Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
Research article
Customer review or influencer endorsement: which one influences purchase
intention more?
Diena Dwidienawati
a
,
*
, David Tjahjana
b
, Sri Bramantoro Abdinagoro
c
, Dyah Gandasari
d
,
Munawaroh
e
a
Management Department, BINUS Business School Undergraduate Program, Bina Nusantara University, Indonesia
b
Universitas Multimedia Nusantara, Indonesia
c
DRM, BINUS Business School, Bina Nusantara University, Indonesia
d
POLBANGTAN Bogor, Indonesia
e
Swiss German University, Indonesia
ARTICLE INFO
Keywords:
Customer review
Influencer
Trust
Purchase intention
Social media
Technology adoption
Business management
Marketing
Digital media
ABSTRACT
E-commerce has grown steadily since internet access became more available in the mid-1990s. Informativeness
plays a key role in online shopping decisions. Potential customers usually collect useful information and do a
comparison before considering the purchase. Electronic word-of-mouth (eWOM) is considered a reliable source of
information. Customer reviews and influencer reviews can be considered eWOM. They represent customers’
sharing of experience and evaluation of a product or service with other potential shoppers. There is abundant
evidence concerning the influence of eWOM on purchase intention. However, there are few studies on influencer
reviews and purchase intention. This study aims to investigate the impact of customer review and influencer
review to purchase intention and the mediating role of trust to those relationship. A quantitative experimental
study (2 1) was conducted. Two hundred respondents from three cities in Greater Jakarta were divided into two
groups to self-rate their opinion on customer review, influencer review, trust and purchase intention. Data
collected was analysed PLS using SmartPLS. The study results showed that influencer review has a positive impact
on purchase intention. On the other hand, customer review failed to show its influence. Trust as a moderating
variable was also not validated in this study.
1. Introdution
E-commerce has grown steadily since internet access became more
available in the mid-1990s. In 2019 in the US, the share of e-commerce
was 11.1% of total retail, up from 5.8% in 2013 [1] and expected to
increase to 13.7% by 2021. In Indonesia, the share of e-commerce of total
retail is only 3% [2]. Considering the preference of the younger customer
for online shopping, the e-commerce sector in Indonesia is expected to
grow significantly.
Informativeness plays a key role in online shopping decisions [3].
Potential customers collect useful information and compare it before
deciding on the purchase [3], and electronic word-of-mouth (eWOM) is
considered a good source of information. Customer reviews and influ-
encer endorsements can both be considered eWOM because they repre-
sent customers sharing their experience evaluation of a product or service
with other potential shoppers [4]. When customers share their good
experiences, it shows customer satisfaction with the product [5]. Liter-
ature on eWOM has generally used customer reviews as an example of
eWOM. The influence of eWOM on purchase intention has been proved
by various authors [6,7,8,9]. Customer reviews play an important role
in online shopping decision-making. Ninety-one percent of customers
claimed that they read customer reviews before making purchasing de-
cisions [10].
On the other hand, the literature on the effectiveness of influencer
endorsements on purchase intention are still limited. While the idea of
someone ‘influential’might be convincing, there is very little evidence
showing that influencers actually improve performance [11]. There are
limited scholarly research examining the relationship between influ-
encers and performance [12]. In December 2019, a search in Science
Direct for the keywords ‘Influencer’and ‘Purchase Intention’for all years
only yielded 32 results, while there were only seven results in Google
Scholar for all years for the same search period and similar keywords.
* Corresponding author.
E-mail address: diena.tjiptadi@gmail.com (D. Dwidienawati).
Contents lists available at ScienceDirect
Heliyon
journal homepage: www.cell.com/heliyon
https://doi.org/10.1016/j.heliyon.2020.e05543
Received 21 May 2020; Received in revised form 11 July 2020; Accepted 16 November 2020
2405-8440/©2020 Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Heliyon 6 (2020) e05543
Despite the limited research available on influencer endorsements,
marketers in many organisations plan to spend heavily on influencers
[13]. This study aims to confirm the influence of customer reviews on
purchase intention and provide evidence of the relationship between
influencer endorsements and purchase intention. The other goal of this
study is to reconfirm the moderating effect of trust in the reviewer on
purchase intention. The premise is that if the customers trust the
reviewer, both for customer reviews and influencer endorsements, they
are more likely to have a purchase intention.
This paper is structured as follows: the next section presents a liter-
ature review on signalling theory, eWOM, customer reviews, influencer
endorsements, trust and hypothesis development. Thereafter, the meth-
odology of the empirical study is described and the findings are pre-
sented. Finally, the conclusion, limitations and recommendations for
further research are given.
2. Literature review
2.1. Signalling theory
Signalling theory has become increasingly popular in strategic
management studies [14]. This theory focuses on the issue of
decision-making and how the parties involved use signals to reduce
uncertainty related to making decisions in cases of incomplete and
asymmetrical information [14]. In signalling theory, the signal can be
both negative and positive. Those signals are important for the receivers
in making decisions.
Figure 1. Research framework.
Figure 2. Path analysis of structural model (bootstrapping) - influencer review.
Figure 3. Path analysis of structural model (bootstrapping) - customer review.
D. Dwidienawati et al. Heliyon 6 (2020) e05543
2
The birth of signalling theory can be traced back to the seminal work
of Spence in 1973 [15] on the labour market, where he introduced
asymmetric information for decision-making in an economic model.
Signalling theory is concerned with reducing the effect of asymmetric
information between two parties [16]. According to signalling theory,
there are three elements: signaller, signal and receiver [17]. The signaller
is the individual who obtains information about an individual, product or
organisation. The signal is the information itself, which can be both
positive and negative. The receiver is the third element in the signalling
timeline. The receiver is the one who lacks information about an indi-
vidual, product or organisation [16].
