Sponsorship transparency on purchase intention

(1)

S

PONSORSHIP

T

RANSPARENCY ON

P

URCHASE

I

NTENTION

ALI, HURIA ABDULKADIR

KALANE, MADITLHARE LEBOHANG ELSY

School of Business, Society &

Engineering

Course: Master Thesis in Business

Administration

Course code: FOA403

15 credits

Supervisor: Konstantin Lampou

(2)

ABSTRACT

Date:

June 9, 2020

Institution: School of Business, Society and Engineering, Mälardalen University

Authors:

Huria Abdulkadir Ali & Maditlhare Lebohang Elsy Kalane

(91/06/16)

(95/09/18)

Title:

Sponsorship Transparency on Purchase Intention

Tutor:

Konstantin Lampou

Keywords:

Purchase intention, Sponsorship transparency, Influencers, Source

credibility, Disclosure

Research

question:

To what extent does influencers’ sponsorship transparency impact consumer

purchase intention in the beauty community?

Purpose:

The purpose of this paper is to examine how consumers’ online purchase

intention is impacted by sponsorship transparency and the use of disclosure.

Method:

An online survey was collected from 135 people from 41 different countries

around the world. The collected data was used to test the proposed four

hypotheses and was analyzed with a linear regression test.

Conclusion:

The findings suggest that the ethical practice of sponsorship transparency

negatively impacts purchase intention; and posits that an explicit sponsorship

disclosure has no positive impact on purchase intention. However, the

characteristic trait of the influencer, i.e. transparency and authenticity and

their source credibility has a significant impact on prompting consumers

toward purchase intention.

(3)

Acknowledgments

We would like to express our utmost appreciation to our supervisor Dr. Konstantin Lampou for his guidance during the process of conducting this study. He steered us in the right direction with his helpful and critical comments. We also want to thank our seminar group members, for their valuable and constructive suggestions.

We would like to express our gratitude to all the respondents that partook in the survey. This study would not have been possible without their insightful contributions. We also thank those who helped us evaluate our survey questions before sending it to potential respondents. Their advice and opinions have been highly valuable and appreciated. Last but not least, we would like to thank our co-assessor, Dr. Peter Thilenius, for his valuable feedback and suggestions.

(4)

List of Figures

Figure 1: Conceptual Model

12 List of Tables

Table 1: Operationalization table for sponsorship transparency survey

17 Table 2: Reliability statistics

22 Table 3: Spearman correlation coefficient of purchase intention on sponsorship transparency,

disclosure hashtag, social media influencer, and source credibility.

23 Table 4: Simple regression analysis of purchase intention – sponsorship transparency, social

(6)

1 1 Introduction

The beauty trade is a multi-billion industry that has been on a steady trajectory of growth in the last decades. According to a survey done by Tire Industry Research in 2018, the global beauty and personal care market value is estimated to reach a value of 716.3 billion U.S. dollars by the year 2025 (Shahbandeh, 2019). In 2019, Euromonitor released a survey that states that the United States of America is currently leading the market valued at 89.7 billion U.S. dollars, followed by China at 60.5 billion U.S. dollars (Shahbandeh, 2020). The numbers should not come as a surprise, as the beauty industry has been subject to many changes in recent years such as high consumption of Korean cosmetic products (Kbeauty) attributed to the “Korean Wave” (Lee, Sung, Phau, & Lim, 2019), green and sustainable alternatives for conscious consumers (Singhal & Malik, 2018) and at the apex, influencers to help grow businesses, especially small and medium-sized enterprises (SMEs) (Konstantopoulou, Rizomyliotis, Konstantoulaki, & Badahdah, 2019).

1.1 Background

Prior literature on cosmetics and beauty products regarding purchase intention has focused on celebrity endorsements (Yan, 2018); other literature has centered around factors influencing purchase intention of cosmetic products (Chin & Harizan, 2017). However, due to the rise of influencers, brands have been using influencer marketing to drive purchase intention (Sexsmith & Angel, 2012) as influencers, or influencer marketing, has transformed marketing and many different industries. However, the transformation is more tangible in the beauty industry. It has become increasingly common for marketers to opt for social media for their business promotional strategies because of its wider reach and considerable low cost (Sexsmith & Angel, 2012). Additionally, using influencers on social media is a one-way marketers attempt to attract online consumers. According to a beauty study report by Pixability, a software company that specializes in insights for YouTube and other social media platforms for influencer marketing, today marketers in the beauty industry employ digital-first strategies (Pixability, 2018); and they utilize different digital platforms such as YouTube, Facebook, and Instagram for awareness, advertising, and impacting purchase intention. Furthermore, the utilization of well-known and public figure influencers will more likely boost the reputation of a business and attract consumers to purchase its products (Djafarova & Rushworth, 2017). However, using already existing influencers or brand evangelism communities where consumers identify with the influencer or brand, there has been a positive impact on purchase intention (Hsu, 2019). The beauty industry is a big industry and it comprises cosmetics such as skin care products, make-up, and hair products.

(7)

2 1.2 Problem Formulation

Previous literature shows that influencers have a significant impact on purchase intention (Lee & Watkins, 2016). The claim is also affirmed by research from three decades ago, as Atkin and Block (1983) stated that the use of celebrity endorsers, the influencers of the time, in advertisements on media has seen increased purchase intention compared to advertisements that do not use ‘influencers’. Continuous studies on influencers and purchase intention have led to the expansion of the literature that agrees that influencers positively impact purchase intention; and they develop the literature by looking at the extent which source credibility, products, and target audience magnify the influence of the influencer to significantly impact purchase intention (Rahmi, Sekarasih, & Sjabadhyni, 2017; Ananda & Wandebori, 2016). However, there have also been other literature that disputes the argument that influencers have a positive impact on purchase intention (Lim, Mohd Radzol, Cheah, & Wong, 2017).

