The YouTube Marketing Communication Effect on Cognitive, Affective and Behavioural Attitudes among Generation Z Consumers

Abstract

YouTube (YT) is the largest online video digital channel with more than 2 billion users, and over a billion hours of YT videos are viewed every day, particularly among young consumers. YT has become a massive marketing communication platform, which serves as a medium to target the lucrative Generation Z cohort (first born in the late 1990s), and influence this generation’s infamously unpredictable purchase decision process. The main aim of this paper was to consider the effect of YouTube marketing communication (YMC) on the traditional and non-traditional attitudinal associations of response hierarchy models. A multi-stage sample technique was used and 3750 high school and college learners (aged 13–18 years old) were surveyed via self-administered questionnaires in South Africa. Structural equation modelling was utilised to consider the hypothesised attitudinal associations. The research determined that YMC had a positive influence on the hypothesised attitudinal associations, and young consumers who used YT for fewer years, logged on more frequently, spent shorter time periods on the platform, viewed higher numbers of commercials, aged 13–14 years old, and from the White population group exhibited the most positive attitudinal responses. Hence, organisations should review their strategies in order to develop more sustainable YMC owing to the heterogeneity evident among young African consumers.

1. Introduction

The tremendous explosion of online video digital channels (OVDCs) have provided Generation Z with uninterrupted entertainment and, consequently, these digital information and communication technology (ICT) conduits have largely replaced traditional television across the globe. OVDC marketing communication (MC) expenditure is forecasted to increase twofold (to $37 billion) over the next half a decade, and will reach an estimated 4.5 billion global users. Over a billion of hours of YouTube (YT) videos are uploaded and viewed on a daily basis, especially by young consumers [1,2,3,4,5]. Younger cohorts, namely Generation Y (Millennials) and Generation Z are spending longer time periods watching OVDCs via mobile devices [5]. Hence, over 65% of retailers plan to increase their OVDC MC budgets, especially when targeting Generation Z (digital natives), since YT is the favourite and fastest growing OVDC among these young consumers [2,6]. The Generation Z cohort spends in excess of $142 billion and influences more than $600 billion of household purchase decisions on an annual basis [7,8]. Therefore, it is vital that originations implement more sustainable YouTube marketing communication (YMC) to target and positively influence young consumer attitudes and the purchase decision process at all levels, especially since the Generation Z cohort is infamously known to be difficult to maintain the attention of and reach with MC [9,10]. Therefore, Generation Z’s attitudes towards YMC are of significant interest to organisations and their brands, since a new body of knowledge (pertaining to this cohort) will assist organisations in increasing their MC effectiveness, as well as provide an inclination of future behavioural predispositions.

Researchers and marketing practitioners have postulated many theories to evaluate the effect of MC. Some maintain that MC is only successful effective if it results in sales [11], while others believe that consumers go through several attitudinal phases prior to purchase, which may include several of the following phases: attention, awareness, knowledge, comprehension, desire, liking, preference, evaluation, retention, purchase intentions, acceptance, action, satisfaction, and others [12]. This investigation is in accord with the second theory, as consumers generally do not react immediately, but take time to develop favourable predispositions towards brands prior to a behaviour response (such as purchase). The hierarchy-of-effects model postulates that consumers go through three attitudinal stages, viz. cognitive, affective, and conative (behavioural) in response to MC. The hierarchy-of-effects model includes six hierarchical phases (two for each attitude stage), namely awareness (cognitive), knowledge (cognitive), liking (affective), preference (affective), purchase intention (behavioural), and purchase (behavioural) [13]. A number of other MC models were developed to ascertain the attitudinal stages that consumers experience prior to purchase, but were developed via traditional media MC [12,14,15,16,17]. Hence, there is a deficiency of theoretical evidence in terms of the influence of OVDC (such as YT) MC on the aforementioned attitudinal associations.

Subsequently, many recent empirical studies considered various aspects of YT usage, MC, and consumer attitudes [18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53,54,55]. A number of research gaps were evident from these YT studies, as well as in other digital ICT channel inquiries. Several researchers assert that there was insufficient YMC information and research in order to use of this OVDC as an effective and sustainable MC medium [21,22,32]. A majority of the above-mentioned inquiries only use usage and demographic factors to describe the research, and do not examine the independent variables further. Hence, a number of inquiries mandate that additional research should be conducted on YT usage and/or demographic variables [20,25,30,56]. Westenberg [24], Kujur and Singh [40], Ahmad et al. [57], and Mukerjee and Shaikh [58] indicate that increased sample sizes should be utilised in social media research. Araújo et al. [26] and Stoilova et al. [59] assert that few studies quantitatively assess MC regarding OVDC among young consumers (e.g., Generation Z). Viertola [4] and Westenberg [24] examined Generation Z’s attitudes and behaviour in terms of YMC, but both of these inquiries used a qualitative approach, which included sample sizes of 20 and 7, respectively. Several investigations also proposed that different cohorts should be considered in terms of future social media research [46,47,54,60].

A number of studies predominantly investigated Generation Y, which in many instances comprised of student samples [21,22,23,25,27,28,30,31,32,33,34,38,39,41,43,44,46,47,54,55]. Several studies also recommended that more diverse or different segments of YT users should be examined, especially since university or college student samples diminished the possibility of generalizing the results due to their unique situational context and life phases [22,29,35,40,46,47,54,57,61]. Additionally, a majority of the aforementioned studies examined various YMC elements in developed countries, especially in the United States (US) and Western Europe. Consequently, a number of researchers maintain that there is a lack of social media and/or YT investigation in developing countries and/or in different cultures [9,20,26,46,47,54,56,58,60,62,63,64]. Furthermore, a number of developing countries are still experiencing rapid social media growth, whereas many developed countries are beginning to reach saturation. The superior ICT infrastructure and faster Internet speeds in developed countries may result in divergent usage and socio-cultural factors in comparison to developing countries (especially in Africa) [1,54,65,66].

Hence, the primary objectives of this research are to assess the effect of YMC on the several attitudinal associations of response hierarchy models among the Generation Z cohort in South Africa (SA), which is based on the attitude-to-advertising (A_ad) theoretical framework. The study also seeks to determine if usage and demographic variables have an influence on the traditional response hierarchy-of-effects model’s attitudinal associations. Furthermore, the usage and demographic variables’ investigation will reveal if heterogeneity exists in young cohorts, as directed by Zhang et al. [56] and Bolton et al. [67].

2. Review of Literature and Development of Hypotheses

2.1. YouTube

YT is the largest OVDC with in excess of 2 billion users. Over a million hours of YT videos are uploaded and viewed on a daily basis, with seven out of ten users using mobile devices to view YT content [1,68]. This OVDC comprise of a number of user-generated content (UGC) videos such as how-to-do, music, educational, and organisational content. YT was formerly created as an entertainment and information video conduit, but has subsequently grown into a huge MC platform, which includes channels, celebrity endorsers, influencers, YouTubers, promotions, advertising, product placement, and testimonials. YT makes content available at no cost, since this video viewing channel derives most of its income from MC, and has experienced steady growth with the rapid proliferation of mobile devices [1,5]. YMC is a quick, targeted, controlled, and cost-effective means to sustainably promote organisations and their brands, especially if the video content is topical. However, younger consumers frequently find YMC to be intrusive, especially in developing countries where bandwidth is expensive and slow, since it takes more time to stream for commercials. YT provides a number of different MC types, which comprise of standard and display, discovery (in-search), overlay, sponsored cards, and video bumper commercials. The video commercials are the commonly used by organisations and their brands, which appear at the beginning and during YT videos. Many of these commercials are skippable after five seconds, but others are not skippable, which can be 30 s in length or longer [1,5,43,68,69]. Organisations can also use product placements, influencers, celebrity endorsers, YouTubers, and testimonials to promote their brands, which are effective in reaching young consumers, and assist in generating traffic and providing exposure for the organisations’ offerings [24,38,43].