In online shopping, potential customers are faced with perceived risks
and uncertainties. They cannot evaluate the products directly and
physically. The availability of other parties who can share their experi-
ences, evaluations and perceptions of products or services in the form of
eWOM information can help potential customers in reducing asymmetric
information. This will reduce potential customer perceived risk and in-
crease potential customers’confidence in making purchase decisions.
2.2. E-WOM
WOM is defined as direct communication from person-to-person of an
opinion about a product and/or service [18]. The diffusion of WOM is an
important mechanism for information to reach a large population,
thereby possibly influencing 1) public opinion; 2) adoption of innova-
tion; and 3) the market share of a new product or brand awareness [19].
WOM is important for customers’buying decisions [4]. EWOM has been
reported to have greater power than traditional WOM [5] because it
reaches a large population rapidly [20].
With the increasing trend of online shopping, since consumers cannot
evaluate the product directly and physically, eWOM has become a
prominent source of information for potential customers [20,21].
Informativeness is one of the most significant factors influencing a
shopper's decisions with respect to online purchases [3]. The online
consumer would like to collect and compare useful information when
considering a product purchase, and informativeness aids the consumer
in evaluating alternative products and in making the best choice [3].
EWOM is differentiated from traditional WOM by the online context as a
way of exchanging information regarding the usage and characteristics of
products and services [4]. It differs from the traditional oral form of
interpersonal communication among friends or acquaintances [10],
which is private in nature, in that eWOM can be shared with others
openly [22].
EWOM is defined as internet-mediated opinions and recommenda-
tions on products and services from peers [23]. EWOM can be “any
positive or negative statement made by potential, actual or former customers
about a product or company that is made available to a multitude of people
and institutions via the internet”[4]. It is an informal communication from
one customer to other potential customers about the usage, ownership, or
characteristics of certain goods and services or their sellers [24].
For online purchasing, eWOM plays an important role in helping
customers make purchasing decisions while considering the perceived
risks involved [10]. EWOM serves as a decision-making tool. Many po-
tential online shoppers wait and observe others before making such de-
cisions, and information from others is known to increase consumer
confidence [24].
Previous studies have shown the consequences of eWOM [10], argued
that positive eWOM can strengthen the relationship between consumer
trust and intention to shop online. Setiawan et al. (2014) showed that
eWOM has an indirect effect on satisfaction and loyalty, mediated by
destination images in the tourism industry [25], while Yoo et al. (2013)
showed a direct influence of eWOM on loyalty [26]. The direct influence
of eWOM on purchase intention has been reported by various authors. Di
Pietro et al. (2012) in [13] showed the relationship between eWOM and
intention to choose a travel destination. EWOM's direct influence on
purchase intention has also been reported by [6,7,8,9]. Other conse-
quences of eWOM include trust in e-commerce. EWOM information has
been reported by [27] to have a positive and significant impact on the
trust of the seller. Trust and online reputation are influenced by eWOM,
as reported by Anderson and Barton, 1989. Credible and trusted eWOM
influences the reputation of e-commerce [28]. Customer reviews and
influencer endorsements can be considered electronic word-of-mouth
(eWOM) because they represent customers sharing their experience
and evaluation of a product or service with other potential shoppers [4].
2.3. Customer reviews
Electronic customer reviews are defined as peer-generated product
evaluations posted on company or third-party websites [29]. Online
shopping, just like traditional shopping, is a social activity. People's
purchase decisions tend to be influenced by their interactions with others
[30]. Online customer reviews can reduce the risks perceived by con-
sumers [10] and improve their degree of satisfaction [24], as well as their
efficiency in making decisions [24]. During their online search for a
product, consumers encounter dozens or even hundreds of pieces of
product information and alternatives. Customer reviews give people
more reason for a decision and increase the decision-maker's confidence.
These customer reviews provide additional information, expert reviews
and personalised advice, which can add value to potential customers
[29]. [10] showed that 91% of participants read customers' reviews
before purchasing a new product or service, while 46% of the partici-
pants in that study said that their decisions were influenced by those
comments. Seventy-seven per cent of online shoppers in the US reported
using consumer-generated reviews and ratings to help them make
decisions.
Online reviews are considered credible because the contents are real
users' own experiences with the products or services. The users are
perceived to have no vested interest and no intention to deceive the
readers [31]. Longer reviews that include details of the products and
specific information about the products increase the quality of the re-
view. The number of peer reviews can also reduce uncertainty concern-
ing the product's quality, while consistency between comments from one
user to another improves the credibility of the review [24]. Customer
reviews have been known to improve sales significantly [24] and to in-
crease the credibility of the website. Reviews make consumer visits more
attractive and increase the time spent on the website [29]. They also
enhance consumer confidence in the product [10].
2.4. Influencer endorsements
Influencer marketing is the most important new approach to mar-
keting. Influencer marketing is a type of marketing that focuses on
certain key individuals rather than the target market as a whole, because
these individuals have been identified as prospective buyers. Thus,
influencer marketing involves marketing activities around these “influ-
encers”[32].
Social and influencer marketing are arguably the hottest concepts in
online advertising at the moment [33]. Most organisations that advertise
online must have a “social”strategy [34]. Deploying influencers, such as
Instafamous people, for branding has become an important aspect of
social media marketing campaigns [35]. The trend of the “Influencer
Marketing Goldrush”is about to explode in popularity as a marketing
tactic. A recent survey of 1,300 marketers found that 74% planned to
invest in influencers over the next 12 months. According to Hubspot,
71% of consumers were more likely to make a purchase when the product
was mentioned in social media, and 92% would trust an influencer's
marketing review of a brand when making a purchasing decision [13].
D. Dwidienawati et al. Heliyon 6 (2020) e05543
3
An influencer is defined as “someone who exhibits some combination of
desirable attributes—whether personal attributes like credibility, expertise, or
enthusiasm, or network attributes such as connectivity or centrality—that al-
lows them to influence a disproportionately large number of others”[19].