Furthermore, there has been development in the literature concerning transparency of influencers, or lack of it, regarding paid endorsements (Kim & Kim, 2020). Although the literature on transparency has progressed from organization transparency to studying the role transparency plays in relationship marketing with customers and the relationship between business-to-consumer (B2C) e-commerce and online purchase intention (McCorkindale & DiStaso, 2014; Alshurideh, Al Kurdi, Vij, Obiedat, & Naser, 2016; Zhou, Wang, Xu, Liu,& Guet, 2018). The literature on the transparency of influencers, i.e. sponsorship transparency, is still in its infancy stage. The concept, sponsorship transparency which in simple terms means influencers are open and honest about being paid to talk about a product or service (Evans, Phau, Lim, & Jun, 2017). According to Evans et al., (2017) persuasion knowledge theory means the consumers use prior knowledge to defend oneself from the persuasive intent of the communication message, and hence the consumers are able to recognize content or a post as an advertisement. Prior findings of the theory of persuasion knowledge suggest when consumers are made aware of the persuasive intent of the communication, they are inclined to resist the persuasion tactic (Ham, Nelson, & Das, 2015). As a result, consumers have a less favorable attitude towards the brand; and this conceives negative attitudes towards sponsorship transparency by marketing practitioners. Moreover, this has led to sponsorship transparency literature being used in native advertising, which can be defined as online advertising that reduces the disruptions that consumers may experience when exposed to in-stream marketing ads(Jung & Heo, 2019). The research body surrounding native advertising and advertisement disclosures, i.e. sponsorship

(8)

3

transparency, found that there was a negative effect on advertisement performance when advertisement disclosures were applied to native online advertisements (Jung & Heo, 2019). Additionally, with native ads being covert by nature, Boerman, van Reijmersdal, and Neijens (2012) found that the presence of advertisement disclosures had an adverse attitude effect towards brands. These findings have contributed to the determination marketing practitioners have with evading the use of sponsorship transparency and advertisement disclosures in influencer marketing and native advertising. Jung and Heo (2019) stated that subtle advertisement disclosures such as provided by and presented by on native online ads have resulted in concerns of misleading and deceiving consumers.

However, due to the proliferation of influencer marketing research studies and increasing interest in sponsorship transparency, more scholars are looking into the impact of influencer sponsorship transparency towards attitudes and behavioral intent (Evans et al., 2017). Further research on sponsorship transparency and how it affects purchase intention is growing (Dhanesh & Duthler, 2019). One of the reasons why sponsorship transparency literature is growing is the regulations the Federal Trade Commission (FTC) has enforced to make practitioners and influencers to be transparent about posts that are compensated by brands (FTC, 2019). Another reason why sponsorship transparency literature is expanding is due to the reluctance of marketers and influencers as they find being transparent about sponsored content leads to low post engagement, negative attitude towards influencers and brands, and unfavourable behavioral intent (Dhanesh & Duthler, 2019). Furthermore, a report by a data analytics company, Klear, reported that sponsored content by influencers is increasing, in 2019, the sponsored posts grew by 48% from 2018 on Instagram (Hutchison, 2019). Additional findings of the report indicated that the most sponsored influencers are micro-influencers, i.e. influencers who have a following between five thousand to thirty thousand potential consumers, and are in demand as they often have a niche market and an engaging audience (Klear, 2019). Furthermore, the report’s findings revealed that 25% of the sponsored content in Instagram stories was related to the beauty industry (Klear, 2019). Among the top three trends they found, the report indicated that 69% of brands will be utilizing influencer’s materials for their promotional strategies because they are affordable and have a highly targeted audience (Klear, 2019). This implies that there will be a shift in marketing where influencers execute promotional strategies for brands, by recommending and endorsing products and services to their audience and being compensated for it. Therefore, the ethical practice of sponsorship transparency of brands and influencers is even more important now as sponsorship transparency will become a norm for sponsored content to protect consumers from being deceived, and brands and influencers from breaking the law as per FTC legislation.

(9)

4

As a result and due to most of the literature focusing on native advertising, persuasion knowledge, and advertising recognition, the authors deem it necessary to explore this concept of sponsorship transparency further in the context of influencer marketing. The researchers are interested in observing whether influencer sponsorship transparency has a significant impact on consumer purchase intention. The researchers will then use the findings and connect them to the beauty community. This study will contribute to the growing literature on influencers, purchase intention, sponsorship transparency, and sponsorship disclosure. In addition to this, it will contribute by using findings to offer practical advice to marketers in the beauty field for better influencer marketing practice as well as sponsorship transparency practice. For the purpose of this paper, the authors will focus on the beauty community on social media, this includes influencers that are referred to as beauty influencers and those who have content on beauty, i.e. make-up, skin-care, and hair products.

1.3 Research Purpose

The aim of this study is to examine the impact of sponsorship transparency and the use of disclosure on purchase intention. Moreover, the aim is to test whether the transparency of the influencer, as a character trait, source credibility and disclosure type has a positive effect on purchase intention. To assist in the investigation of this study, a quantitative method was utilized. Additionally, quantitative data was collected via an online survey.

1.4 Research Question

Therefore, in this research study the authors investigate the impact of influencers’ sponsorship transparency towards purchase intention and answering the following research question:

RQ: To what extent does influencers’ sponsorship transparency impact consumer purchase intention in the beauty community?

(10)

5 2 Literature Review

This section defines and discusses the concepts of social media influencer and transparency, purchase intention, sponsorship transparency, and disclosure based on previous studies. This literature review was conducted in order to develop and elaborate on these concepts. Sponsorship transparency and disclosure are discussed in detail as the concepts are relatively new within the influencer body of literature. Additionally, the literature review was conducted to assist the researchers to formulate hypotheses for this research study in an effort to answer the aforementioned research question.

2.1 Social Media Influencers and Transparency

In today’s online business world, using influencers has become one of the most frequently employed strategies by marketers (Nugraha, Kusumawardani, & Octavine, 2018). According to De Veirman, Cauberghe and Hudders (2017), influencer marketing is when a company uses an influencer to market its brand, product, or service rather than reaching customers through advertisements by the company itself. Utilizing influencers who are public figures or well-known will potentially boost a business reputation and attract more customers to purchase a product or service (Djafarova & Rushworth, 2017). Influencers are not necessarily traditional celebrities but can also be non-traditional celebrities who are famous online such as beauty bloggers, vloggers, and Instagram celebrities (Hwang & Zhang, 2018).

With the increase of social media usage, social media influencers are emerging more than ever before. Social media platforms such as Facebook, YouTube, and Instagram have made it easier for influencers to aptly publicize and promote products to their online followers. According to Tuten and Solomon (2015), social media influencers are individuals that others view as valid and reliable sources before deciding to purchase a product or service. These influencers consistently engage with their online followers by frequently updating them with the latest information (Liu, Jin, Briones, & Kuch, 2012). The influencers usually have large numbers of followers on their social media accounts and this makes their opinions highly valuable and likable by consumers (De Veirman et al., 2017). However, Lindh and Lisichkova (2017) cite a report by Markerly indicating that a “vast number of influencers does not necessarily mean real engagement or more purchases, as the more followers the influencer acquires, the less the engagement outreach and brand exposure as the target segment becomes too broad”. Additional literature indicates that it is more appealing

(11)

6

for followers to be able to directly interact and engage with the influencers on social media and follow their daily lives (Djafarova & Trofimenko, 2018).