Organisations can also upload their television commercials and other public relations material via brand channels, which have changed the manner in which young consumers view commercials, since these can be viewed at their leisure. YT brand channels have become cost-effective conduits, since commercials can be uploaded at no cost, and could potentially be viewed by millions of young consumers if they are sufficiently entertaining to go viral. Furthermore, YT provides comprehensive analytics/metrics, whereby organisations can obtain detailed information regarding who is viewing their channel content, as well as other MC content in terms of demographic, geographic, video usage, and other consumer information [1,5,69,70,71]. However, many organisations have large MC budgets, which are based on YT metrics, but lack important information about young consumers’ sequential attitudinal responses that culminates in purchase decisions [71]. Several studies (as outlined in the introduction) have considered the effectiveness of YMC via quantitative (surveys and analysis of content) and qualitative (interviews and focus groups) approaches. Hence, this investigation utilised the former approach, viz. a survey, to quantitatively assess Generation Z’s attitudinal associations in terms of YMC in an African developing country, which will provide valuable information on the attitudinal phases that occur prior to the ultimate purchase.

2.2. Generation Z Cohort

Generation Z was first born a couple of years prior to the turn of the century and onwards—this cohort is the first true 21st century generation. This cohort has not known life without personal computers, mobile devices, social media, the Internet and the omnipresent “search” (e.g., via Google). The Generation Z cohort carry their technology with them and are continually connected, frequently engaging in a range of activities at the same time, and have an incessant need for instant gratification and technology dependence [72,73]. This cohort seeks to be continually connected, and Internet access is considered to be a basic need, whereas increased speed is viewed as an effective digital conduit, which is used to enhance efficiency and communication, particularly through mobile apps. Generation Z wants content to be instantaneously accessible in audio (e.g., podcasts) and video, as well as segmented text to aid skimming, which is prevalently for online site reading (e.g., blogs and social media). This generation is generally independent, individualistic, and ambitious, as well as exude diversity in terms of culture, population group, gender, and sexual orientation [24,66,72,73]. Hence, organisations should embrace these traits and incorporate this cohort’s uniqueness in developing more sustainable MC, so as to support the diversity values that are prevalent among young consumers.

Two thirds of South African’s are younger than 36 years—Generation Z comprises 41% and Millennials comprise 25%. Generation Z did not experience apartheid, and so are also known as the “Born Frees”. This cohort has experienced rapid political, social, and economic change, which has made this generation more resilient, ambitious, independent, and optimistic. A majority has the highest education level in their households, and is pressured to succeed and make a contribution to their families owing to better education and job prospects, notwithstanding apartheid’s long-lasting negative socio-economic effect [74,75,76]. The oldest of the Generation Z cohort has recently completed high school, and has either entered the job market or is at university. This generation has a significant influence on household purchase decisions. The increased wealth of the “Black Middle Class” (formerly known as “Black Diamonds”) has grown by more than 60% over the last decade (aided by Employment Equity, Affirmative Action, and Broad-Based Black Economic Empowerment initiatives), which has added to the purchasing power of this dynamic cohort [66,74,76,77].

South African Generation Z members are captivated by MC that enable them to co-create and are more favourable disposed to organisations that allow them to make a decision, choose, and/or vote. Young African consumers no longer scrutinise American or European trends on what to purchase, but value and embrace present and future African culture, diversity, influences, intricacies, and insights as differentiation points [66,73,76]. Stoilova et al. [59] affirm that studies are limited in most countries, which consider the influence of YT among the Generation Z cohort. Balakrishnan and Manickavasagam [20] assert that there is little information about demographical and usage influences on attitudes in terms of social media MC. Furthermore, a majority of quantitative inquiries only considered the influence of various YMC elements on Generation Y (Millennials), whereas the several inquiries that did assess young consumers YT attitudes and behaviour, were qualitative in nature [4,24].

Hence, this study seeks to ascertain young consumer attitudes towards YMC, and to examine the impact of YMC on usage and demographic variables on the Generation Z cohort’s attitudes in an African developing country.

2.3. Consumer Attitudes and Hypothesis Development

As previously mentioned, a number of advertising frameworks were developed to evaluate different attitude phases, which consumers go through before the ultimate purchase, but these frameworks were formulated through traditional advertising. As mentioned in prior text, Lavidge and Steiner [13] developed arguably the most famous response hierarchy frameworks, viz. the hierarchy-of-effects model, which proposes different traditional attitudinal responses, namely: first awareness and then knowledge (cognitive attitudinal responses); first liking and then preference (affective attitudinal responses); and first intention-to-purchase and then purchase (behavioural attitudinal responses). The attitude stages of the hierarchy-of-effects model are often equated to the communications effects pyramid model, whereby consumers also go through phases, which ends in purchase; however, it becomes incrementally more challenging to achieve the upper echelon phases [12,78]. A number of organisations use one or more of the consumer attitudinal responses to develop MC objectives.

Several recent studies use Ducoffe’s [79] advertising value attitude model and/or Brackett and Carr’s [80] web advertising attitude model (or an adaption of these models) to consider consumers’ attitudes regarding various online ICT channels, which include informativeness/knowledge (cognitive attitudinal responses), entertainment (affective attitudinal responses), irritation, credibility, and additional attitude elements [21,22,25,33,36,44,45,49,81]. Moreover, several other social media-related studies considered various buyer behavioural responses, as well as purchase intentions and/or purchase (behavioural attitudinal responses) [19,30,49,51,52,54,55,62,64,81,82,83].

However, a majority of the above-mentioned studies were conducted in developed countries (as mentioned in prior text), which have begun to reach social media usage saturation, whereas developing countries (particularly in Africa) are still experiencing steady social media usage growth, which is largely due to the prolific growth of mobile devices [2,5,20,66]. Most of the aforementioned studies sampled Generation Y (students), whereas few have quantitatively considered Generation Zs A_ad in terms of YT. There is also a dearth of quantitative research in terms of YMC, which include all of the attitude elements regarding Lavidge and Steiner’s [13] hierarchy-of-effects model among the Generation Z cohort. Hence, this study will consider the traditional cognitive, affective, and behavioural attitudinal associations among young consumers due to YMC (refer to Figure 1):

• H1. YMC has a favourable effect on traditional attitudinal associations amid the Generation Z cohort.
• H1a. YMC has a favourable effect on the awareness→knowledge association amid the Generation Z cohort.
• H1b. YMC has a favourable effect on the liking→preference association owing amid the Generation Z cohort.
• H1c. YMC has a favourable effect on the intention-to-purchase→purchase association amid the Generation Z cohort.

Additionally, previous research on social media and other online ICT platforms only considered the hierarchy-of-effects model in the original sequence, and were limited to the traditional associations between the attitudinal phases [46,47,54,84,85,86,87,88,89,90,91,92,93,94,95,96,97]. However, the attitudinal phases can be attained in a divergent sequence, and several other response hierarchy models, namely Wolfe et al.’s [98] AAPIS model and Aspinwall’s [99] consumer acceptance theory have also emphasised the importance of preference as a mediator between cognitive and behavioural attitudinal responses [12,78]. The abovementioned theories propose two non-traditional attitudinal responses, namely knowledge→preference, and preference→purchase. Each attitude phase of the response hierarchy frameworks should be attained, but a number of phases can theoretically be accomplished at the same time, and/or in a different order as proposed by the original hierarchy-of-effects framework [12,78,84,88,89,90]. However, this theory has not yet been empirically tested in terms of YT, so the awareness→liking and awareness→intention-to-purchase associations will also be explored. Therefore, this study will consider several non-traditional attitudinal associations, which are outlined in the following hypotheses (refer to Figure 1):

• H2. YMC has a positive effect on non-traditional attitudinal associations among the Generation Z cohort.
• H2a. Awareness has a positive effect on liking owing to YMC among the Generation Z cohort.
• H2b. Awareness has a positive effect on intention-to-purchase owing to YMC among the Generation Z cohort.
• H2c. Knowledge has a positive effect on preference owing to YMC among the Generation Z cohort.
• H2d. Preference has a positive effect on purchase owing to YMC among the Generation Z cohort.

