with visiting restaurants. Also, increased consumer knowledge of review fraud by vendors who can respond to consumer comments was associated with increased visiting of restaurants. The typical reviews on the online restaurant review Web site of UrbanSpoon do not have restaurant manager responses. It is possible that consumer readers of reviews may focus on the negative reviews. Managers of restaurants should consider posting professional responses to explain why the negative experience occurred to the customer who wrote the negative review. This may allow the consumer who reads reviews to be reassured and have increased trust about the restaurant and thus consider visiting the restaurant, regardless of whether the consumer believes that there is increased risk for review fraud. There are international implications regarding the approach to be considered for restaurant managers posting of responses. Managers should realize that many reviews are read by not only local consumers but also business people and tourists from other locations visiting the area. Managers may want to attract interest to those from collectivistic societies by posting responses for positive reviews implying that consumers patronizing their restaurant are those who fit in to the local culture. Also, managers should not ignore negative reviews but rather should post a response, as consumers from authoritarian cultures may be more likely to believe the manager posting the response to a negative review rather than the consumer who posted the negative review. End of article

 

About the authors

Joshua Fogel, Ph.D., is a tenured Professor in the Department of Business Management at Brooklyn College of the City University of New York. His major research focus is on the study of consumer behavior and Internet information topics.
E-mail: joshua [dot] fogel [at] gmail [dot] com

Mohit Kumar, BBA, graduated with a bachelor’s degree from Brooklyn College of the City University of New York. His research interests include consumer behavior and health topics. He is currently a medical student at SUNY Downstate College of Medicine.

 

References

Syed Shah Alam and Nazura Mohamed Sayuti, 2011. “Applying the Theory of Planned Behavior (TPB) in halal food purchasing,” International Journal of Commerce and Management, volume 21, number 1, pp. 8–20.
doi: http://dx.doi.org/10.1108/10569211111111676, accessed 20 April 2017.

Michael Anderson and Jeremy Magruder, 2012. “Learning from the crowd: Regression discontinuity estimates of the effects of an online review database,” Economic Journal, volume 122, number 563, pp. 957–989.
doi: http://dx.doi.org/10.1111/j.1468-0297.2012.02512.x, accessed 20 April 2017.

Kevin Askew, John E. Buckner, Meng U. Taing, Alex Ilie, Jeremy A. Bauer, and Michael D. Coovert, 2014. “Explaining cyberloafing: The role of the theory of planned behavior,” Computers in Human Behavior, volume 36, pp. 510–519.
doi: http://doi.org/10.1016/j.chb.2014.04.006, accessed 20 April 2017.

Soonyong Bae and Taesik Lee, 2011. “Gender differences in consumers’ perception of online consumer reviews,” Electronic Commerce Research, volume 11, number 2, pp. 201–214.
doi: http://doi.org/10.1007/s10660-010-9072-y, accessed 20 April 2017.

Alexander Benlian, Ryad Titah, and Thomas Hess, 2012. “Differential effects of provider recommendations and consumer reviews in e-commerce transactions: An experimental study,” Journal of Management Information Systems, volume 29, number 1, pp. 237–272.
doi: http://doi.org/10.2753/MIS0742-1222290107, accessed 20 April 2017.

BrightLocal, 2013. “Local customer review survey 2013,” at https://www.brightlocal.com/2013/06/25/local-consumer-review-survey-2013, accessed 19 February 2015.

Ting-Pong Vincent Chang, Jo Rhodes, and Peter Lok, 2013. “The mediating effect of brand trust between online customer reviews and willingness to buy,” Journal of Electronic Commerce in Organizations, volume 11, number 1, pp. 22–42.
doi: http://doi.org/10.4018/jeco.2013010102, accessed 20 April 2017.

George Christodoulides, Nina Michaelidou, and Eymorfia Argyriou, 2012. “Cross-national differences in e-WOM influence,” European Journal of Marketing, volume 46, numbers 11–12, pp. 1,689–1,707.
doi: http://dx.doi.org/10.1108/03090561211260040, accessed 20 April 2017.

Dimensional Research, 2013. “Customer service and business results: A survey of customer service from mid-size companies” (April), at https://d16cvnquvjw7pr.cloudfront.net/resources/whitepapers/Zendesk_WP_Customer_Service_and_Business_Results.pdf, accessed 19 February 2015.

Yi-Wen Fan and Yi-Feng Miao, 2012. “Effect of electronic word-of-mouth on consumer purchase intention: The perspective of gender differences,” International Journal of Electronic Business Management, volume 10, number 3, pp. 175–181.

Jillian J. Francis, Martin P. Eccles, Marie Johnston, Anne Walker, Jeremy Grimshaw, Robbie Foy, Eileen F.S. Kaner, Liz Smith, and Debbie Bonetti, 2004. “Constructing questionnaires based on the theory of planned behaviour: A manual for health services researchers,” at http://openaccess.city.ac.uk/1735/, accessed 24 April 2015.