According to Park (2013), social media influencers can highly impact the consumers’ purchase intention and their attitudes. It is also argued by Zeljko, Jakovic and Strugar (2018) that non-traditional influencers in the beauty industry like beauty vloggers are more powerful than traditional celebrities such as actors and singers since they are easily accessible and engage with their followers almost instantly and frequently. Furthermore, consumers consider these influencers to be more credible than mainstream celebrities because they can relate to them (Djafarova & Rushworth, 2017). Because of this, influencers are able to persuade their followers to buy the same products they advertise and promote on their social media platforms (Hwang & Zhang, 2018). However, parasocial relationships between the influencers and followers are essential since followers must have a strong interest in the life of the influencer in order to be impacted by their advertisements and promotions (Djafarova & Trofimenko, 2018).

Lohtia, Donthu and Guillory (2013) stated that it is crucial for influencers to have their audiences’ trust to effectively attract more followers, promote brands, and impact purchase intention. Trust is when the followers are feeling confident about not being lied to and this is earned through regularly providing information that is credible (Hayes, Singer, & Ceppos, 2007). One of the valuable attributes that an influencer should possess to encourage and increase trust amongst the followers is transparency (Enli, 2016). This requirement for openness and trust has always been a necessity for actors in all areas of public life including the beauty industry hence, transparency is neither a new idea nor exclusive to influencers (Heald, 2006). The aim of transparency in influencers is to enhance trust and credibility by being honest and making information accessible to the audience which eventually impacts purchase intention (Allen, 2008).

Further research on social media influencers has shown that transparency plays a strategic role in the characterization of the influencer’s authenticity (Audrezet, de Kerviler, & Guidry Moulard, 2018). An additional study investigating self-branding on social media states “authenticity and realness have a significant presence” on followers of social media influencers (Liu & Suh, 2017). Furthermore, in a different research context Meier (2009) stated that transparency is the “golden rule” for bloggers, i.e. before they were called influencers, to establish credibility and authenticity with their audience.

(12)

7 2.2 Purchase Intention

Opinions formed by influencers on social media platforms such as Instagram and YouTube are considered to have significant importance and impact on consumers and followers (Pornpitakpan, 2004). When online consumers trust the influencers they welcome promotions and recommendations proposed, which could reshape their buying decisions (Konstantopoulouet al., 2019). The concept of purchase intention can sometimes be easily confused with consumer attitudes, where attitude is a summary of evaluations and intentions (Spears & Singh, 2012). However, according to Lu, Chang and Chang(2014), purchase intention is ‘’the willingness of the consumer to buy a product at a particular time or situation’’.

Previous studies have found that perceived source credibility is one of the main factors that can increase the persuasive power of a message, which impacts purchase intention (Ohanian, 1990). According to Ohanian (1990), the term source credibility is usually used to imply the positive attributes of a communicator that have an impact on the receiver’s acceptance of a message. Among the multiple dimensions of perceived source credibility that have been projected in several scholarly works, trustworthiness and expertise are two aspects that most commonly appear (Hautz, Füller, Hutter, & Thürridl, 2014). When a source is perceived as trustworthy and knowledgeable on a certain product or service, the message delivered will become more effective in impacting and changing the audience’s attitude and intention (Pornpitakpan, 2004).

In general, followers perceive non-traditional celebrities including beauty vloggers and bloggers to be more credible than traditional celebrities such as actors and singers (Djafarova & Rushworth, 2017), because they are considered to be more honest and transparent in delivering the information (Allen, 2008). Furthermore, Andsager, Bemker, Choi and Torwel (2006) argue that consumers discern beauty vloggers to be more credible since they are more likely to identify themselves with them as fellow ordinary consumers rather than celebrities. On the other hand, those vloggers must be qualified enough to provide valid and accurate information, as Ohanian (1990) explains that one should be an expert, qualified, experienced, and knowledgeable. Moreover, research has shown that being trustworthy and expert on a product or service has an impact on the consumers’ purchase intention (Smith, Menon, & Sivakumar, 2005). Recent studies emphasize that trust is critical in purchase intention (Chang & Fang, 2013). Lindh et al., (2020) support the statement by stating that it is important for consumers to develop trust before purchasing.

(13)

8 2.3 Sponsorship Transparency and Disclosure

Sponsorship transparency and disclosure concepts are usually linked to persuasion knowledge and native advertising literature (Boerman et al., 2012; Wojdynski & Evans, 2014). It is only in recent years that these concepts have shifted towards literature concerning influencers as the Advertising Standards Authority of U.K. (ASA) and Federal Trade Commission (FTC) have found it imperative to create guidelines to protect global consumers (ASA, 2018; FTC, 2019). The ASA and FTC guidelines are created to make advertisement recognizable (ASA, 2018) on platforms such as Instagram and YouTube; protect consumers against advertising deception and ensure that both the influencer and the endorsement partner comply with the law through sponsorship transparency (FTC, 2019). To get a better understanding of what sponsorship transparency is, Wojdoynski, Evans and Hoy (2018) describe it as the extent to which a sponsored message is clear to the consumer that it is paid for and the sponsor is visible for consumers to see. Sponsorship transparency research is currently contributing to the development of influencer marketing (De Jans, Van de Sompel, De Veirman, & Hudders, 2020), particularly when it concerns endorsements by influencers. De Jans et al., (2020) found that when influencers endorse a product or brand, it resulted in consumers having a greater brand liking among adolescents due to source admiration, meaning that the more the followers admired the influencer, the greater they liked the brand they were endorsing. Additional developments include the observation by Lu et al., (2014) between sponsorship recommendations and purchase intention. They found that there is a positive impact towards purchase intention when consumers are exposed to sponsored content provided that they have a positive attitude toward the influencer (Lu et al., 2014).

Before sponsorship transparency was legalized by the FTC, influencers would be paid and in return would post content showcasing and reviewing products in an attempt to convince the consumers to like or purchase the products (De Jans et al., 2020). Due to this, it became unclear whether content regarding products, especially recommended products, were endorsements or personal opinions of the influencer (De Jans et al., 2020). It is this lack of clarity that necessitated sponsorship transparency as consumers felt misled. Previous literature confirms that lack of sponsorship transparency led to consumers feeling deceived and as a result had decreased trust in the sponsors and its contents (Evans, Wojdynski, & Hoy, 2018). Woodroof, Howie, Syrdal and VanMeter (2020) further add that the consumers’ distrust applied to influencers as well. This means that the lack of sponsorship transparency meant consumers were trusting influencers less. Prior research indicates that consumers’ distrust is based on the realization that they were deceived into thinking the post was not an advertisement, therefore were unable to activate their

(14)

9

advertising literacy, meaning they were unable to critically evaluate the persuasive attempt of the content (Hudders, Pauw, Cauberghe, Panic, Zarouali, & Rozendaal, 2017).