2.3.1. Usage Variables’ Hypotheses

Many quantitative attitudinal YT inquiries only use usage variables as a means of sample description, and did not perform cross-examine of these variables versus the attitudinal scales. Duh and Struwig [66] and Zambodla [100] affirm that there are similarities from a worldwide viewpoint among younger cohorts, but that there are also a number of differences, especially between the South African and global counterparts. Bolton et al. [67] agree that further empirical investigation is required to determine if usage variable heterogeneity exist within the younger cohorts. This study expands on Duffett’s [88,89,90] MC research on social network sites and instant messaging among young consumers (i.e., Generation Z), which considered access, number of years of usage, frequency, and durations of log-on. Hence, several hypotheses will be outlined with regards to the influence of YT usage variables on young consumers’ traditional attitude associations as a result of YMC.

Generation Z display the highest levels of sophistication owing to the pervasive smartphones and other mobile multimedia devices, which are primarily used for viewing videos, posting pictures, instant text messaging, playing online games, entertainment, social connection, seeking information, and generally perceiving the world [10]. Smartphone ownership has steadily escalated in emerging nations, especially among younger consumers. Two thirds of South African unique browsers coming from mobile devices and 70% of YT views are derived from mobile devices. Young consumers are spending longer time periods watching YT videos via the omnipresent mobile devices, but few have considered whether computer access versus mobile device access have an influence on YMC [1,24,68,73]. A number of studies investigated the influence of access on the cognitive, affective and/or behavioural attitudes due to different social media conduits MC among the Generation Y and/or Z cohorts, which resulted in divergent results. Several of these studies ascertained that access via mobile devices and/or mobile apps gave rise to more positive attitudes [84,87,88,89,90,94,95,96,97] and other inquiries found that social media access (i.e., computers versus mobile devices) did not have an influence on attitudes [85,86], but none of the aforementioned investigations considered the influence of access due to YMC. Three recent studies found that mobile devices resulted in more positive cognitive, affective and/or behavioural attitudes among Millennials owing to YMC [46,47,54], but did not consider the Generation Z cohort. For this reason, the study considers the following hypotheses:

• H3. The impact of traditional attitudinal associations differs based on how Generation Z access YMC.
• H3a. The impact of the awareness→knowledge association differs based on how Generation Z access YMC.
• H3b. The impact of the liking→preference association differs based on how Generation Z access YMC.
• H3c. The impact of the intention-to-purchase→purchase association differs based on how Generation Z access YMC.

Several studies only used YT experience (the number of years’ usage) to describe the research sample [18,33,34,83]. A number of social media-related and YT investigations considered the effect of experience on consumer attitudes among Millennials and the Generation Z cohort due to MC, but produced mixed findings. Some studies established that less experienced consumers displayed positive A_ad [47,85,90], while some inquiries yielded the opposite result, in that more experienced consumer displayed more positive A_ad [46,54,84,88,89]. Other investigations determined that there were no significant differences among consumer attitude due to YT and/or social media MC [86,87]. Therefore, the resulting hypotheses are as follows due to these diverse results:

• H4. The impact of traditional attitudinal associations differs according to the experience (years) of the Generation Z cohort owing to YMC.
• H4a. The impact of the awareness→knowledge association differs according to the experience (years) of the Generation Z cohort owing to YMC.
• H4b. The impact of the liking→preference association differs according to the experience (years) of the Generation Z cohort owing to YMC.
• H4c. The impact of intention-to-purchase→purchase association differs according to the experience (years) of the Generation Z cohort owing to YMC.

A number of investigations only used YT frequency of log-on to describe the research sample [18,23,33,34,43,45]. Various social media-related and YT MC investigations found that frequency of log-on gave rise to varied attitudes among Generation Y and Z. Several inquiries ascertained that young consumers who logged-on more frequently displayed positive A_ad [4,19,24,26,27,46,54,88], while other studies resulted in converse findings with lower frequencies of log-on generating more favourable A_ad [46,54,84,88,89]. Some studies found no significant differences among young consumer attitudes owing to social media and/or YT MC [86,87]. Hence, the ensuing hypotheses are as follows owing to the mixed results:

• H5. The impact of traditional attitudinal associations differs according to the frequency of log-on by the Generation Z cohort owing to YMC.
• H5a. The impact of the awareness→knowledge association differs according to the frequency of log-on by the Generation Z cohort owing to YMC.
• H5b. The impact of the liking→preference association differs according to the frequency of log-on by the Generation Z cohort owing to YMC.
• H5c. The impact of the intention-to-purchase→purchase association differs according to the frequency of log-on by the Generation Z cohort owing to YMC.

Various studies only considered the duration of YT log-on in terms of descriptive statistics [18,30,33,34,43,45,83]. Many social media-related and YT investigations ascertained that higher durations led to the most positive Millennial and Generation Z attitudes [4,46,47,84,85,86,87,88,89,90,101]. Yet, some inquiries determined that lower durations of log-on by the younger generations produced more favourable A_ad [21,26]. However, other research yielded no significant differences due to frequency of log-on among young consumer attitudes due social media and/or YT MC [54]. Additionally, a study reported that Millennials who spent between 1–3 h exhibited more positive attitudes towards YMC [20]. Therefore, the resultant hypotheses are as follows:

• H6. The impact of traditional attitudinal associations differs according to the duration of log-on by the Generation Z cohort owing to YMC.
• H6a. The impact of the awareness→knowledge association differs according to the duration of log-on by the Generation Z cohort owing to YMC.
• H6b. The impact of the liking→preference association differs according to the duration of log-on by the Generation Z cohort owing to YMC.
• H6c. The impact of the intention-to-purchase→purchase association differs according to the duration of log-on by the Generation Z cohort owing to YMC.

Few YMC-related studies considered the number of YT commercials influence on young consumers’ attitudes. However, three recent inquiries established that a higher number of YT commercials (viewed by the Generation Y cohort members) resulted in the most favourable cognitive, affective, and/or behavioural attitudes due to YMC [46,47,54]. Yet, none of these studies investigated the Generation Z cohort, so for this reason, the research considers the following hypotheses:

• H7. The impact of traditional attitudinal associations differs according to the number of YT commercials viewed by the Generation Z cohort.
• H7a. The impact of the awareness→knowledge association differs according to the number of YT commercials viewed by the Generation Z cohort.
• H7b. The impact of the liking→preference association differs according to the number of YT commercials viewed by the Generation Z cohort.
• H7c. The impact of the intention-to-purchase→purchase association differs according to the number of YT commercials viewed by the Generation Z cohort.