Richard Ghiselli and Jing Ma, 2015. “Restaurant social media usage in China: A study of industry practices and consumer preferences,” Worldwide Hospitality and Tourism Themes, volume 7, number 3, pp. 251–265.
doi: http://dx.doi.org/10.1108/WHATT-03-2015-0014, accessed 20 April 2017.

Namitaz Gupta, 2015. “Urbanspoon to Zomato: D-Day (2 of 4)” (1 June), at http://blog.zomato.com/post/120426713871/urbanspoon-to-zomato-d-day-2-of-4, accessed 14 October 2015.

Nan Hu, Ling Liu, and Vallabh Sambamurthy, 2011. “Fraud detection in online reviews,” Decision Support Systems, volume 50, number 3, pp. 614–626.
doi: http://doi.org/10.1016/j.dss.2010.08.012, accessed 20 April 2017.

Nan Hu, Indranil Bose, Noi Sian Koh, and Ling Liu, 2012. “Manipulation of online reviews: An analysis of ratings, readability, and sentiments,” Decision Support Systems, volume 52, number 3, pp. 674–684.
doi: http://doi.org/10.1016/j.dss.2011.11.002, accessed 20 April 2017.

IBM, 2013. “IBM SPSS statistics for Windows, version 23.0,” at https://www.ibm.com/us-en/marketplace/spss-statistics, accessed 20 April 2017.

Mohammad Reza Jalilvand and Neda Samiei, 2012a. “The effect of electronic word of mouth on brand image and purchase intention: An empirical study in the automobile industry in Iran,” Marketing Intelligence & Planning, volume 30, number 4, pp. 460–476.
doi: http://dx.doi.org/10.1108/02634501211231946, accessed 20 April 2017.

Mohammad Reza Jalilvand and Neda Samiei, 2012b. “The impact of electronic word of mouth on a tourism destination choice: Testing the theory of planned behavior (TPB),” Internet Research, volume 22, number 5, pp. 591–612.
doi: http://dx.doi.org/10.1108/10662241211271563, accessed 20 April 2017.

Matthew L. Jensen, Joshua M. Averbeck, Zhu Zhang, and Kevin B. Wright, 2013. “Credibility of anonymous online product reviews: A language expectancy perspective,” Journal of Management Information Systems, volume 30, number 1, pp. 293–324.
doi: http://dx.doi.org/10.2753/MIS0742-1222300109, accessed 20 April 2017.

Kevin K.Y. Kuan, Kai-Lung Hui, Pattarawan Prasarnphanich, and Hok-Yin Lai, 2015. “What makes a review voted? An empirical investigation of review voting in online review systems,” Journal of the Association for Information Systems, volume 16, number 1, article number 1, at http://aisel.aisnet.org/jais/vol16/iss1/1, accessed 20 April 2017.

Jiunn-Woei Lian and David C. Yen, 2014. “Online shopping drivers and barriers for older adults: Age and gender differences,” Computers in Human Behavior, volume 37, pp. 133–143.
doi: http://dx.doi.org/10.1016/j.chb.2014.04.028, accessed 20 April 2017.

Kwek Choon Ling, Lau Teck Chai, and Tan Hoi Piew, 2010. “The effects of shopping orientations, online trust, and prior online purchase experience toward customers’ online purchase intention,” International Business Research, volume 3, number 3, at http://www.ccsenet.org/journal/index.php/ibrarticle/view/6506, accessed 20 April 2017.

Xianghua Lu, Sulin Ba, Lihua Huang, and Yue Feng, 2013. “Promotional marketing or word-of-mouth? Evidence from online restaurant reviews,” Information Systems Research, volume 24, number 3, pp. 596–612.

Michael Luca and Georgios Zervas, 2016. “Fake it till you make it: Reputation, competition and Yelp review fraud,” Management Science, volume 62, number 12, pp. 3,412–3,427.
doi: http://dx.doi.org/10.1287/mnsc.2015.2304, accessed 20 April 2017.

Stephan Ludwig, Ko de Ruyter, Mike Friedman, Elisabeth C. Brüggen, Martin Wetzels, and Gerard Pfann, 2013. “More than words: The influence of affective content and linguistic style matches in online reviews on conversion rates,” Journal of Marketing, volume 77, number 1, pp. 87–103.
doi: http://dx.doi.org/10.1509/jm.11.0560, accessed 20 April 2017.

W. Glynn Mangold and Katherine Taken Smith, 2012. “Selling to millenials with online reviews,” Business Horizons, volume 55, number 2, pp. 141–153.
doi: http://doi.org/10.1016/j.bushor.2011.11.001, accessed 20 April 2017.