De Jans et al., (2020) mention that previous research has indicated that influencer marketing was effective in boosting purchase intention. This is an important notation because influencers fear that sponsorship transparency will make them less effective. However, in their research, Evans et al., (2018) found that sponsorship transparency lessened negative results towards advertising, the brand, and purchase intention. Thus, refuting the claim that influencers will be ineffective if they are sponsorship transparent. To abide by FTC and ASA guidelines, influencers have to use disclosures to make consumers aware that the content is sponsored (Jung & Heo, 2019).

2.3.1 Disclosure

The purpose of disclosure is to clearly identify the communication as advertising (Wojdynski & Evans, 2016). As per FTC regulations, influencers are required to use disclosures in posts that incorporate sponsoring (De Jans et al., 2020). Instagram has made it easy for influencers to meet these regulations as they are able to tag a business, which automatically expresses the disclosure as “Paid partnership with [brandname]” in the post (De Jans et al., 2020). However, influencers frequently disclose sponsored posts with hashtags such as #sponsored and #ad; and if they choose to be ambiguous they disclose sponsored posts as ‘thank you to [brandname]” (De Jans et al., 2020; Woodroof et al., 2020). The #sponsored and #ad disclosures indicate that the post is sponsored by a brand and that it is an advertisement that the influencer was paid to endorse among their followers. The disclosures mentioned here make it easier for consumers to differentiate between a normal influencer post and a sponsored influencer post (Wojdynski & Evans, 2016). The #[brand]partner is used less by influencers, and it is a hashtag that meets all the FTC guidelines as it discloses the sponsor, indicates that the influencer was paid and that the post is an advertisement (FTC, 2019; Woodroof et al., 2020).

Disclosure research has indicated that revealing a post is sponsored activates both conceptual and attitudinal persuasion knowledge, i.e. advertising literacy (Boerman et al., 2012). To understand the research claim, persuasion knowledge needs to be defined; Evans et al., (2017) define it as the knowledge that assists consumers to recognize, evaluate and recall the persuasion attempts in order to use appropriate coping tactics when exposed to an advertisement. This means when consumers see a disclosure on a post, they are able to activate persuasion knowledge that allows them to critically evaluate the advertisement and decide whether to make

(15)

10

a purchase because they need/want what is being advertised (conceptual) or they trust both the influencer and what is being advertised (attitudinal).

Previous disclosure literature has indicated that the longer consumers see the disclosure the better it will be for brand recall (Boerman et al., 2012). Additional literature suggested that where the disclosure is located is critical, as it makes it easier for consumers to recognize that it is an advertisement (Boerman, van Reijmersdal, and Neijens, 2014). It is important to note these studies because Evans et al., (2017) mention that disclosure can influence affective, cognitive, and behavioral outcomes in different capacities. This counters the argument that sponsorship transparency and use of disclosures will make influencers less effective due to less engagement and thus affect purchase intention (De Jans et al., 2020). Woodroof et al., (2020) agree as the results of the study indicated that when persuasive knowledge was activated due to disclosures, there was a significant indirect effect on purchase intention.

(16)

11 3 Hypotheses Development and Conceptual Model

The following chapter consists of the hypotheses development and conceptual model using the literature review as a foundation. Additionally, new literature will be utilized to ensure that the hypotheses are concrete with the support of scholarly articles.

3.1 Hypotheses Development

Influencers are entities that assist potential customers to make a buying decision by influencing their opinion through networking (posts) on social media platforms (More & Lingam, 2019). Marketers and brands have recognized the ability of influencers and the direct influence they have on their followers by making recommendations that result in a product search, product purchase, and use of the product (Uzunoğlu & Misci Kip, 2014). Due to companies recognizing this business opportunity and utilizing influencers as intermediaries to drive product messages for purchase intention, this has led to consumers being exposed to sponsored content as a form of advertisement without their knowledge. In order to protect consumers from advertisements that seem like a regular post recommending a product; the FTC (2019) has made it a law for influencers to indicate if the post is an advertisement by disclosing that it is sponsored and that influencers were compensated by using hashtags. Lu et al., (2014) claim that there is a positive relationship between the attitude towards a sponsored post and purchase intention. De Jans et al., (2020) from Computers in Human Behaviour journal further claim that the sponsored influencers “have an important influence on their followers’ consumption decision”. Based on these arguments, the study suggests that influencers using sponsorship transparency will positively impact purchase intention. Therefore, the following hypothesis is proposed.

H1: The use of sponsorship transparency has a positive impact on purchase intention in the beauty community.

In the last three decades, source credibility has been deemed as one of the key factors affecting purchase intention (Ohanian, 1990). Previous literature on purchase intention, celebrity endorsements, and influencers has supported this claim (Hui, 2017; Rahmi et al., 2017). This indicates that an influencer only has influencing power when they are deemed credible by their followers. Sokolova and Kefi (2020) further support these claims by stating that the credibility of the influencer has a positive effect on purchase intention. Additional studies on sponsorship transparency have indicated that influencers being transparent positively impacts the influencers’

(17)

12

source credibility (Evans et al., 2018). Additionally, influencers who have branded themselves as transparent and authentic have a significant influence, i.e. influencers who characterize themselves as transparent outside commercial benefit (Liu & Suh, 2017). The arguments stated above infer that influencers that are transparent and authentic by characterization and have source credibility positively influence purchase intention. The proposed hypotheses:

H2.1: Social media influencer transparency positively influences purchase intention in the beauty community.

H2.2 Source credibility positively influences purchase intention in the beauty community.

Though much of the literature on sponsorship disclosure types focuses on persuasion knowledge and advertisement recognition in native advertising, there has been much interest in the effect of sponsorship disclosure types on purchase intention. In examining the effect of impartial sponsorship transparency, Stubb and Colliander (2019) found impartial or explicit sponsorship disclosures had no effect on purchase intention. In addition to this Evans et al., (2017) in their investigation at how different sponsorship disclosure languages (types) affects behavioural intent, they found that an explicit hashtag such as #PaidAd and #Sponsored had no effect on purchase intention than that of a vague sponsorship disclosure such as #SP. An additional previous study investigated the impact of a clear sponsorship disclosure type, such as #ad and #sponsored, in comparison to an ambiguous sponsorship disclosure type, and they found that consumers are unable to distinguish between the two (Woodroof et al., 2020). It can be said that due to consumers being unable to distinguish between the sponsorship disclosure types, this may be the root reason for the lack of any significant impact on purchase intention. With prior studies, the aim is to add more to the literature that follows Evans et al., (2017) in sponsorship disclosure types, by looking into how different sponsorship disclosure type, #[brand]partners can result in a positive impact purchase intention. Thus, we hypothesize the following,

H3: The use of the #[brand]partner in a post has a positive impact on purchase intention in the beauty community

.