Influencers are at the forefront of social trends. They could be innovators
who create new ideas, concepts or content that regularly grab the
attention of social media. They could be early adopters or those who
discover trends before anyone else, and enliven them with their own
creativity, spreading them further on social media [33]. They are in-
dividuals who are effective in spreading messages about new products,
starting and popularising new trends and driving up sales [12]. Influ-
encer marketing is the process of developing a relationship between such
influential individuals and potential buyers [32].
Influencers are known to have high numbers of followers, which
enables them to have a wide reach [35]. They are considered by their
followers to have learnt about different topics and products. With their
popularity and their role as opinion leaders, influencers are more likely to
influence sales. A Marketing Dive study stated that 41% of marketers
claimed that marketing campaigns using influencers were more suc-
cessful than traditional ones [33].
2.5. Trust
Trust plays an important role in all transactions, considering the un-
certainty and the risk involved. In e-commerce, trust is crucial and one of
the most influential factors [30]. Consumers are unlikely to conduct an
online transaction if they do not trust the seller [36]. Trust is defined as
“the willingness of a party to be vulnerable to the actions of another party
based on the expectation that the trustee will perform a particular action
important to the trust, irrespective of the ability to monitor or control that other
party”[37].
In e-commerce, trust is a belief in the good faith of sellers that con-
sumers have after reviewing their characteristics [36]. Another definition
of trust in the context of e-commerce is as a subjective belief that the
seller will fulfil their obligations, as those obligations are understood by
the consumer [36]. Trust is a “willingness to rely on an exchange partner
(i.e., reliable person who keeps promises [38]. Trust can be bestowed upon a
person, a product, an organisation, an institution or a role [9]. In this
study, trust is a subjective belief of potential customer to a person who
provides customer review or endorsement.
Trust has been known to facilitate business transactions between two
parties who are lack of prior experiences in mutual confidence. It means
not only reducing perceived risk, but also enhancing the customer's
perceived value. Trust has a moderating effect on the process and
behaviour [39] and it can help to reduce the anxiety, vulnerability and
uncertainty that may be caused by the transaction, resulting in greater
satisfaction. Trust can create a positive attitude towards transaction
behaviour, which will lead to transaction intention [30], and is
important in creating expected and satisfying results in online trans-
actions [9]. It has been shown in various studies that trust positively
influences customers' online purchase intentions, and the higher the
degree of consumers' trust, the higher the degree of consumers' purchase
intention [9]. Trust in e-reviews also has a positive impact on choice
[28,40].
2.6. Hypothesis development
Customer reviews and influencer endorsements are part of electronic
word-of-mouth (eWOM) because they represent customers who are
sharing their experiences and evaluations of products or services with
other potential shoppers [4]. Peer-generated reviews have been known to
help potential customers in making purchase decisions. They reduce the
perceived risk and increase confidence in and attitude towards the
products [10].
Positive eWOM can strengthen the relationship between a consumer's
trust and intention to shop online [10]. Bokunewicz and Shulman (2016),
referring to Di Pietro et al. (2012) stated that there is a positive rela-
tionship between eWOM and intention to choose travel destinations. The
authors in [6,7,8] and [9] agreed that eWOM has a direct influence on
purchase intentions (see Figure 1).
H1: Customer reviews have a positive influence on purchase intention
Influencers have a high number of followers. High numbers of fol-
lowers enable influencers to have a wide reach [35]. With high likability
and their role as opinion leaders, influencers are more likely to influence
sales. A Marketing Dive study stated that 41% marketers claimed that
marketing campaigns using influencers are more successful than tradi-
tional ones [33].
H2: Influencer endorsement has a positive influence on purchase
intention
Online consumers tend to gather information before making decisions
because learning about the products and services from other consumers’
experiences could reduce the perceived risk and increase the confidence
of online shoppers. Customer reviews and influencer endorsements are
considered trusted and credible sources of information.
Trust plays an important aspect in the transaction, considering the
uncertainty and risk involved. In e-commerce, trust is crucial and one of
the most influential factors [30]. Consumers are unlikely to make an
online transaction if they do not trust the seller [36]. Trust can create a
positive attitude toward transaction behaviour, which leads to trans-
action intention [30].
The resulting hypotheses are:
H3: Trust moderates customer review, leading to purchase intention
H4: Trust moderates influencer endorsement, leading to purchase
intention
3. Methodology
3.1. Research design
This study carried out experiments with a two (customer review vs
influencer endorsement) by one (purchase intention) between-subjects
design (2 1). Two hundred respondents were assigned to one of two
conditions. One hundred respondents received the customer review and
the other 100 had the influencer endorsement. A purposive sampling
method was used for an online survey via respondents’mobile phones.
Each participant was identified with an individual number. All re-
spondents were university students from Jakarta, Bogor and Tangerang.
They represented young consumers.
3.2. Experimental stimulus product
The smartphone was selected as the experimental stimulus for two
reasons. First, the respondents were interested in knowing other con-
sumers' views on the target product. Second, smartphones appealed to
and were therefore readily accessible and purchased by the re-
spondents. The present research, therefore, selected a smartphone that
met those two conditions. Smartphone penetration is very high in
Indonesia and thus consumers often purchase this product. Moreover,
previous studies also refer to online reviews of high-end products, such
as smartphones, as a relevant stimulus product. In the current study, the
smartphone had a fictional brand name of Smart Ace. The study's design
was based on the experimental study done by [31]. In this study, all the
customers' names, influencers, website names and brand names used
were fictitious. The study was an academic study, not for commercial
purposes, which was also indicated at the beginning of the
questionnaires.
D. Dwidienawati et al. Heliyon 6 (2020) e05543
4
3.3. Online consumer review and influencer endorsement
Various review comments posted on company and consumer websites
create an online customer review. Based on the preliminary study,
customer reviews are more trusted if pictures are posted, there are many
reviews and there is a mixture of negative and positive reviews. There-
fore, a fictitious list of customer reviews was designed, which included
those criteria (Appendix 1).