2.3.2. YouTube Demographic Variables’ Hypotheses

The Generation Z exhibit many analogous characteristics and traits around the world, but developing nations generally have inferior digital ICT infrastructure compared to developed nations, which may result in divergent attitudes owing to poorer social-economic conditions. Additionally, there is also both a large disparity of wealth and in access to online platforms in many developing countries. Hence, demographical variables may have an impact on the cognitive, affective and/or behavioural attitudes due to YMC owing to heterogeneity within the Generation Z cohort in emerging countries [20,66,67,100,102]. Furthermore, many YT studies frequently use demographic usage variables (age, gender, population group, income and education) as a research description tool; however, no cross-examination is performed regarding various attitudinal scales. Hence, this inquiry endeavours to ascertain if several demographic variables effect young consumers’ traditional attitudinal associations owing to YMC. Different investigations only used gender to describe the research sample [4,18,19,23,24,25,30,33,35,44,45,48,50,51,52,55,103]. Males displayed more favourable attitudes regarding social media and/or YT MC in a number of inquiries among Millennials [21,27,43,47], whereas females were found to have more favourable in other studies [26,83,84,85,88,89,90]. Additionally, no significant differences were determined by other research [32,46,49,54,86,87]. However, most of these studies examined the Generation Y cohort and/or social media MC (not YMC), so the ensuing hypotheses are as follows:

• H8. The impact of traditional attitudinal associations differs according to the gender of the Generation Z cohort owing to YMC.
• H8a. The impact of the awareness→knowledge association differs according to the gender of the Generation Z cohort owing to YMC.
• H8b. The impact of the liking→preference association differs according to the gender of the Generation Z cohort owing to YMC.
• H8c. The impact of the intention-to-purchase→purchase association differs according to the gender of the Generation Z cohort owing to YMC.

Bolton et al. [67] and Zambodla [100] posit that younger cohorts are heterogeneous, since older cohort members are likely to exhibit divergent psychographic attributes, lifestyles, attitudes, values, needs, interests, preferences, desires, consumption of media, and shopping tendencies in comparison to their younger counterparts. Sharma [104] stresses that further inquiry should investigate young cohorts A_ad in terms of different MC formats across different age groups. Numerous social media-related and/or YT studies only considered age as a descriptive research sample factor [4,18,19,22,23,24,25,28,33,34,35,43,44,45,48,50,51,52,55,103,105], while other studies produced differing findings among various age groups within the Generation Y and Z cohorts. Some social media and/or YT studies established that younger age groups displayed more favourable A_ad within the Generation Y cohort [26,84], whereas older age groups exhibited more positive A_ad in other research [20,21,27,46,47,49,54]. Several studies found no significant differences between different age groups within the Generation Y cohort [32,85,86,87]. Additionally, three social media MC inquiries found that older age groups within Generation Z generational cohort showed more positive A_ad [88,89,90], but did not consider YMC. Hence, the subsequent hypotheses are considered by this study:

• H9. The impact of traditional attitudinal associations differs according to Generation Z’s age owing to YMC.
• H9a. The impact of the awareness→knowledge association differs according Generation Z’s age owing to YMC.
• H9b. The impact of the liking→preference association differs according Generation Z’s age owing to YMC.
• H9c. The impact of the intention-to-purchase→purchase association differs according Generation Z’s age owing to YMC.

SA is a culturally diverse country (referred to as “Rainbow Nation”), which has 11 official languages and comprise of four main population groups, viz. Black (79.4%), Coloured (9.3%), White (8.7%) and Asian/Indian (2.6%) [75]. However, Black South African consumers still have a lower average income per capita compared to the other population groups due to the injustices of apartheid era that ended a little over 25 years ago. However, as mentioned in prior text, there is a fast growing “Black Middle Class” segment, which has captured the attention of organisations, especially in terms of the younger cohorts, owing to their huge influence on household purchases and growing purchasing power [7,8,66,74,76,102,106]. Several social media MC investigations used race as a descriptive research variable among Millennials [19,23,25,44,50], while other inquiries generated divergent findings. Black and Coloured consumers showed more positive attitudes in terms of social media MC in several studies among the Generation Y and Z cohorts [84,85,86,88,89,90], whereas a studies found that White consumers ethnic group displayed more positive attitudes [26]. Rodriguez [32] found that race (Hispanic/Latinos versus Caucasians) had an influence YT advertising among US and Dominican Republic Millennial respondents. However, none of these studies examined if race had an impact YMC among the Generation Z cohort. Furthermore, a number of researchers suggested that additional social media and/or YT inquiry was needed in other countries, especially owing to cultural (or ethnic) group differences [9,20,26,56,58,60,62,63,64]. Therefore, the following hypotheses are examined by this research:

• H10. The impact of traditional attitudinal associations differs according to Generation Z’s race owing to YMC.
• H10a. The impact of the awareness→knowledge association differs according to Generation Z’s race owing to YMC.
• H10b. The impact of the liking→preference association differs according to Generation Z’s race owing to YMC.
• H10c. The impact of the intention-to-purchase→purchase association differs according to Generation Z’s race owing to YMC.

3. Materials and Methods

3.1. Sampling and Data Collection

A multi-stage sample technique was utilised, which comprised of 13–18 year old college and high school learners who used and observed YMC. Firstly, Western Cape Province was selected in SA. Secondly, systematic sampling was used to select 25 colleges and/or high schools from the Western Cape Education Department’s (WCED) list (ethical clearance was received from the WCED and Cape Peninsula’s University of Technology ethics committee). Thirdly, appointments were made telephonically to conduct the research (the next school on the WCED list was selected if a school declined to grant permission to perform the survey). About 160 learners from each college and high school, who had used and observed YMC, were chosen to partake in the research on a voluntary basis [107].

Twelve fieldworkers (roughly one per two schools/colleges) gathered the data through the self-administered questionnaires, which resulted in a very good response rate, since the fieldworkers waited for the respondents to fill-out and submit the completed questionnaires [108]. However, the study was completely anonymous and confidential since no names or contact particulars were gathered, so it was not possible to follow-up on respondents. Consequently, 50 incomplete questionnaires were discarded. Hence, in this manner 3750 self-administered questionnaires were fully completed by college and high school learners, which adequately represented the Western Cape’s research population in terms of demographic variables (gender, age and race) [109]. The broad range of the research objectives (hypotheses) necessitated the large sample size.

Table 1. Usage and demographic variable frequencies.

Usage Variables		n	%
Access	Mobile device	1430	38.1
	PC	428	11.4
	Mobile device and PC	1892	50.5
Experience (years)	≤1 year	512	13.7
	2 years	890	23.7
	3 years	977	26.1
	4 years	643	17.1
	≥5 years	728	19.4
Frequency of log-on	Daily	2428	64.7
	2–4 times a week	663	17.7
	Once a week	343	9.1
	2–4 times a month	170	4.5
	Once a month	146	3.9
Duration of log-on	≤1 h	1351	36.0
	2 h	962	25.7
	3 h	549	14.6
	4 h	300	8.0
	≥5 h	588	15.7
YT commercial viewership #	None	1333	35.5
	1–5	824	22.0
	6–10	634	16.9
	11–15	397	10.6
	≥16	562	15.0
Demographic variables
Gender	Male	1831	48.8
	Female	1919	51.2
Age	13–14	808	21.5
	15–16	1370	36.5
	17–18	1572	41.9
Race	White	901	24.0
	Black	1060	28.3
	Coloured	1519	40.5
	Indian/Asian	270	7.2

offers a summary of usage and demographic descriptive statistics of the South African Generation Z respondents.

3.2. Measures

A questionnaire, which was self-administered, was suitable for this study, since no detailed written responses were required. All of the questions were close-ended and could be physically disseminated to a large number of respondents without much explanation [107]. The questionnaire was adapted from Duffett’s research on social media MC among the Generation Z cohort [88,89,90], which included two filter questions and three main sections. The filters questions determined if respondents had used YT and were exposed to MC on this OVDC. Section 1 collected data regarding respondents’ YT media usage variables (independent variables) via multiple-choice questions, which included: how YT was accessed, YT experience (years), YT frequency of log-on, duration of log-on, and YT commercial viewership (numbers) (refer to Table 1). Section 2 included six Likert scale constructs, namely awareness (four-item), knowledge (four-item), liking (six-item), preference (six-item), intention-to-purchase (four-item), and purchase (six-item) (refer to

Table 2. Attitudinal stages (factor loadings, AVE, CR, and Cronbach’s α).