3.2 Conceptual Model

With the previous literature as groundwork, the researchers developed four hypotheses to evaluate the relationship between the concepts.

(18)

13

Figure 1: Conceptual Model

Based on prior studies, the conceptual model above (see figure 1) and hypothesis one (H1) assumes that sponsorship transparency has a positive influence on purchase intention in the beauty community on social media. This hypothesis is based on previous literature on sponsorship transparency (Woodroof et al., 2020; Dhanesh & Duthler, 2019). Previous studies’ results have suggested that transparency of influencers and source credibility positively impacts purchase intention (Sokolova & Kefi, 2020). Therefore, the hypotheses two (both H2.1 and H2.2) undertake the assumption stated above in the beauty community. Based on previous literature, results show that there is a negative relationship between a clear sponsorship disclosure type and purchase intention; however, hypothesis three (H3) assumes that #[brand]partner disclosure is an explicit disclosure and it does influence purchase intention in the beauty community (Evans et al., 2017; Woodroof et al., 2020; Stubb & Colliander, 2019).

(19)

14 4 Methodology

The purpose of this study is to examine the relationships between influencers, purchase intention, and sponsorship transparency. Therefore, this chapter will explore the suitable research approach, research design, and data collection method required to be able to test the hypotheses above and accomplish the purpose of the study. Additionally, this chapter will discuss sampling, operationalization of the constructs, as well as reliability and validity.

4.1 Research Approach

According to O'Reilly (2009), an inductive approach is where the researchers’ preconceptions lead to emerging theories, however, with the deductive approach, the hypotheses are derived from existing theories and the validity of these existing theories are empirically tested through hypotheses. Furthermore, a deeper definition of a deductive research approach is an approach that is used to describe the relationship between theory and research (Bryman & Bell, 2015). Based on the purpose and research question of this study, a deductive approach was appropriate as the authors have tested hypotheses derived from existing theories (Dahlberg & McCaig, 2010), meaning in this case; theories on different aspects such as purchase intention, sponsorship transparency, and disclosure. Therefore, this study has been conducted with references to the hypotheses that have emerged from the existing theories discussed in the literature review.

Bryman and Bell (2015) argue that the main focus of quantitative research is not to describe how things are, rather explain why things are in a certain way. They also mention that quantitative research bases its study on the collection of quantifiable data and analysis. Thus, quantification was used as the main tool for analyzing the collected data. Additionally, the research approach for this study requires data that could represent a larger population and make the result generalizable. Therefore, the researchers have used the deductive approach under quantitative methods.

4.2 Research Design

According to Bryman and Bell (2015), research design provides a framework and acts as a supporting foundation for gathering and analyzing data, thus, a correct choice and implementation

(20)

15

of it; is of great importance in collecting quantitative data. Furthermore, Bryman and Bell (2015) stated that the choice of research design reflects on the priority that is being given to a range of dimensions of the research process such as causality linkage between different variables. This helped in generating the required data to show and describe how sponsorship transparency and the use of disclosures can impact consumer purchase intention, hence this paper emphasizes descriptive purpose. Consequently, a cross-sectional design was applied in this research study to gather the relevant quantitative data needed for the study.

The cross-sectional design, which is also referred to as a social survey, entails the collection of data on more than one case at a single point in time in order to collect quantifiable data (Bryman & Bell, 2015). Thus, the aim of this research design is to measure the chosen variables to determine the correlation between them (Eggert & Helm, 2003). For this reason, the cross-sectional design is suitable for the study to find out the correlation between sponsorship transparency of an influencer and the purchase intention of a consumer. One of the advantages of using this design is that it allows researchers to gather a great deal of information quite quickly in the form of surveys from a large pool of participants (Eggert & Helm, 2003). It is, therefore, suitable for the research considering the time at hand to collect the required data.

4.3 Data collection method

Online survey research was used in this study to collect data from participants. According to Bryman and Bell (2015), survey research is a data collecting method that predominantly uses questionnaires and falls within the quantitative cross-sectional design category. Therefore, it was appropriate to use this method to gather the required data since the cross-sectional design was chosen as the research design. When conducting the survey research, self-completion questionnaires were implemented to collect the data from respondents. This method is helpful in gathering large amounts of data from many respondents in a short period of time. Additionally, in the construction of the questionnaires, the emphasis was given on closed questions. To ascertain the percentage of the sample who agreed to participate in the survey, the response rate is established. The research consisted of 135 participants out of which 116 were confirmed to have completed the survey. The response rate was calculated and found to be 87.9%. The survey was open from 1 May for pilot testing and after modifications, the survey took place from 2 May 2020 to 22 May 2020.

(21)

16

Closed questions are useful in ensuring what is intended to be measured as it directly gives respondents answer options to choose from (Bryman & Bell, 2015). The questionnaires were distributed by sending and posting them on different online platforms such as Facebook, Instagram, e-mail, and WhatsApp. Online platforms were chosen due to their ability to have a wide scope and reaching a great number of individuals (Ioanas & Stoica, 2014). The survey consisted of 12 close-ended and multiple-choice questions. The questions were in the form of rating questions and a 7-point Likert scale was used. According to Joshi, Kale, Chandel and Pal (2015), Likert scale is a set of items or statements offered for a real or hypothetical situation under research where respondents are asked to show their level of agreement (from strongly disagree to strongly agree) with the given items on a metric scale. It is also considered that the 7-point Likert scale is reliable since it provides more range of options which in turn increases the probability of meeting the objective reality of people (Joshi et al., 2015). One screening question was also included to check if the respondents follow at least one influencer on social media, this was to make sure that the information gathered is correct and relevant. Additionally, definitions of the terms such as ‘influencers’ and ‘sponsorship transparency’ was provided for the respondents to avoid any misunderstanding of the concepts.

4.4 Sampling

When conducting any research, it is rarely practical to study and receive answers from a whole population, therefore sampling becomes essential; even more so in quantitative research (Bryman & Bell, 2015). According to Bryman and Bell (2015), a sample is the segment of the population that is selected for investigation. Since it is important to target the right sample to receive answers from, for this study, the data set contained individuals that are 18 years of age and above, who also follow at least one influencer on a social media platform. To confirm this, control questions were asked in the survey questions to make sure the targeted people were reached, and accurate data was gathered.