Fictitious influencers were designed by mimicking two influencers
(one male and one female) with more than 100,000 followers on Insta-
gram. One of their typical product reviews was copied, but the attributes
were changed to suit the smartphone (Appendix 2).
3.4. Procedures and participants
The experiment was online and sent to participants directly via WA
application, which is one of the most popular communication applica-
tions in Indonesia. The participants were 200 students at four Indonesian
universities, one in Jakarta, one in Bogor and two in Tangerang. Selection
of the respondents was done using the purposive sampling method. There
were five lecturers from four universities involved in this study. Each
lecturers had to have two classes. Each class was assigned to one treat-
ment. For example, lecturer A had two classes, Class X and Class Z. All
students from Class X were asked to respond to the Influencer Ques-
tionnaire and all students from Class Z were asked to respond to the
Customer Review Questionnaire. Each university had 50% male re-
spondents and 50% female respondents, with 25% male customer re-
views and 25% female customer reviews, and 25% male influencer
endorsements and 25% female influencer endorsements.
For customer reviews, participants were asked to imagine they were
searching for information about a smartphone, and the web pages they
searched had a consumer-generated information aggregator (htt
p://gadgetplace.com) for the product called Smart Ace (Appendix 1).
For the influencer endorsements, participants were invited to imagine
visiting Instagram pages of an influencer to find their review of the
product (Appendix 2).
In part one, after reading the first page of the experimental website
or Instagram page, in which the fictitious customer reviews and influ-
encer endorsements were shown, the participants were asked to give
information about their gender, year of birth, campus location and
hometown. Afterward, respondents were guided to part two of the
survey, which related to variable statements. Participants were asked to
rate the reviews concerning usefulness, trust and intention to purchase
the reviewed product, attitudes, items for manipulation check, and
other online review-related questions. All content of the questionnaires
used casual Bahasa Indonesia, which is common language used by
younger people.
3.5. Measurements
All measurement scales used in this study (Table 1), except items used
in the manipulation check and demographics, used a framework from a
previous study. Customer review and influencer endorsement measures
used four items, each modified from [41], trust measures used three
items modified from [36] and purchase intention measures used four
items modified from [42].
A six-point Likert scale (from 1 strongly disagree to 6 strongly agree)
was used for respondentsto rate their opinion. With a six-point Likert scale,
the mid-point is omitted to avoid a social desirability bias [43]. Additional
demographic information, such as age, gender, educational background
and respondent's origin, was also requested for descriptive analysis.
All data collected were analysed using descriptive analysis and PLS
with SmarPLS. There are two stages of analysis. The first stage is the outer
model, which is to determine the validity and reliability of each research
indicator. The second stage is the inner model, which aims to determine
the relationship between latent variables.
4. Results
4.1. Descriptive analysis
This study compared the two groups, which were Influencer
Endorsement and Customer Review. From a total of 187 participants, 104
were in the Influencer Endorsement group and the Customer Review
group. The majority (98.4%) of respondents were born between 1995
and 2005 (Generation Z). They were 49.7% male and 48.1% female.
Ninety-two per cent of respondents were university students. Their
campuses included Jakarta (12.3%), Tangerang (39.6%), Bogor (39.6%)
and others (5.9%) (see Table 2).
The descriptive analysis showed that, for all questions, the average
score was more than three out of six. The highest average was the
Influencer Endorsement, with an average of 4.63. The lowest score was
Purchase Intention from Customer Reviews at 3.46 (see Table 3).
4.2. Measurement model analysis
This study aimed to examine the effect of influencer endorsements
and customer reviews on trust moderated purchase intention. Data were
Table 1. Measures.
Customer Reviews CR01: I often read customer review to know product impression by others
(Jalilvand, 2012) CR02: To make sure I buy the right product, I often read customer review
CR03: I frequently gather information from customer review to help me choose the right product
CR04: When I buy a product, customer review make me confident in purchasing the product
Influencer Endorsement IR01: I often read influencer review to know product impression by others
(Jalilvand, 2012) IR02: To make sure I buy the right product, I often read influencer review
IR03: I frequently gather information from influencer review to help me choose the right product
IR04: When I buy a product, influencer review make me confident in purchasing the product
Trust TR01: This product review has integrity
(Ponte, 2015) TR02: This product review is reliable
TR03: This product review is trustworthy
Purchase Intention PI01: After reviewing the comment, the likelihood of purchasing this smartphone is high
(Lien, 2015) PI02: If I am going to purchase smartphone, I would consider this smartphone
PI03: The probability that I would consider purchasing this smartphone is high
PI04: My willingness to purchase this smartphone is high
D. Dwidienawati et al. Heliyon 6 (2020) e05543
5
Table 2. Respondent characteristic.
Respondent Characteristic Influencer Review Customer Review Total
Number % Number % Number %
Year of Birth 1995–2005 103 99 81 97,6 184 98,4
Gender Male 55 52,9 38 45,8 93 49,7
Female 47 45,2 43 51,8 90 48,1
No Answer 2 1,9 2 2,4 4 2,1
Educational Background High School Graduate 3 2,9 2 2,4 5 2,7
University Students 94 90,4 78 94 172 92
University Graduate 7 6,7 3 3,6 10 5,3
Campus location, if student Jakarta 10 9,6 13 15,7 23 12,3
Tangerang 44 42,3 30 36,1 74 39,6
Bekasi 0 0 2 2,4 2 1,1
Bogor 40 38,5 34 41 74 39,6
Others 5 4,8 4 4,8 9 4,8
Place of Origin Jabodetabek 61 58,7 - - 61 32,6
Others 43 41,3 - - 43 23
No Answer - - 83 100 83 44,4
Total 104 100 83 100 187 100
Table 3. Descriptive analysis of variables.