Attitude Stages	Factor Loadings	AVE	CR	Cronb.’s α
Awareness
YMC are effective in creating awareness of brands	0.734	0.550	0.829	0.743
YMC alerts me to new company offerings	0.800
I have become aware of new YMC	0.806
YMC get my attention towards certain brands	0.611
Knowledge
Ads on YT are effective in providing information about brands	0.690	0.585	0.849	0.775
YMC are a good source of knowledge	0.790
I use YMC to find new information about products	0.830
YMC provide me with valuable product knowledge	0.730
Liking
YMC has made me like the brands more	0.757	0.546	0.877	0.851
YMC adds to the enjoyment of using YT	0.818
YMC are likeable and pleasant	0.800
YMC are entertaining and fun	0.806
YT has a positive influence on me liking advertised products	0.644
YMC has made me like the products more	0.574
Preference
I look for products that are advertised on YT	0.660	0.536	0.874	0.838
YMC are relevant to me and my interests	0.680
Ads on YT are effective in stimulating my preference in brands	0.790
YMC are effective in gaining my interest in products	0.790
I prefer brands that are promoted on YT	0.760
YMC have a positive effect on my preference for brands	0.707
Intention-to-purchase
I will buy products that are advertised on YT in the near future	0.819	0.541	0.821	0.725
I desire to buy products that are promoted on YT	0.861
YMC increase purchase intent of featured brands	0.626
I would buy products that are advertised on YT if I had the money	0.599
Purchase
I purchase products that are featured on YT	0.634	0.565	0.885	0.848
YMC positively affect my purchase behaviour	0.755
Ads on YT help to make me loyal to the promoted products	0.810
YMC favourably affect my purchase actions	0.840
I purchase products that are promoted on YT	0.780
YMC positively affect my buying actions	0.660

). Section 3 collected data regarding Generation Z’s respondents’ demographic variables (independent variables) through multiple-choice questions, which included gender, age, and race (refer to Table 1).

4. Results and Data Analysis

4.1. Measurement Model

Confirmatory factor analysis (CFA) was used to assess to the various validity and reliability factors of the measurement model via IBM’s SPSS and AMOS. The appropriateness of the principle component factor analysis was assessed via a sampling adequacy test, namely the Kaiser-Meyer-Olkin (KMO) test. The KMO value was 0.865, which is indicative of good reliability of the factors. The Bartlett’s Test of Sphericity was significant at 0.000, which is also indicative of robust factor analysis [110]. The reliability of the attitudinal constructs was evaluated by means of Cronbach’s Alpha (α) and composite reliability (CR) measurement values. Table 2 shows that Cronbach’s α values were between 0.725 and 0.851 and the CR values ranged from 0.821 to 0.885, which all exceeded the universally accepted reliability measure score of 0.70 as suggested by Bagozzi and Yi [111]. Hence, all six attitudinal constructs exhibited acceptable reliability. Next, the attitudinal constructs’ convergent validity was considered by assessing factor loadings and the average variance extracted (AVE). The scores of the factor loadings were 0.574–0.861, and the AVE scores were 0.536–0.585, which is suggestive of convergent validity as the scores were all above 0.5 as suggested Bagozzi and Yi [111].

Discriminatory validity was assessed by comparing the square root of AVE values for each attitude construct against the other construct correlations.

Table 3. Correlations between attitude constructs and square root of AVE.

Awareness	0.742
Knowledge	0.430	0.765
Liking	0.119	0.548	0.739
Preference	0.329	0.163	0.448	0.732
Intention-to-purchase	0.035	0.141	0.048	0.429	0.736
Purchase	0.047	0.110	0.113	0.327	0.244	0.751

shows that all construct AVE square root values were larger than correlation values, which is indicative of discriminant validity [112].

The hypothesised relationships between the attitude constructs were investigated through structural equation modelling (SEM), and multi-group SEM, which was used to consider the influence of the usage and demographic variables on the traditional response hierarchy-of-effects model attitudinal associations. The goodness-of-fit statistics was assessed via seven model-fit measures, and all the indices produced a very good fit, according to Hooper’s [113] minimum acceptable thresholds, in terms of the hypothesised SEM model indices (χ²/df = 1.199, RMSEA = 0.007, NFI = 0.992, TLI = 0.998, CFI = 0.999, GFI = 0.994, and SRMR = 0.014). A configural invariance test obtained adequate goodness-of-fit statistics (χ²/df = 1.199, RMSEA = 0.007, NFI = 0.992, TLI = 0.998, CFI = 0.999, GFI = 0.994, and SRMR = 0.018) via the analysis of the freely estimated model over multi-groups.

A metric invariance test was used to constrain the two models to be equal and the chi-square difference test between the fully constrained and unconstrained models showed them to be invariant (p = 0.932). The data for the research was gathered from the self-reported responses of the Generation Z sample. Therefore, a bias test (common method) was executed to compare the common method factor (CMF) (unconstrained) model to the zero CMF (constrained) model. The chi-square test showed that there was a significant different at p < 0.05; hence, the unconstrained CMF model was used, as there was significant variance that was shared. The Cook’s Distance measure was used to establish if outliers were evident regarding Generation Z’s attitudinal responses, however there were no responses that exhibited abnormal Cook’s Distance. A multi-collinearity test was conducted to assess if the attitude constructs were overly correlated with each other, which would adversely influence the reliable estimates of the regression coefficients [114]. The tolerance was greater than 0.1 (0.545–0.994) and the variation inflation factors were less than 3 (1.006–1.834) for the all of attitude constructs, which showed that the attitude constructs were not overly correlated with each other.

4.2. Hypothesis Testing

The path coefficients for the SEM and multi-group SEM, in considering the hypotheses, are evaluated below.

• H1: Traditional attitudinal associations. Figure 2 exhibits the SEM analysis in terms of significance, standardised path coefficients, and variance for the attitudinal association hypotheses. The standardised path coefficients showed significant positive influences for: awareness→knowledge (β = 0.673, p < 0.001), liking→preference (β = 0.751, p < 0.001), and intention-to-purchase→purchase (β = 0.705, p < 0.001) traditional attitudinal associations. Therefore, H1a, H1b, and H1c were supported (refer to Figure 2). Additionally, 45% of the variance due to knowledge was elucidated by awareness amid Generation Z owing to YMC. Knowledge and liking were found to explain 57% of preference’s variance, and preference and intention-to-purchase explained 50% of a purchase’s variance.
• H2: Non-traditional attitudinal associations. No significant influences were found for the awareness→liking (β = 0.036) and awareness→intention-to-purchase (β = 0.037) standardised path coefficients. However, the standardised path coefficients revealed significant positive influences for: knowledge→preference (β = 0.099, p < 0.05), and preference→purchase (β = 0.065, p < 0.05) non-traditional attitudinal associations. Accordingly, H2a and H2b were not supported, whereas H2c and H2d were supported (refer to Figure 2).
• H3: Access. The standardised path coefficients indicated that the devices used to access YT did not have a significant influence on the traditional cognitive (awareness→knowledge), affective (liking→preference), and behavioural (intention-to-purchase→purchase) attitudinal associations due to YMC. Hence, H3a, H3b, and H3c were not supported.
• H4: Experience (years). The standardised path coefficient indicated that awareness had a favourable effect on knowledge amid young consumers who used YT for ≤1 year (β = 0.694, p < 0.05 and p < 0.001) versus those who used YT for 3 years (β = 0647, p < 0.05) and ≥5 years (β = 0.660, p < 0.001). As a result, H4a was supported, while H4b and H4c were not supported.
• H5: Frequency of log-on. The standardised path coefficient also indicated that awareness had a favourable effect on knowledge for young consumers who logged-on to YT daily (β = 0.687, p < 0.05) versus those who logged-on to YT once a week (β = 0.531, p < 0.05). The standardised path coefficient showed that the liking→preference association was more favourable for young consumers who logged-on 2–4 times a month (β = 0.820, p < 0.05) versus those who logged-on to YT once a week (β = 0.717, p < 0.05). Additionally, the standardised path coefficients revealed that intention-to-purchase had a positive influence on purchase for young consumers who logged-on 2–4 times a month (β = 0.765, p < 0.05) versus those who logged-on 2–4 times a week (β = 0.678, p < 0.05). Consequently, H5a, H5b, and H5c were supported.
• H6: Duration of log-on. The standardised path coefficient revealed that the awareness→knowledge association was more positive amid young consumers who spent ≤1 h (β = 0.710, p < 0.05) versus those who spent 3 (β = 0618, p < 0.05) and 4 (β = 0.593, p < 0.05) h logged-on to YT. So, H6a was supported, whereas H6b and H6c were not supported.
• H7: Number of YT commercials viewed. The standardised path coefficient indicated the liking→preference association was more positive for young consumers who viewed 1–5 (β = 0.766, p < 0.05), 6–10 (β = 0.763, p < 0.05), and ≥16 (β = 0.775, p < 0.05) versus those who viewed no (β = 0.689, p < 0.05) YT commercials. The standardised path coefficients also revealed that intention-to-purchase had positive influence on purchase for young consumers who viewed 11–15 (β = 0.733, p < 0.05) versus those who viewed no (β = 0.730, p < 0.05) YT commercials. Accordingly, H7a was not supported, while H7b and H7c were supported.

