Jour of Business Resench 194 (2020) 122-142 Contents lists available a ScenceDisect Journal of Business Research journal homepage: ww.elsevier comilocatelbusres How social media influencers’ narrative strategies benefit cultivating influencer marketing: Tackling issues of cultural barriers, commercialised content, and sponsorship disclosure Shuang Zhou , Marta Blazquez.”, Helen McCormick, Liz Barnes ® 1 The hci of Teter al ting, The Hag Rog Pc Univers ag a, Koda Hag Kang, Oi " prof er Te Ur of Meter Sacrie Set Bg Mate M388, OF “Maver Ron hina, Machete Bon Obert, Ron B,C Shr, Machete HS 686, Sa mtn neces Calera bes Socal media inuncers (SMS) ae ares involving er, there ar ath opportuni and ies with infencer marking, SM camaive strategies can be of gent “ali s lghqality eWOM content i vital to mata the timaltorgnist-esponse (0-8) fan cr matting ess vwlthin the context of China's Kory markt. A qualitative app re Chinese consume perce Influencer keting to promt product. Hoe Influence marketing effective. This paper adapts ror toexpone the vali of Sle ariative strategist overcome che toward capabilities of these SMI’ mcratvestnteies Cre siidenityconstacion) to eal with the ses of uli bares, and sponsorship scl in SMe WOM. Bae on the researc fndigs tht poner develope m eonceptal mode! t iste hove SM" native sracgex Benefit the clinton of inf nce marketing om social mein 1, Introduction, Collaborating with social media influencers (SMls) ro leverage thelr electonie word-of-mouth (eWOM) has been utilised by brands as a strategic marketing communication instrument (Aulie2et, De Keres, ‘& Wowlaic, 2020). SMls are knowledgeable and expert tn particular topics, and thelr WOM concerning brands and product is pefceived by consumers at authentie and more tustworthy than marketer created messages (Kn) Kin, 2021). AB- significantly growing industry, the Gane prep of sep + Conse secepte f pnsnip ‘dha ins eMOM crete kar uae tee Jounal fines Row 194 C021) 122-142 5. Research framework ‘Ths study develops the research framework 9.1) to underpin the approach to explore the value of SMls" narrative strategies fr resolving problems with inluencer marketing. This framework i bull on the $0: R framework (Wehrabian & Rossel, 197°) to ilusteate the defined stim, organism and responses in the present study and euides data collection and analysis. The stimuli are the three SMIs" narrative sta- tceies, including brand steibute evaluation, brand love inspiration, ad selidentity constriction. This study explores the ofganism and response, which are about consumers’ perceptions and behavioural r- actions of SMI’ narrative strategies and the capabilities of thes stra tegies to overcome thre infiencer marketing sues 6, 5, Research methodology 6.1. Focus groups Focus group interviews with consumers were employed to unde: stand their perceptions toward the capabliies of SMe" narrative stra {egies to tackle the three Influencer marketing Issues, Rather than focusing on staistial generalisabilty, this qualitative method helps ‘obtain more in-depth insights into consumers’ thoughts, perceptions, And altitudes about the dee issues and the underlying reasons behind ther particular perceptions, which are dificult or imposible to detect ‘through quantitative research methods (Kiueyer & Casey, 2015) 6.1.1. Focus group discussion protocol ‘Before data collection, four types of open-ended questions (Fable ©) were formulated and utilised in focus group discussions, including ‘pening questions, introductory questions, key questions, and ending ‘questions, They ace derived fom the lteratute review and closely connected with four research questions and defined varlables in the research framework, These open ended questions are primary questions designed to guide participants’ discussions, Secondary questions, which can be elther openended or closed, are used to follow up primary ‘questions or probe for greater deal on participants’ answers to pimary questions, such as “why do you feel that way?” and “how strongly do you feel about this: strongly, very strongly?” 