Questions Influncer Review Customer Review
Trust to Influencer/Customer Review 4,34 3,64
Influencer/Customer Review 4,63 4,13
Purchase Intention 4,13 3,46
Table 4. Measurement model analysis.
Latent Variable Indicator Loading Fcator T-Stat AVE Remark CR Remark
(>0,5) (>0,5) (>0,7)
Influencer Review
Trust to Influencer Review IFTR01 0,886 25,261 0,82 Valid 0,932 Reliable
IFTR02 0,937 69,836 Valid
IFTR03 0,893 31,774 Valid
Influencer Review IF01 0,819 14,177 0,751 Valid 0,9 Reliable
IF02 0,899 32,383 Valid
IF03 0,88 37,516 Valid
Purchase Intention PI01 0,702 10,685 0,601 Valid 0,883 Reliable
PI02 0,827 19,497 Valid
PI03 0,723 9,963 Valid
PI04 0,812 14,023 Valid
PI05 0,806 16,415 Valid
Customer Review
Trust to Customer Review CRTR01 0,812 11,234 0,766 Valid 0,907 Reliable
CRTR02 0,883 21,14 Valid
CRTR03 0,927 45,939 Valid
Customer Review CR01 0,865 3,984 0,778 Valid 0,913 Reliable
CR02 0,883 3,805 Valid
CR03 0,898 3,876 Valid
Purchase Intention PI02 0,881 25,971 0,775 Valid 0,932 Reliable
PI03 0,815 10,909 Valid
PI04 0,921 43,065 Valid
PI05 0,899 29,668 Valid
D. Dwidienawati et al. Heliyon 6 (2020) e05543
6
processed using structural equation modelling with partial least square
(PLS), which is a variant-based alternative. In structural equation
modelling, PLS uses two stages of evaluation. The first stage is the outer
model, which is to determine the validity and reliability of each research
indicator. The second stage is the inner model, which aims to determine
the relationship between latent variables.
All indicators from all related variables showed loading factors >0.5.
The composite reliability (CR) for all variables was also >0.7. The AVE
score for all variables was also >0.5. Therefore, it can be concluded that
the measurements were reliable and valid (Table 4). Discriminant val-
idity test with a Fornell-Larcker criterion showed that the root square of
AVE of all variables was the highest between variables, compared to
other variables. This showed that all tested variables had good discrim-
inant validity.
4.3. Structural model analysis
In this study, several structural model tests were carried out, namely
prediction relevance (Stone-Geisser's Q
2
) and R-square test. The Q2 test
result showed that Q
2
for all relationships between variables was more
than zero. This was evidence that the model had predictive relevance.
The R-square of the Influencer group was 0.505. This shows that the
Influencer Endorsement variable moderated by Trust in Influencer
Endorsement had an influence of 50.5% on Purchase Intention, while the
remaining 49.5% was influenced by other variables. The coefficient of
determination (R-square) found in the Customer group was 0.323, which
shows that the Customer Review variable moderated by Trust in
Customer Review had an influence of 32.2% on Purchase Intention, while
the remaining 67.7% was influenced by other variables (Figures 2and 3).
4.4. Hypothesis testing
Table 5 shows that the influence of Influencer Endorsement on Pur-
chase Intention had T-Sat 6.299 and P-value 0.000. Both figures were
higher than the standard reference; therefore, H1 was accepted. The in-
fluence of Customer Review on Purchase Intention had T-Stat 0.357,
which was lower than T-table (1.96), and P-value 0.722, which was
higher than 0.05. This showed that Customer Review did not signifi-
cantly influence purchase intention. Therefore, H2 was rejected. The
analysis between groups showed that T-value was 3.802, which was
higher than T-table 1.96. Therefore, it can be concluded that the two
groups were different.
Hypothesis testing for the moderating effect of Trust in Influencer
Endorsement to Purchase Intention showed that T-Stat was 0.881 and P-
value 0.379. Therefore, H3 was rejected. T-Stat and P-value of the
moderating effect of trust in Customer Review to Purchase Intention were
0.001 and 0.999 respectively. H4 was rejected based on those results.
Furthermore, the effect of trust was tested to see whether it had a
direct effect on purchase intention. Trust, both in Influencer Endorse-
ment and Customer Review, showed T-Stat higher than T-table (T-Stat
was 3.789 for Trust in Influencer Endorsement and 5.481 for Trust in
Customer Review) and the P-value was <0.05 (P-value ¼0.000 for both
Trust in Influencer Review and Trust in Customer Review). From the
above hypothesis testing, it can be seen that trust does not have a
moderating effect, but it has a direct effect. Therefore, it can be
concluded that trust is a predictor variable, not a moderator.
5. Discussion
The results of this study provide further evidence that eWOM in-
fluences purchase intention. This shows that in online shopping decision-
making, information from ‘experienced’customers is considered valu-
able by potential customers. In online shopping, eWOM plays an
important role in reducing the asymmetric information of potential
customers. Previous studies showed that eWOM had a positive influence
on purchase intention [10]. and [24] have shown how eWOM helps
customers by reducing uncertainty and increasing their purchase confi-
dence. The evaluations of ‘experienced’customers can reduce the un-
certainty and perceived risk, and therefore eWOM influences
decision-making for purchases.
However, this study also showed that not all eWOM influences pur-
chase intention. Influencer endorsement has a positive impact on pur-
chase intention, but the customer review failed to show an influence on
purchase intention. There are several explanations as to why the
customer review did not significantly influence purchase intention. The
first issue is trust. Information is valued based on the reliability. One of
the source of reliability is trusted source. In this study, the information
that participants got from other customers’reviews seemed not to be very
well trusted. The average score for trust in customer review was only 3.64
out of 6, unlike score for trust from influencer which was higher (4.3).