Refer to

Table 4. Significance and standardized beta coefficients of YouTube usage and demographic variables on the traditional cognitive, affective, and behavioural attitudinal associations.

Independent Variables		Awareness→ Knowledge		Liking→ Preference		Intention-to-Purchase→Purchase
Usage Variables		β	Sig	β	Sig	β	Sig
Access	Mobile device (1)	0.655	-	0.740	-	0.694	-
	PC (2)	0.647		0.766		0.738
	Mobile device and PC (3)	0.694		0.756		0.708
Experience (years)	≤1 year (1)	0.694	p < 0.05 (1)–(3) p < 0.001 (1)–(5)	0.747	-	0.725	-
	2 years (2)	0.691		0.740		0.700
	3 years (3)	0.647		0.731		0.704
	4 years (4)	0.682		0.761		0.691
	≥5 years (5)	0.660		0.783		0.709
Frequency of log-on	Daily (1)	0.687	p < 0.05 (1)–(3)	0.757	p < 0.05 (4)–(3)	0.707	p < 0.05 (4)–(2)
	2–4 times a week (2)	0.641		0.757		0.678
	Once a week (3)	0.531		0.717		0.736
	2–4 times a month (4)	0.682		0.820		0.765
	Once a month (5)	0.665		0.778		0.633
Duration of log-on	≤1 h (1)	0.710	p < 0.05 (1)–(3 & 4)	0.759	-	0.723	-
	2 h (2)	0.657		0.732		0.702
	3 h (3)	0.618		0.741		0.686
	4 h (4)	0.593		0.751		0.667
	≥5 h (5)	0.698		0.766		0.711
YT commercial viewership #	None (1)	0.689	-	0.730	p < 0.05 (2, 3 & 5)–(1)	0.689	p < 0.05 (4)–(1)
	1–5 (2)	0.643		0.766		0.702
	6–10 (3)	0.718		0.763		0.731
	11–15 (4)	0.647		0.733		0.733
	≥16 (5)	0.646		0.775		0.699
Demographic variables
Gender	Male (1)	0.669	-	0.754	-	0.717	-
	Female (2)	0.677		0.748		0.694
Age	13–14 (1)	0.717	p < 0.05 (1)–(2 & 3)	0.724	-	0.742	p < 0.05 (1)–(2)
	15–16 (2)	0.675		0.757		0.691
	17–18 (3)	0.649		0.756		0.694
Race	White (1)	0.717	p < 0.05 (1)–(2, 3 & 4)	0.726	-	0.733	p < 0.05 (1)–(2)
	Black (2)	0.646		0.765		0.674
	Coloured (3)	0.658		0.752		0.710
	Indian/Asian (4)	0.639		0.756		0.693

for an overview of the significance of YT usage variables on traditional attitudinal associations and

Table 5. Hypothesis testing.

Hypothesis	Sub-Hypothesis	Significance	Support
H1	H1a	p < 0.001	Yes
	H1b	p < 0.001	Yes
	H1c	p < 0.001	Yes
H2	H2a	-	No
	H2b	-	No
	H2c	p < 0.05	Yes
	H2d	p < 0.05	Yes
H3	H3a	-	No
	H3b	-	No
	H3c	-	No
H4	H4a	p < 0.001	Yes
	H4b	-	No
	H4c	-	No
H5	H5a	p < 0.05	Yes
	H5b	p < 0.05	Yes
	H5c	p < 0.05	Yes
H6	H6a	p < 0.05	Yes
	H6b	-	No
	H6c	-	No
H7	H7a	-	No
	H7b	p < 0.05	Yes
	H7c	p < 0.05	Yes
H8	H8a	-	No
	H8b	-	No
	H8c	-	No
H9	H9a	p < 0.05	Yes
	H9b	-	No
	H9c	p < 0.05	Yes
H10	H10a	p < 0.05	Yes
	H10b	-	No
	H10c	p < 0.05	Yes

for an overview of the hypothesis testing results.

• H8: Gender. The standardised path coefficients indicated that gender did not result in a significant effect on the cognitive, affective and behavioural attitudinal associations of young consumers due to YMC. Hence, H8a, H8b, and H8c were not supported.
• H9: Age. The standardised path coefficient indicated that the awareness →knowledge was more favourable for young consumers who were aged 13–14 (β = 0.717, p < 0.05) versus those who were aged 15–16 (β = 0.675, p < 0.05) and 17–18 (β = 0.649, p < 0.05) years. The standardised path coefficients also revealed that intention-to-purchase had a positive influence on purchase for young consumers who were aged 13–14 (β = 0.742, p < 0.05) versus those who were aged 15–16 (β = 0.691, p < 0.05) years. Consequently, H9a and H9c were supported, whereas H9b was not supported.
• H10: Race. The standardised path coefficient indicated that the awareness→knowledge association was more favourable for young White consumers (β = 0.717, p < 0.05) versus young Black (β = 0.646, p < 0.05), Coloured (β = 0.658, p < 0.05), and Indian/Asian (β = 0.639, p < 0.05) consumers. The standardised path coefficients also revealed that intention-to-purchase had a positive influence on purchase for White (β = 0.733, p < 0.05) versus Black (β = 0.674, p < 0.05) young consumers. Therefore, H10a and H10c were supported, while H10b was not supported.

Refer to Table 4 for an overview of the significance and standardised beta coefficients of YT demographic variables on traditional attitudinal associations and Table 5 for an overview of the hypothesis testing results.