Two different sampling methods are mentioned by Bryman and Bell (2015); probability sampling and non-probability sampling. They further explained that probability sampling is a sample that is selected randomly whereas non-probability sampling is not selected using a random selection method; which means some units in the population are more likely to be selected than others. In this research, the non-probability convenience sampling was applied since the authors have

(22)

17

already non-randomly chosen the individuals they want to target based on the aforementioned criterion by using the screening questions. The convenience sampling is a sampling method where the chosen respondents are the easiest to reach, however, they are still part of the sample criterion (Bryman & Bell, 2015).

4.5 Pilot-testing

In order to assure the reliability and validity of the structure of the questionnaire, it is important to pretest the questions when conducting a survey research (Bryman & Bell, 2015). Therefore, a pilot test was carried out to test the design of the survey, since it is essential that all surveys are pre-tested before the actual survey is conducted (Adam, Khan, Raeside, & White, 2007). Apart from the structure of the questionnaire and the design of the survey, the reliability and validity of the data from the respondents primarily depend on the rigor of the pilot-testing conducted beforehand (Saunders, Lewis, & Thornhill, 2009). One of the benefits of pilot-testing is to make sure the respondents can understand and respond to the questions without any problem and ambiguity (Fink, 2017).

For the above-mentioned reasons, the survey questions were distributed to different respondents and a pilot test was conducted which included qualified people and active followers of influencers on social media. This was done to ensure if the wording and sequence of the questions including the general layout of the survey were clear, understandable, and easy to follow. Additionally, to know if the survey was time-consuming or not to answer for the respondents. Based on the feedback received, it showed that the layout of the questionnaire was clear to follow and understand the questions. A slight modification had to be made on one of the questions to make it more clear for the participants. In the beginning the question was ‘’name your favourite influencer’’, this had to be changed to ‘’name any influencer you follow’’. This modification was made based on the feedback after the pilot testing, because the former question was limiting the participants since many people do not have one favourite influencer but follow many.

4.6 Operationalization

Operationalization entails formulating measures of constructs or concepts that the researcher is interested in (Bryman & Bell, 2015). This is to ensure that the constructs are trustworthy. To operationalize our constructs, all the measures used in this research study were adapted from the literature. The influencer transparency has been operationalized using (Woodroof et al., 2020; Dhanesh & Duthler, 2019) with the purpose of testing whether an influencer who is perceived as

(23)

18

transparent can impact purchase intention, and sponsorship transparency using (Wojdynski, Evans, & Hoy, 2018) with the purpose of investigating whether sponsorship transparency has a positive effect on purchase intention. To operationalize the disclosure hashtag (#[brand]partner) the authors utilized (Wojdynski et al., 2018) to test if the #[brand]partner has a positive impact on purchase intention. The operationalization of source credibility construct follows (Sokolova & Kefi, 2020), the word ‘blogger’ was adapted and changed to an influencer for the purpose of this study. The purpose of this construct is to test whether influencers that are deemed credible by their followers can positively impact purchase intention. For the purchase intention construct, it was operationalized using (Sokolova & Kefi, 2020; Dhanesh & Duthler, 2019). Following the consistency of previous literature, all the items are reflective, with the exception of the disclosure hashtag construct being semi-formative, and assessed with a 7-point Likert scale. Table 1 summarizes the questionnaire.

Table 1: Operationalization table for sponsorship transparency survey

Construct Item description (Likert Scale, 1= strongly disagree, 7= strongly agree)

Source(s)

Social media influencer

transparency

The influencer is honest about the post he/she was paid to make.

I can rely on an influencer to post only products he/she believes in.

I can rely on an influencer to post only the products he/she personally uses.

I am aware that this influencer is paid to endorse certain brands and products. I can tell when an influencer is paid to endorse a certain brand/ product or not.

(Woodroof et al.,, 2020; Dhanesh & Duthler, 2019)

(24)

19

Sponsorship transparency

This influencer clearly conveyed the product or brand that was being promoted.

The influencer made the name of the advertiser very obvious.

The influencer said it was an advertisement.

The influencer said it was sponsored.

(Wojdynski et al., 2018).

Disclosure Hashtag (#[Brand]partner)

It was clear who sponsored the post The post was labelled as advertising. This post was trying to fool consumers into thinking it wasn’t advertising.

The hashtag showed that the post was an endorsement

(Wojdynski et al., 2018).

(25)

20

Purchase intention I would like to try the brands endorsed by this influencer.

I would buy other products of this brand because of this influencer.

I would actively seek out the product/service shown by this influencer in order to purchase it.

I would purchase the products promoted by the influencer in the future.

I would encourage people close to me to buy the products/brands promoted by this influencer.

(Sokolova & Kefi, 2020; Dhanesh & Duthler, 2019)

Source credibility I find this influencer expert in his/her domain.

I find this influencer efficient in his/her job. I find this influencer trustworthy.

I think this influencer cares about his/her followers.

(Sokolova & Kefi, 2020)

4.7 Reliability and Validity

According to Bryman and Bell (2015) reliability considers the question of whether the results of the research study are repeatable and stable. In an effort to measure whether the questions are repeatable or not, reliability evaluates the consistency of the survey questions (Rasinski, 2008). To indicate that the study has internal consistency, it is measured by a statistic named alpha (Rasinski, 2008). When a study is reliable it means that it can be repeatable, stable, and unbiased (Gavin, 2008).

(26)

21

Reliability measures the consistency of the study, and Lewis-Beck, Bryman and Liao (2004) state that internal reliability presents an estimated evaluation of the consistency of the responses towards the items that are measured. The internal reliability, also referred to as internal consistency, is evaluated using Cronbach alpha (Bryman & Bell, 2015). The Cronbach alpha test is used to reduce random measurement error so that the error is not highly correlated with the true score among the responses of the items (Salkind, 2010; Lewis-Beck et al., 2004). Bryman and Bell (2015) add that Cronbach alpha measures whether the multiple-item concept is coherent and whether the items are related. To depict this, Bryman and Bell (2015) mention that the coefficient varies between zero (0) and one (1). The coefficient of 0 means that there is no correlation, indicating that there is no internal consistency, and the coefficient of 1 means perfect correlation, indicating a complete internal consistency (Bryman & Bell, 2015). Though there is a debate among scholars regarding a suitable coefficient, Salkind (2010) mentions that an acceptable coefficient is between 0.70 to 0.79. However, an acceptable coefficient value is between 0.80 to 0.89 and is often employed as a rule of thumb; any value higher than 0.89 means high reliability and internal consistency (Bryman & Bell, 2015). Additionally, Taber (2017) states that the Cronbach alpha value below 0.55 is not satisfactory, and an alpha value above 0.55 is described as moderately sufficient. For the purpose of this study, the authors use the argument of Salkind (2010), where the acceptable coefficient value is 0.70. With the semi-formative construct, the internal reliability is lower than the acceptable coefficient value of 0.70.