This means that most of the participants only slightly agreed that they
trusted the customer reviews [12]. argued that a feeling of closeness
creates a strong attachment, which leads to higher trust. In this study, it
seems that the participant “customers”could not really relate or closely
attach to the other customers who had written the reviews; therefore, any
information given did not increase their confidence enough to purchase.
The other explanation relates to the information itself. Informative-
ness is considered the most significant factor influencing customers' de-
cisions in online shopping [3]. To become valuable, information should
be of a high quality. High-quality information may help customers to
know a product better and make better decisions [3]. In online customer
reviews, quality is often the issue. Most of the time, customers only write
a short, unspecific review. Many will only comment with ‘like’and give a
rating without any further explanation [29]. confirmed that a review that
is deep, long and specific helps to increase confidence.
The last possible explanation relates to the product itself. In this
study, a premium mobile phone was used as the target product. It was
considered expensive for most of the respondents in this study. To spend
a lot of money, getting full information is important [44]. also confirmed
that the types of products influence customers’attitudes toward online
shopping. The more expensive the product is, the more risk the customer
is taking. Consumers need more information to reduce their uncertainty,
and customer reviews alone are not enough as an information source to
support their decision-making.
Unlike a customer review, it seems that an influencer endorsement
has a positive influence on purchase intention. An influencer has a high
number of followers; therefore, their message will reach a large number
of people [35]. stated that this wide reach can leverage the power of
WOM and, at the same time, also create a huge amount of discussion.
‘Group’opinion and consensus can be informally built. The high number
of followers also influences the level of this likeability. When the majority
of group members ‘like’a post, it influences the opinion of other fol-
lowers, which will become more similar to the rest of the group.
Table 5. Hypothesis testing –influencer and customer review to purchase intention.
Relationship Path T-Stat P Values T total Conclusion
Influencer Review->Purchase Intention 0,535 6,299 0,000 1,96 H1 accepted
Customer Review->Purchase Intention -0,045 0,357 0,722 1,96 H2 rejected
Moderating Effect 1 (Influencer Review*Trust) ->Purchase Intention 0,073 0,881 0,379 1,96 H3 rejected
Moderating Effect 1 (Customer Review*Trust) ->Purchase Intention 0 0,001 0,999 1,96 H4 rejected
D. Dwidienawati et al. Heliyon 6 (2020) e05543
7
Furthermore, influencer endorsement also involves the role of lead-
ership. When a group of people consider someone as ‘their’leader, they
will trust this person. Trust has been known to be a determinant of
decision-making. In this case, the perceived leadership role borne by the
influencer fosters a level of trust in his/her followers. The followers may
know that the influencer is paid to post product endorsements; however,
they may also consider that the influencer has his/her reputation and
their followers' trust to keep, and therefore he/she will only endorse a
good product. The followers also feel that they have a good connection
with the influencer. A study from [21] showed that strong and weak ties
have a different impact on purchase intention. Unlike customers, whom
the consumers consider ‘unknown’, followers feel they have a ‘strong tie’
to the influencer. Therefore, their opinions matter and they are trusted.
The results of the study show that not all types of eWOM are effective
at influencing purchase intention. The type of product influences what
type of eWOM is effective. In this study, for luxury products, such as an
expensive smartphone, customer review was not effective. The infor-
mation from customer reviews was less trusted and was therefore
considered less valuable. For luxury products, influencer endorsement is
considered more valuable by potential customers. Besides the factor of
trust, it seems that, for luxury products, ‘opinion leader’status also plays
an important role. An influencer is considered ‘the leader of his/her
pack’, and therefore their opinion matters.
Trust has been studied extensively, particularly its influence on
intention. It has been studied for its role as a direct variable [30]ora
moderator variable [39]. Trust plays an important aspect in the trans-
action, considering the uncertainty and risk involved. In e-commerce,
trust is crucial and one of the most influential factors [30]. Consumers are
unlikely to make an online transaction if they do not trust the seller [36].
Trust can create a positive attitude towards transaction behaviour, which
leads to transaction intention [30].
This study failed to show the moderating effect of trust on customer
reviews and influencer endorsements to purchase intention. This study
also failed to show that the higher the level of trust in the review, the
higher the purchase intention. However, the alternative model showed
that trust was not a moderator but an independent variable, which
influenced purchase intention directly.
This phenomenon can be explained by referring to signalling theory.
“Signals support the identification of an interacting partner as either trust-
worthy or untrustworthy during the relationship”[45]. Trust acted as a
moderator when the partner was assessed during the relationship by
observing the partner's observable signal. In this study, there was no
contact between the influencer and the respondent. Therefore, the rela-
tionship between trust in the influencer's endorsement and purchase
intention was not a moderating effect, but a direct effect. Further study is
needed to see whether the role of trust is independent or mediating.
6. Conclusion
Informativeness plays a key role in online shopping. Electronic word-
of-mouth (eWOM) is considered a good source of information. Customer
reviews and influencer endorsements can be considered eWOM because
they represent customers sharing with other potential shoppers their
experience and their evaluation of a product or service. This study shows
that not all types of customer review are effective in influencing purchase
intention. For luxury products, influencer endorsement, but not customer
review, influences purchase intention. This study, however, failed to
show the moderation effect of trust on the relationship of influencer
endorsement to purchase intention.
6.1. Theoretical implication
The results of this study provide further evidence that eWOM in-
fluences purchase intention, although not all types of eWOM influence
purchase intention. This study showed that influencer endorsements had
a positive impact on purchase intention, but customer reviews failed to
show an influence on purchase intention. Trust as a moderating variable
was also not validated in this study. This study confirms that not all types
of eWOM can influence purchase intention. The influence of different
types of eWOM depends on different types of products.
This study also contributes to the literature on influencer endorse-
ments, which is currently still limited. Further research into influencer
endorsements’effects on purchase intentions should be carried out
extensively, considering the huge sums of money involved in this
activity.
6.2. Managerial implication
Influencer marketing is a hot trend and 41% of marketers claimed
that they will use influencers in their upcoming marketing campaigns.