5. Discussion

5.1. YMC Influence on Young Consumers’ Attitudes

As discussed in prior text, several other studies [21,22,25,33,36,44,45,49,50,83] utilised Ducoffe’s [79] and Brackett and Carr’s [80] models to consider A_ad regarding various online ICT conduits, which mainly included informativeness/knowledge (cognitive attitudinal responses), entertainment (affective attitudinal responses), irritation, credibility, and/or other attitude components. Other studies also considered various consumer attitudinal responses towards YT and YMC. Lou et al. [50] revealed that relevant information and entertainment value of YT marketing resulted in favourable attitudes and increased purchase intentions among consumers. Arora and Agarwal [26] and Firat [49] established that information, entertainment, and/or purchase intention had a favourable influence on YT advertising value among Indian and Turkish Millennials, respectively. Göbel et al. [28] found that unfamiliar brands, without persuasion knowledge activation presented through user channels, led to more favourable attitudes towards the covert YMC and a greater intention to share among German students. Gupta et al. [29] found that YMC mostly gained Indian viewers’ attention and interest regarding the AIDA (attention, interest, desire, and action) model via message characteristics and appeals. Hansson and Stanic [30] reported that entertainment of sponsored YT content had a favourable influence on purchase intentions and attitudes among Generation Y viewers in Sweden. Rasmussen [41] ascertained a significant relationship in terms of YT celebrities’ popularity versus respondents’ feelings of knowing and enjoyment among US students. Rasmussen [41] also found that YT celebrities were seen as knowledgeable and commanded a strong influence on purchase via product recommendation.

However, most of the aforementioned inquiries were conducted in developed countries and/or only considered students (Generation Y). This research demonstrated that young African consumers shared similar positive attitudinal associations as a result of YMC. However, few of the above-mentioned studies include the full hierarchy-of-effects model’s cognitive, affective, and behavioural attitudinal associations, and the non-traditional attitudinal associations were also not examined, which were advocated by several researchers [12,78,98,99].

5.2. YMC Usage Variables’ Influence on Young Consumers’ Attitudes

Awareness was found to have a positive influence on knowledge among inexperienced (used YT for a year or less) South African Generation Z cohort members due to YMC. A supposition is that young consumers, who have used the OVDC for a longer (in terms of years), have become habituated to and were less likely to pay attention to YMC versus less experienced users. Several inquiries also determined that online and/or social media MC had a greater favourable influence on more inexperienced young consumers’ attitudinal responses, but these studies considered different online ICT platforms and Millennial consumers [85,115].

Generation Z cohort members who logged-on to YT more frequently showed a favourable cognitive attitudinal association. This finding is supported by a number of other investigations that also found social media and/or YT users who logged-on everyday display more favourable attitudinal responses [4,19,24,26,27,46,54,90]. Conversely, young consumers who logged on less frequently resulted in more favourable affective and behavioural attitudinal associations owing to YMC, and this finding is also reinforced by a number of inquiries [46,54,84,88,89]. This result is a fair notion, since it can be posited that young consumers who log-on more frequently are more likely to notice and respond to YMC, but higher attitudinal level responses (affective and behavioural) are not dependent on high log-on frequencies.

Awareness was also found to have a positive influence on knowledge among young consumers who spent shorter periods of time logged-on YT due to MC among the South African Generation Z cohort, which is supported by several studies [21,26]. One researcher agreed that YT users who spent shorter periods of time viewing YMC were more likely to exhibit more positive attitudinal responses [21], whereas another revealed that <1 h and >3 h result in the least positive attitudinal responses amid OVDC users [20]. Several studies found that younger consumers who spent longer periods of time on social media revealed more favourable attitudinal responses due to MC [88,89,90], but these studies considered different online ICT platforms.

South African Generation Z cohort members who viewed a greater number of YT commercials had more favourable affective and behavioural attitudinal associations in comparison to those who were not exposed to any YMC. This finding is an apt deduction, since a higher number of commercials viewed should result in more positive attitudinal responses, which is the one of the primary objectives of advertising. These findings are in consensus with other research, which found similar results regarding YMC in terms of cognitive, affective, and/or behavioural attitudinal responses [46,47,54], but these studies were conducted among the Generation Y cohort. However, there still tends to be a lack of research on the influence of the number of YT commercials viewed on attitudinal associations among young consumers.

5.3. YMC Demographic Variables’ Influence on Young Consumers’ Attitudes

Younger Generation Z consumers (aged 13–14 years old) were found to have more favourable cognitive and behavioural attitudinal associations, than older Generation Z consumers (15–16 and 17–18 years old) due to YMC. Other studies on social media and/or YT MC also found that younger age groups exhibited more favourable attitudinal responses [26,84], but these inquiries assessed different cohorts, viz. Millennial consumers. However, several social media MC studies examined Generation Z, in terms of the influence of age on attitudes, and found that older groups showed more favourable A_ad [88,89,90], but these studies did not investigate YMC. Nevertheless, the disparity between age groups is a reasonable supposition, since younger consumers would generally be more susceptible to YMC owing to less experience on this OVDC, which is reflected in a prior finding of this study. Hence, this inquiry confirms that heterogeneity is evident among different age groups within the Generation Z cohort owing to YMC.

Awareness resulted in a more positive influence on knowledge among young White consumers in comparison to young Black, Coloured, and Indian/Asian consumers due to MC on this OVDC. Intention-to-purchase was also found to have a positive influence on purchase among White versus Black young consumers owing to YMC. These finding are supported by a study, which also found more favourable attitudes among White consumers [26], but only considered Millennials and a different social media platform. However, the disparity between population groups is a reasonable notion, since despite the Black middle class spending power, education and employment opportunities having grown substantially since the end of apartheid (a quarter of a century ago), the pervasive socio-economic effects of apartheid still endures today. The income per capita of a majority previously disadvantaged individuals (Black, Coloured and Indian/Asian) lags behind the White counterparts. More than quarter of Black South Africans still lives in relative poverty, which is expected to significantly increase due to job losses as a consequence of the COVID-19 coronavirus pandemic. The White minority exhibit the highest income, employment, and education levels, which affords them greater access to social media, OVDC, and exposure to MC [66,74,75,106,109].

Therefore, this study affirms that several usage and demographic variable have a significant influence the traditional response hierarchy-of-effects model attitudinal associations owing to YMC amid young consumers, and is suggestive of heterogeneity in the Generation Z group in an African developing country.

6. Conclusions and Contributions to Knowledge and Practice

Generation Z is not only an important target group of organisations owing to the steady growth of these young consumers buying power, but also due to their huge influence on household purchases [7,8]. However, young consumers are extremely technologically savvy and commonly antagonistic regarding above-the-line and digital media MC [10,116], but OVDC have provided organisations with a sustainable means to more accurately target and reach Generation Z cohort (owing to detailed segmentation, metrics and analytical capabilities of YT) [71]. Additionally, the huge growth of mobile device usage in developing countries (such as SA) and broad range MC types available on OVDC platforms has extended the accessibility of organisations and their brands among young consumers [56,66]. YT has facilitated a mechanism for organisations to tailor product offerings, which will more accurately meet the interests, desires, needs, and wants by influencing young consumers’ attitudes via more sustainable MC available on this OVDC platform. Therefore, as previously discussed, a number of organisations have allocated significant portions of their MC budgets to YT [9,10].

Accordingly, young consumers’ attitudes towards YMC are crucial to organisations in creating more sustainable MC campaigns, as well as provide an understanding of behavioural predispositions in the future. This research affirms that YMC has a positive effect on the traditional awareness→knowledge, liking→preference, and intention-to-purchase→purchase) attitudinal associations, as well as on two non-traditional attitude associations, namely knowledge→preference and preference→purchase.

The favourable cognitive attitudinal relationship, due to YMC, functions as the initial catalyst stage of the hierarchy-of-effects model among young consumers. Accordingly, organisations should maintain sustainable MC via advertising, promotions, and product placement in an endeavour to stimulate awareness among Generation Z, which will ultimately initiate positive cognition and lead to the higher level hierarchical consumer responses. Therefore, organisations should aim to create credible, entertaining, and informative YMC content [21,22,25,33,36,44,45,49,50,83] in order to raise awareness on this OVDC platform.