Validity takes into account the integrity of the conclusions that have been generated from the research study (Bryman & Bell, 2015). Researchers usually measure internal or external validity. Internal validity measures the extent to which a research design permits researchers to posit that there is a causal relationship between variables or there is an absence of a causal relationship (Cramer & Howitt, 2014). In contrast, external validity measures whether the results can be generalized to the reality in the actual world (Rasinski, 2008). Though there are other ways to measure the validity of a study, it is important to know that it is one of the important criteria in a research paper.

To measure whether the study is consistent, accurate, and trustworthy, it is crucial to verify the reliability and validity (Bryman & Bell, 2015). The validity of the survey was verified by using validated statements from prior studies that were utilized to create a survey via Google Forms. Spearman correlation was also used to measure the correlation between the constructs as a way to ensure convergent validity (Bryman & Bell, 2015). In order to define correlations between

(27)

22

variables, Hennig and Cooper (2011) claim that a study must have at least 100 participants. Therefore, it was made sure that this research consisted of at least 100 respondents. The correlation value is between zero (0) and one (1) which will be either positive or negative (Bryman & Bell, 2015). Zero or the value closer to zero indicates a weak relationship and one or a value close to one indicates a strong relationship, and the positive and negative aspect indicates the direction of the relationship (Bryman & Bell, 2015). The goal of the researchers is to ensure reliability to get stable results and trustworthiness. The data will be analyzed using the SPSS statistics system, and reliability will be tested using Cronbach’s alpha.

4.8 Data Analysis

The data analysis was conducted using analytical software. The authors utilized the renowned Statistical Package for the Social Sciences (SPSS) for quantitative analysis.

Bryman and Bell (2015) state that it is important to employ only the number of usable data, this means using data that participants fully participated in. It is therefore imperative to determine the responsive rate percentage of the sample thus far. Bryman and Bell (2015) indicate that the problem with a poor response rate is sampling bias which affects the quality of the survey. Therefore, it is good to use a sample with a high response rate as it shows that the survey was completed. To analyse the current data, known and accepted software SPSS by IBM was used (Bryman & Bell, 2015). Data was collected and imported as a Microsoft Excel File to SPSS. The relationship between sponsorship transparency, disclosure hashtag, social media influencer transparency, purchase intention, and source credibility was evaluated. For each of the constructs which are discussed in this study, the relevant question items were tested. The constructs and the question items tested are as follows: Sponsorship transparency-4, Disclosure hashtag-4, Social media influencer-5, Purchase intention-4 and source credibility-4. These question items were transformed and measured and reported into different constructs (Table 2).

To test the strength of the relationship between the independent and dependent variables, a linear regression test was conducted (Allen, 2017). For a simple linear regression, only one independent predictor is used to test the dependent variable (Allen, 2017), for this study a simple linear regression test was used to test H1, H2.1, H2.2 and H3. In order to keep the risk minimum, the acceptable significance level value is p < 0.05, therefore if there is a p-value less than 0.05, the model is significant (Hinton, Brownlow, & McMurray, 2014).

(28)

23 4.9 Ethical Considerations

When conducting research and studies that involve people, there is a high possibility that potential ethical issues can occur, therefore it is important for researchers to take this into consideration (Bryman & Bell, 2015). For this reason, it was made sure that the anonymity of the respondents was maintained for confidentiality purposes. Moreover, personal anonymity in research is significant in gaining reliable information (Fox, Murray, & Warm, 2003). Respondents were also informed that completing the survey was strictly voluntary and under no obligation to participate if they choose not to. Another important aspect that was taken into consideration was the well-being and protecting respondents from any harm that may be caused since the research method is internet-based. Hence, the responses were kept anonymous and were not used for any commercial purposes. The data gathered from the survey was only used for the purpose of this study and will not be accessible to a third party for other objectives.

5 Results and Analysis

This chapter will present the obtained results collected through an online survey. The results will be discussed with a more in-depth discussion taking place in the next chapter.

5.1 Data Analysis

The survey had responses from 135 participants from 41 countries. South Africa was the country with the highest participants with 32.6%, followed by Sweden and the United States of America at 12.9% and 5.3%, respectively. A detailed description of the data sample of the participants’ home country is specified in Appendix 1. The majority of the participants were between 19-29 years old (76.5%), and the gender distribution of females was 65.2%, while the males were at 34.1%. The age and gender distribution table can be found in Appendix 2. The survey indicated that Instagram and YouTube are the two platforms respondents follow influencers the most with 56.1% and 22.7%, respectively. Refer to Appendix 3 to see the distribution table. The survey indicated that 37 participants follow influencers who are part of the beauty community, and 27.4% are influencers who are part of the beauty community on social media, i.e. post beauty content and 72.6% do not post beauty content (see Appendix 4).

The findings from the Cronbach alpha test indicate sponsorship transparency has a value of 0.748, social media influencers transparency has a value of 0.714, purchase intention has a value of 0.896, and source credibility has a value of 0.815 (See Table 2). This indicates a high and

(29)

24

acceptable reliability. Disclosure hashtag results show the value of 0.309. This indicates the reliability is not satisfactory which suggests the respondents’ response varied to the items in these constructs, additionally this could be because the construct is semi-formative.

Table 2: Reliability statistics

Reliability Statistics

Cronbach's Alpha N of Items

Sponsorship transparency 0,748 4

Disclosure hashtag 0,309 4

Social media influencer transparency 0,714 5

Purchase intention 0,896 4

Source credibility 0,815 4

5.1.1 Correlation

This research study employs the Spearman’s correlation coefficient analysis to determine the relationship and direction of the relationship between the constructs illustrated in the conceptual model. The correlation coefficient r of purchase intention and sponsorship transparency is 0.057 (see table 3). This indicates that there is a positive relationship between purchase intention and sponsorship transparency. A correlation test was conducted to assess the relationship between disclosure hashtag and purchase intention, resulting in a correlation coefficient r of 0.232*. This reveals that the two variables correlate, indicating a positive relationship between disclosure hashtag and purchase intention. The third correlation between social media influencer transparency and purchase intention shows there is a correlation coefficient r of 0.303**. This implies that the variables relate to each other and that there is a positive relationship between the variables. Lastly, source credibility and purchase intention were tested, producing a correlation coefficient r of 0.457**. This signifies a positive relationship between the variables. All three correlations resulted in a p-value <0.05, which shows that the results are statistically significant and valid. However, purchase intention and sponsorship transparency were not significant.