However, studies that show the effectiveness of influencer endorsement
are still limited. This study provides empirical evidence on how influ-
encer endorsement positively influences purchase intention. This evi-
dence can provide insight to marketers that using influencer
endorsement in their marketing campaigns is effective. This study also
showed that not all types of eWOM are effective. For perceived expensive
products, influencer endorsement is effective, but not customer review.
6.3. Limitations and further research
This study has some limitations. The first limitation was the use of a
fictitious brand. For expensive products, such as smartphones, the brand
influences decision-making. Therefore, the next study should consider
the use of a real, exclusive brand. The second limitation was the use of
fictitious websites and influencers. The next study should consider using
real websites, customer reviews and influencers to give a stronger
impression. The third limitation was the study location. The respondents
were from Greater Jakarta. Broader coverage should be explored for
future studies. Further study to investigate how different generations are
influenced by different types of eWOM could also be interesting. Despite
previous study results on the relationship of trust to purchase intention,
this study failed to show a significant relationship. The alternative model
showed that trust has a direct effect on purchase intention. Further study
to reconfirm the role of trust directly on purchase intentions should be
explored.
Declarations
Author contribution statement
D. D. Tjiptadi: Conceived and designed the experiments; Performed
the experiments; Wrote the paper.
D. Tjahjana: Performed the experiments; Analyzed and interpreted
the data.
S.B. Abdinagoro: Conceived and designed the experiments; Per-
formed the experiments.
D. Gandasari: Performed the experiments; Wrote the paper.
M. Zainal: Performed the experiments.
Funding statement
D.D. Tjiptadi was supported by Binus University and D. Tjahjana was
supported by University Multimedia National.
Declaration of interests statement
The authors declare no conflict of interest.
Additional information
No additional information is available for this paper.
D. Dwidienawati et al. Heliyon 6 (2020) e05543
8
Appendix 1
D. Dwidienawati et al. Heliyon 6 (2020) e05543
9
Appendix 2
D. Dwidienawati et al. Heliyon 6 (2020) e05543
10
References
[1] Statistica, E-commerce Share of Total Retail Sales in United States from 2013 to
2021, Statistica, 2019. https://www.statista.com/statistics/379112/e-commerce-
share-of-retail-sales-in-us/PREMIUM.
[2] Statistica, E-commerce Share of Total Retail Sales in Indonesia from 2015 to 2017,
Statistica, 2019. https://www.statista.com/statistics/970740/indonesia-e-c
ommerce-share-of-retail-sales/PREMIUM.
[3] H. Zhu, Z. Yang, C.X.J. Ou, H. Liu, R.M. Davison, Investigating the impacts of
recommendation agents on impulsive purchase behaviour 2 theoretical
development, in: Australasian Conference on Information Systems, 2015, pp. 1–12.
[4] T. Hennig-Thurau, K.P. Gwinner, G. Walsh, D.D. Gremler, Electronic word-of-mouth
via consumer-opinion platforms: what motivates consumers to articulate themselves
on the internet? J. Interact. Market. 18 (1) (2004) 38–52.
[5] H.G. Pereira, M. Cardoso, P. Dionísio, The determinants of website purchases: the
role of e-customer loyalty and word-of-mouth, Int. J. Electron. Market. Retailing. 8
(2) (2017) 136–156.
[6] K. Ohk, M. Kim, “Who’s leading China’s E-commerce industry? The antecedents and
consequences of E-WOM focusing on one person media, J. Theor. Appl. Inf. Technol.
96 (5) (2018) 1323–1333.
[7] A.S. Yusuf, A. Razak, C. Hussin, A.H. Busalim, Influence of e-WOM engagement on
consumer purchase intention in social commerce, J. Serv. Market. 32 (4) (2018)
493–504.
[8] H. Cheng, S.-W. Huang, “Exploring antecedents and consequence of online group-
buying intention : an extended perspective on theory of planned behavior, Int. J.
Inf. Manag. 33 (1) (2013) 185–198.
[9] K.C. Ling, L.T. Chai, T.H. Piew, “The effects of shopping orientations, online trust
and prior online purchase experience toward customers’online purchase intention,
Int. Bus. Res. 3 (3) (2010) 63–76.
[10] C.M.K. Cheung, M.K.O. Lee, D.R. Thadani, The impact of positive electronic word-
of-mouth on consumer online purchasing decision, in: World Summit on Knowledge
Society, 2009, pp. 501–510.
[11] G. Satell, 3 reasons to kill influencer marketing, Harv. Bus. Rev. September (2014).
[12] S.V. Jin, A. Muqaddam, E. Ryu, Instafamous and social media influencer marketing,
Market. Intell. Plann. 37 (5) (2019) 567–579.
[13] J.F. Bokunewicz, J. Shulman, Influencer identification in Twitter networks of
destination marketing organizations, J. Hosp. Tour. Technol. 8 (2) (2017).
[14] D.D. Bergh, B.L. Connelly, D.J. Ketchen, L.M. Shannon, Signalling theory and
equilibrium in strategic management research: an assessment and a research
agenda, J. Manag. Stud. 51 (8) (2014) 1334–1360.
[15] S. Hilal, J.A. Densley, D.S. Jones, A signalling theory of law enforcement hiring,
Polic. Soc. 27 (5) (2017) 508–524.
[16] B.L. Connelly, S.T. Certo, R.D. Ireland, C.R. Reutzel, Signaling theory: a review and
assessment, J. Manag. 37 (1) (2011) 39–67.
[17] M. Rahman, M.
A. Rodríguez-Serrano, M. Lambkin, Brand management efficiency
and firm value: an integrated resource based and signalling theory perspective, Ind.
Market. Manag. 72 (February 2019) (2018) 112–126.
[18] C.J. Clare, G. Wright, P. Sandiford, A.P. Caceres, Why Should I Believe This?