Additionally, several organisations also utilise YouTubers, testimonials, influencers, and celebrity endorsers (who frequently have millions of young impressionable followers) to disseminate positive information and endorse the organisations via their YT channels. Several studies assert that YouTubers have a significantly positive relationship with the younger cohorts and have a major influence on these followers’ consumer decision-making process [34,38,41,43]. Westenberg [24] confirm that YouTubers exert a substantial influence on product and brand awareness among teenagers. Therefore, YouTubers are perceived as an important information source and use their authority as a means to influence consumer decision-making process, and to positively affect their young followers’ purchases by recommending specific products and services. Hence, organisations should employ YouTubers in an effort to stimulate awareness and disseminate credible product information, which is the ignition phase in stimulating the Generation Z cohort’s other attitudinal phases and, ultimately, influence future affective and behavioural responses.

The favourable affective attitudinal relationship, owing to YMC, also serves as the penultimate stage of the hierarchy-of-effects model among Generation Z consumers. Hence, organisations and marketers should aim to stimulate liking, which positively influences preference and ultimately leads to purchase among this young generation. Therefore, it is important that organisations constantly upload credible and informative material, which has a high entertainment value, since it generally arouses positive emotional responses among young consumers. The YT content could make use of affective advertising appeals and execution styles, such as humour, dramatization, fantasy, nostalgia, status (ego), and testimonials, which are presentation formats designed to evoke emotional attitudinal responses [21,29,33,40,44,56].

Some organisations also employ celebrity endorsers, influencers, and YouTubers to establish a relationship with young followers via regular interaction, which fosters emotional connections and is also perceived as a form of branded entertainment. Parasocial interaction is the notion whereby viewers develop an attraction and intimate feelings when they interact with celebrity endorsers, influencers, and YouTubers for a sustained period, and, therefore, aim to emulate analogous traits [41,55,62]. Rasmussen [41], Sokolova and Kefi [55], and Hwang et al. [62] reveal that YouTubers and social media celebrities are perceived as friends (who share their opinions) as a result of the parasocial interaction, but in reality are actually brand ambassadors. The respondents of the aforementioned studies found that young consumers developed emotional responses of familiarity, personality enjoyment, empathy and knowing as a result of the parasocial interaction. Consequently, YouTubers and social media celebrities are perceived as trustworthy and reliable information sources, which result in favourable behavioural responses. Hence, organisations should engage with YouTubers to promote their products and services as a sustainable MC strategy, to take advantage of the many positive benefits, and to evoke favourable affective attitudinal among young consumers.

The favourable behavioural attitudinal relationship, due to YMC, is the final stage of the hierarchy-of-effects model. Organisations could consider several important marketing tactics, viz. review, and YouTubers to stimulate positive behavioural responses among young consumers. Mir and Rehman [103] reveal that the number of reviews and views (comments and posts placed on YT by viewers) result in a favourable effect on perceived brand usefulness and credibility, which lead to favourable purchase intentions among young consumers. However, it is important that reviews and other UGC are not observed to originate from the organisation, but rather from other viewers so as to maintain the content’s credibility. Hence, organisations should encourage young consumers to upload reviews about their recent purchases, but should also prudently monitor the UGC to address to any negative comments. The consumer reviews could also be used to improve on promotional offers and even develop more innovative products and services.

Mir and Rehman [103] suggest that organisations should employ YouTubers to promote their products by embedding advertising messages in YT since younger generations generally perceive UGC as more credible. Rasmussen [41] found that American students were likely to buy products, which were suggested by YouTubers. Sokolova and Kefi [55] reveal that credibility and parasocial interaction by YouTubers and Instagram influencers result in favourable purchase intention associations among French Generation Z and Y respondents. Hwang et al. [62] report that social media celebrity parasocial relationships have a positive influence on purchase intentions among Chinese consumers. Thus, it is recommended that organisations should increase their use of YouTubers to endorse their products to have a positive influence on behavioural attitudinal responses, which is not only more cost-effective than traditional celebrities, but also an effective avenue to reach the younger digitally savvy cohorts.

Furthermore, this study found that young consumers who used YT for fewer years: logged on more frequently; spent shorter time periods on YT; viewed higher numbers of YT commercials; were aged 13–14 years old; and were from the White population group. These consumers displayed the most positive attitudinal responses to YMC. Hence, organisations and their brands should consider adjusting their YMC tactics to match the demographic and usage variable findings, so as to encourage positive attitudinal associations revealed by this inquiry [71].

Traditional advertising was used to develop a number of response hierarchy models to consider consumers’ attitudinal responses, which concludes with some form of behavioural response (e.g., purchase). Advertising practitioners postulated a number of theories to examine MC effectiveness via the analysis of various consumer hierarchical attitudinal stages, which culminate with the final purchase, but these theories were generally all formulated via traditional advertising mediums [12,78]. Hence, the appropriateness of traditional advertising theories is questionable regarding YMC, and is of significant importance to academics and MC practitioners due to the phenomenal growth of OVDC and other digital ICT channels over the past ten years. Few inquiries investigated YMC in terms of response hierarchy advertising theories or assessed the usage and demographics variables effect on the various attitudinal associations (traditional and non-traditional). Hence, as discussed in prior text, a number of inquiries have exposed social media MC research gaps, which mandate additional inquiry in terms of: YT [21,22,32,117,118]; developing countries [9,20,26,46,47,54,56,58,60,62,63,64]; different generational cohorts [46,47,54,60]; increased sample sizes [24,40,57,58]; quantitative inquiries [26,59,119], different attitude relationships [85,86,87,88,89,90]; and cross-examination of independent (e.g., usage and demographic) variables [20,25,30,56].

Consequently, this research has added to the limited knowledge repository (regarding the aforementioned gaps) by conducting research: on the largest OVDC platform (viz. YT); in a developing country; on the Generation Z cohort; using a large same size; adopting a quantitative research approach; considering traditional and non-traditional attitudinal associations; and conducting cross-examination of several usage and demographic variables. Therefore, this investigation study has contributed significant theoretical and practical discourse regarding hierarchy-of-effect model theory, attitude-to-advertising theory and cohort analysis. There is an overall dearth of academic inquiry regarding the impact of usage and demographic variables on attitudinal associations, owing to YMC, in a solitary generational cohort from an emerging nation perspective. Additionally, the study also showed that heterogeneity exists within this young generation. Hence, these findings offer novel information that will improve organisations’ prospect of taking advantage of digital marketing opportunities, via more sustainable MC strategies, by providing more exact estimates of Generation Z’s notoriously volatile purchase decision process and consumer behaviour.

7. Limitations and Further Research

This inquiry has several limitations, which provides an avenue for additional research in terms of YT and other social media platforms among different cohorts. Different SNS could be analysed to establish if similar attitudinal associations exist as a result of MC on these online ICT conduits. The inquiry was constrained to a single African developing country, whereas other developing and developed countries necessitate further research to ascertain the occurrence of analogous attitudinal, usage and demographic associations. Different socio-demographic variables (e.g., occupation, income, and education levels) influence on the attitudinal associations could be investigated by further inquiry [34]. The inquiry took a cross-section of the research population, whereas further research could explore qualitative and longitudinal approaches. The inquiry collectively assessed YMC, which comprises of promotions, advertising, YouTubers (influencers and celebrity endorsers), product placement, and testimonials, but did consider these different MC formats (e.g., display commercials, in-search discovery, skippable, non-skippable, overlay, sponsored cards, and bumpers) on individual basis or for specific brands, which affords opportunities for additional inquiry [120]. Only Generation Z was considered, whereas further research could examine other generations to discover if heterogeneity was also prevalent within these cohorts. The data for this study was collected prior to the global COVID-19 coronavirus pandemic, which may have influenced consumer attitudes across various online ICT platforms, so this study could be repeated to assess if there was a change in attitudes among young consumers due to YMC.