Aurelio G. Mauri and Roberta Minazzi, 2013. “Web reviews influence on expectations and purchasing intentions of hotel potential customers,” International Journal of Hospitality Management, volume 34, pp. 99–107.
doi: http://doi.org/10.1016/j.ijhm.2013.02.012, accessed 20 April 2017.

Cheol Park and Thae Min Lee, 2009. “Antecedents of online reviews’ usage and purchase influence: An empirical comparison of U.S. and Korean consumers,” Journal of Interactive Marketing, volume 23, number 4, pp. 332-340.
doi: http://doi.org/10.1016/j.intmar.2009.07.001, accessed 20 April 2017.

Sarah Perez, 2012. “UrbanSpoon: Traffic up 80% in 2011, mobile growth faster than Web,” TechCrunch (7 February), at http://techcrunch.com/2012/02/07/urbanspoon-traffic-up-80-in-2011-mobile-growth-faster-than-web, accessed 5 February 2015.

Enrique Bonsón Ponte, Elena Carvajal-Trujillo, and Tomás Escobar-Rodríguez, 2015. “Influence of trust and perceived value on the intention to purchase travel online: Integrating the effects of assurance on trust antecedents,” Tourism Management, volume 47, pp. 286–302.
doi: http://doi.org/10.1016/j.tourman.2014.10.009, accessed 20 April 2017.

Nathalia Purnawirawan, Patrick de Pelsmacker, and Nathalie Dens, 2012. “Balance and sequence in online reviews: How perceived usefulness affects attitudes and intentions,” Journal of Interactive Marketing, volume 26, number 4, pp. 244–255.
doi: http://doi.org/10.1016/j.intmar.2012.04.002, accessed 20 April 2017.

Yoo-Kyoung Seock and Jan M. Hathcote, 2010. “A cross-cultural comparison of Hispanic American and White American adolescents’ use of reference agents for apparel shopping,” Family & Consumer Sciences Research Journal, volume 39, number 1, pp. 45–56.
doi: http://doi.org/10.1111/j.1552-3934.2010.02044.x, accessed 20 April 2017.

Kendall L. Short, 2013. “Buy my vote: Online reviews for sale,” Vanderbilt Journal of Entertainment & Technology Law, volume 15, number 2, pp. 441–471, and at http://www.jetlaw.org/journal-archives/volume-15/volume-15-issue-2/buy-my-vote-online-reviews-for-sale/, accessed 20 April 2017.

Xiao Tong, 2010. “A cross-national investigation of an extended technology acceptance model in the online shopping context,” International Journal of Retail & Distribution Management, volume 38, number 10, pp. 742–759.
doi: http://dx.doi.org/10.1108/09590551011076524, accessed 20 April 2017.

UrbanSpoon, 2015. “Find the best restaurants, cafés, and bars,” at http://www.urbanspoon.com, accessed 19 February 2015.

Katharina S. Valta, 2013. “Do relational norms matter in consumer-brand relationships?” Journal of Business Research, volume 66, number 1, pp. 98–104.
doi: http://doi.org/10.1016/j.jbusres.2011.07.028, accessed 20 April 2017.

Qian Xu, 2014. “Should I trust him? The effects of reviewer profile characteristics on eWOM credibility,” Computers in Human Behavior, volume 33, pp. 136–144.
doi: http://doi.org/10.1016/j.chb.2014.01.027, accessed 20 April 2017.

Kem Z.K. Zhang, Christy M.K. Cheung, and Matthew K.O. Lee, 2014. “Examining the moderating effect of inconsistent reviews and its gender differences on consumers’ online shopping decision,” International Journal of Information Management, volume 34, number 2, pp. 89–98.
doi: http://doi.org/10.1016/j.ijinfomgt.2013.12.001, accessed 20 April 2017.

Ziqiong Zhang, Qiang Ye, Rob Law, and Yijun Li, 2010. “The impact of e-word-of-mouth on the online popularity of restaurants: A comparison of consumer reviews and editor reviews,” International Journal of Hospitality Management, volume 29, number 4, pp. 694–700.
doi: http://doi.org/10.1016/j.ijhm.2010.02.002, accessed 20 April 2017.

 


Editorial history

Received 13 December 2016; accepted 21 April 2017.


Copyright © 2017, Joshua Fogel and Mohit Kumar. All Rights Reserved.

Intentions to use an online restaurant review Web site and purchase behavior after reading reviews
by Joshua Fogel and Mohit Kumar.
First Monday, Volume 22, Number 5 - 1 May 2017
http://www.firstmonday.dk/ojs/index.php/fm/article/view/7250/6174
doi: http://dx.doi.org/10.5210/fm.v22i15.7250





A Great Cities Initiative of the University of Illinois at Chicago University Library.

© First Monday, 1995-2017. ISSN 1396-0466.