Deciphering the Qualities of a Credible Online Customer Review, ”Leeds Beckett
University, 2012.
[19] E. Bakshy, J.M. Hofman, D.J. Watts, W.A. Mason, “Everyone’saninfluencer:
quantifying influence on twitter, in ACM Web Search and Data Mining 11 (February)
(2011) 65–74.
[20] S. Dixit, A. Jyoti Badgaiyan, A. Khare, “An integrated model for predicting
consumer’s intention to write online reviews, J. Retailing Consum. Serv. 46
(October) (2019) 112–120.
[21] J.J. Wang, L.Y. Wang, M.M. Wang, Understanding the effects of eWOM social ties
on purchase intentions: a moderated mediation investigation, Electron. Commer.
Res. Appl. 28 (2018) 54–62.
[22] J.R. Siqueira, N.G. Pe~
na, E. ter Horst, G. Molina, “Spreading the word: how
customer experience in a traditional retail setting influences consumer traditional
and electronic word-of-mouth intention, Electron. Commer. Res. Appl. 37 (40)
(2019) 100870.
[23] Q. Xu, “Computers in Human Behavior Should I trust him ? The effects of reviewer
profile characteristics on eWOM credibility, Comput. Hum. Behav. 33 (2014)
136–144.
[24] Q. Yan, S. Wu, L. Wang, P. Wu, H. Chen, G. Wei, “Electronic Commerce Research
and Applications E-WOM from e-commerce websites and social media : which will
consumers adopt ? Electron. Commer. Res. Appl. 17 (2016) 62–73.
[25] P.Y. Setiawan, E.A. Troena, N. Armanu, The effect of e-WOM on destination image ,
satisfaction and loyalty, Int. J. Bus. Manag. Invent. 3 (1) (2014) 22–29.
[26] C.W. Yoo, G.L. Sanders, J. Moon, Exploring the effect of e-WOM participation on e-
Loyalty in e-commerce, Decis. Support Syst. 55 (3) (2013) 669–678.
[27] I. Alfina, J. Ero, A.N. Hidayanto, M.R. Shihab, The impact of cognitive trust and E-
wom on purchase intention in C2C E-commerce site, J. Comput. Sci. 10 (12) (2014)
2518–2524.
[28] S. Castellano, V. Dutot, Investigating the influence of E-word- of-mouth on E-
reputation, Int. Stud. Manag. Organ. 47 (1) (2017) 42–60.
[29] S.M. Mudambi, D. Schuff, “What makes A helpful online Review ? MIS Q. 34 (1)
(2010) 185–200.
[30] B. Lu, W. Fan, Social presence, trust, and social commerce purchase intention: an
emperical research, Comput. Hum. Behav. 56 (2016) 225–237.
[31] H. Jeong, D. Koo, Combined effects of valence and attributes of e-WOM on
consumer judgment for message and product, Internet Res. 25 (1) (2015) 2–29.
[32] M. Ranga, M. Dayanand, D. Sharma, M. Dayanand, Influencer marketing - a
marketing tool in the age of social media, Abhinav Int. Mon. Refereed J. Res.
Manag. Technol. 3 (8) (2014) 16–21.
[33] Talkwalker, What Is Influencer Marketing ?”Talkwalker.Com, 2018 [Online].
Available: https://www.talkwalker.com/blog/what-is-influencer-marketing-
and-why-your-brand-needs-it.
[34] P. Lagr, O. Capp, B. Cautis, S. Maniu, Algorithms for online influencer marketing,
ACM Trans. Knowl. Discov. Data 13 (2019) 1.
[35] M. De Veirman, V. Cauberghe, L. Hudders, Marketing through Instagram
influencers: impact of number of followers and product divergence on brand
attitude, Int. J. Advert. 36 (2017) 798–828.
[36] E.B. Ponte, E. Carvajal-Trujillo, T. Escobar-Rodríguez, Influence of trust and
perceived value on the intention to purchase travel online: integrating the effects of
assurance on trust antecedents, Tourism Manag. 47 (2015) 286–302.
[37] Y. Zhu, H. Chen, Service fairness and customer satisfaction in internet banking,
Internet Res. 22 (4) (2012) 482–498.
[38] R. Ladhari, N. Souiden, B. Dufour, The role of emotions in utilitarian service
settings: the effects of emotionalsatisfaction on product perception and behavioral
intentions, J. Retailing Consum. Serv. 34 (2017) 10–18.
[39] H.H. Chang, K.H. Wong, “Information &Management Adoption of e-procurement
and participation of e-marketplace on firm performance : trust as a moderator, Inf.
Manag. 47 (5–6) (2010) 262–270.
[40] K.L. Sidalia, H.S.A. Spillera, The impact of online reviews on the choice of holiday
accommodations, in: Information and Communication Technologies in Tourism,
2009, pp. 87–98, 2006.
[41] S. Amed, S. Mukherjee, P. Das, B. Datta, Triggers of positive eWOM: exploration
with analytics, Market. Intell. Plann. 37 (4) (2019) 433–450.
[42] C. Lien, M. Wen, L. Huang, K. Wu, “The effects of brand image, price, trust and
value on purchase intentions,”asia pacific manag, Rev. 20 (4) (2015) 210–218.
[43] J. Nadler, R. Weston, E. Voyles, “Stuck in the Middle : the use and interpretation of
mid-points in items on questionnaires, J. Gen. Psychol. (April) (2015) 10–12.
[44] J. Lian, T. Lin, Effects of consumer characteristics on their acceptance of online
shopping: comparisons among different product types, Comput. Hum. Behav. 24
(2008) 48–65.
[45] T. Ebert, Trust as the Key to Loyalty in Business to Consumer Exchanges, First,
Gabler Edition Wissenschaft, Frankfurt, 2009.
D. Dwidienawati et al. Heliyon 6 (2020) e05543
11