(30)

25

Table 3: Spearman correlation coefficient of Purchase intention on sponsorship transparency, disclosure hashtag, social media influencer and source credibility.

Correlations

Spearman’s rho Correlation Coefficient Sig. (2-tailed)

Sponsorship transparency - Purchase intention

0.057 .541

Disclosure hashtag - Purchase intention

0,232* .012

Social media influencer transparency - Purchase intention

0,303** .001

Source credibility - Purchase intention 0,457** .000

**. Correlation is significant at the 0.01 level (2-tailed). *. Correlation is significant at the 0.05 level (2-tailed).

5.1.2 Regression

To test the four hypotheses of the study, a linear regression analysis was undertaken. The results indicate that the first hypothesis (H1) and the last hypothesis (H3) are not supported; and the results support the proposed hypotheses, H2.1 and H2.2. Based on the results in table 4, sponsorship transparency and purchase intention have a t-value of 0.832 with an insignificant value of .407, which is above 0.05. This indicates that sponsorship transparency does not have a positive effect on purchase intention. Therefore, hypothesis 1 is not supported. The simple linear regression test for H2.1 for social media influencer transparency and purchase intention has a t-value of 3.534 with a significant t-value of .001, which is below the acceptable p-t-value of 0.05. This indicates that social media influencer transparency does positively impact purchase intention. The findings in table 4, show that source credibility and purchase intention have a t-value of 6.246 and a significant value of .000, which signifies that H2.2 is supported as the p-value is below the 0.05 value. This asserts that source credibility does positively influence purchase intention. The regression test indicated that the disclosure hashtag “#[brand]partner” does not have a positive

(31)

26

impact on purchase intention with a t-value of 1.772 and a p-value of 0.079. Therefore, hypothesis 3 is not supported (see table 4).

Table 4: Simple regression analysis of purchase intention – sponsorship transparency, social media transparency, source credibility and disclosure hashtag

Regression

Hypothesis t Sig.

Sponsorship transparency – Purchase Intention

H1 0.832 .407

Social media influencer transparency – Purchase Intention

H2.1 3.534 .001

Source credibility – Purchase Intention H2.2 6.246 .000

(32)

27 6 Discussion

The following section will discuss the findings from the survey. The discussion will be divided into the following concepts: sponsorship transparency, social media influencer transparency, source credibility, and disclosure hashtag. The concept of purchase intention will be integrated into the other three concepts as it is a common element for all.

6.1 Sponsorship transparency

The results indicate that sponsorship transparency does not have a positive impact on purchase intention in the beauty community. The findings are not in line with the researchers who agree that sponsorship transparency does have a positive effect on purchase intention (Lu et al., 2014; Dhanesh & Duthler, 2019). Though the researchers expected positive results for the proposed hypothesis, an explanation of this could be the result of a negative attitude towards the idea of sponsorship; as consumers are afraid of the interference of sponsors, such as having an unfavourable effect on the authenticity of the influencer (Meenaghan, 2001). Studies that show sponsorship transparency has a positive influence towards purchase intention mention that the pre-existing relationship between the influencer and the consumer (follower) plays a role; as Dhanesh and Duthler (2019) mention that due to the established trust, satisfaction and commitment between the influencer and the followers, an awareness of the paid endorsement does not negatively affect the consumers (followers) behavioural intent. Therefore, the results imply that though there is an existing relationship between influencers and followers, the followers’ negative perception of sponsorship is superseding the established trust, satisfaction, and commitment. Other studies mention that sponsored influencer content generates source admiration from followers (De Jans et al., 2020), meaning that the followers admire the influencer when the influencer uses sponsorship transparency. This could explain why the hypothesis is not supported, implying that the followers focus on the influencer and the perception of making money, rather than the endorsed product or service and the influencers’ intent to persuade them to buy.

Prior results that support the proposed hypothesis as mentioned in the literature review state that consumers’ attitude towards the influencer plays a role in leading the consumers’ decision to purchase intention (De Jans et al., 2020). A study by Evans et al., (2018) supports and concurs with the results of De Jans et al., (2020); as Evans et al., (2018) mention that the consumers’ perception upon ad recognition as a result of sponsorship transparency incites a positive attitude

(33)

28

towards behavioural intent such as purchase intention. This is because the inclusion of sponsorship transparency mitigates the negative attitude consumers have towards the advertisement. This implies that the proposed hypothesis is rejected because the consumers might have had an unfavourable attitude and perception towards the influencers they follow such as assuming that the sponsored content is solely for compensation, rather than a reflection of the influencer and therefore, they cannot trust the recommended product or service. Additionally, the proposed hypothesis could be rejected because the influencer used a disclosure that consumers respond negatively to practice sponsorship transparency; as Stubb and Colliander (2019) mention that impartial sponsorship disclosure has no effect on purchase intention. Further discussion on sponsorship disclosure is under 6.3.

Though the ethical practice of sponsorship transparency is for the protection of consumers; it is clear that consumer perception and consumer attitude towards the influencer and how they practice sponsorship transparency play a role and has a negative effect on purchase intention. This key finding offers influencers, advertising practitioners and managers insight to strike a balance that ensures the initial consumer perception and attitude that consumers have towards influencers and purchase intention when they recommend a product without compensation is still there. Particularly, when they endorse a product and practice sponsorship transparency. This insight may lead to significant mitigation of negative effect sponsorship transparency has on purchase intention (Evans et al., 2018)

6.2 Influencer transparency and source credibility

The results indicated that both social media influencer transparency and source credibility have a positive effect on purchase intention in the beauty community. This was in line with studies that found that influencer’s transparency was an important driver of product evaluation and the intention to purchase the product; and how the credibility of the influencer who is found to be trustworthy can positively influence purchase intention (Woodroof et al., 2020; Dhanesh & Duthler, 2019; Sokolova & Kefi, 2020). This is also in line with the results of using celebrities for beauty products in an effort to impact purchase intention (Yan, 2018). The results showed that source credibility was the strongest contributor to purchase intention, and social media influencer transparency was less strong, but still a contributor to purchase intention.

In relation to social media influencer transparency, the results indicate that consumers are cognizant that the influencers are paid to endorse products and are honest when they post about