6.1.2. Focus group techniques out techniques, photo elicitation, fee listing, ranking, and choosing among alternatives, were used infocus group discussions. These tec niques are ‘exercises’ or ‘actvity-oriented questions’ that prove op portunities to elicit participants’ answers and promote discussion (reger & Casey, 2015), 6.1.2.1. Photo elctation. The photo elicitation technique (0, 200%) enables the focus group moderator to explain what is meant by SMls, WOM and narrative strategies, introduce three SMI’ narrative state ales (rand ateibute evaluation, brand love inspitation, andselfkleaity onstruction), and indice participants” discussion by tiggering the thoughts and attitudes towards particular subjects, The messages used for photo elicitation were selected and printed before focus group in terviews and then presented to the participants throughout each focus group discussion. Messages were selected from the posts published by the ten mort influential Chinese fashion bloggers (Pane BNP Paso 2017) on Weibo and WeChat. These messages are about products of vatioue Werterm luxury fashion brande and were copied directly without ‘modifying the content. The recognisable information about bloggers was removed to avold Influence on participant’ perceptions, The other ‘message selection evteria are representing the thee narrative strategies accurately and easly be read and comprehended “Three messages introducing different luxury brands were selected to present each narrative strategy, thereby eliminating posible influencesJounal fines Row 194 C021) 122-142 ‘Consumer perception of Sis" rare sry — Conwue erie capailty Consumer art Srausmmatvesrsenyic SP behavioural responses + Break cultural bariors * Balance commercia- personal + Diminish negative perception toward sponsorship disclosure Fig. 1. Hesereh feamewoek, from consumer biased perceptions and attitudes toward particular lux tury brands. These three messages are related to men’s and women’s luxury products, and each message is generally composed of 200-600 ‘words and several pitures. There are variations in the three messages to present each strategy to get rich results about participants’ perceptions For example, these messages embrace the two types of luxury brand ‘meanings: social meanings and individualised meanings. The selected messages also involve the non-sponsored! and sponsored messages, commercial elements, and other elements to effectively stimulate par tilpants” perceptions and diseussions on the moderators questions. 6.1.22. Free listing. A free listing task (Kreger & Casey, 2015) was ted to ask participants to list all their thoughts and feelings seilst reading the bloggers’ messages portraying three SMI’ narrative strate ‘ies, Participants were provided with pieces of paper for ving their answers. When everyone completed the tsk, the moderator asked par ticipants to diseuss what they have written down about each strategy ‘The key items and adjectives participants used for expressing percep tons toward each narrative strategy were recorded bythe moderator on Paper and quickly categorised based on thelr meanings and polaites, thereby allowing the moderator to ack probing questions about why parteipants ad sich perceptions and feelings. 6.1.23, Ranking, Alter the free listing task, a ranking task was used to ‘ask partieipants to rank the three SMIs'naratve strategies according to their liking fom the most preferred tothe least preferred. When par ticipants Finished the task, the moderator asked them to report their responses about the most and least liked narrative strategies. The ‘moderator wrote doven participants’ responses immediately, counted the results, and selected the most frequently mentioned strategies in thelr responses. Then patiipants were asked for thelr confirmation and ‘opinions on these strategies as the results of the most liked and most dlisliked narrative strategies inthe group's viewpoint. 6.1.24, Choosing among alternatives A task of choosing among alter: natives was applied to obtain partieipants" preference of two categories fof luxury brand meanings: social meanings and individualised mean: ings. The moderator introduced these meanings by giving examples taken from bloggers’ messages and then asked participant to select one ofthe two alternatives. Ater making ther selection, participants were asked to report their answers, provide reasons for their selection, and discuss the advantages and disadvantages of the two types of brand smesnings. 6.1.3. Sampling ‘The target population was primarily defined by country, gender, age ange, socal media usage experience, and interests in luxury brands. ‘The place of residence was also use to identify potential participants, ar consumers living in diferent tie cities are supposed to have different usury consumption behaviours (Fung Business Intelligence Cento, 2015; Liv eta, 2016). This esearch considered consumers living in two cities, Shanghai and Chengdu, which are Tier 1 and Tier 2 cities, respectively, and among the three largest luxury fashion matkete in Ghina (Fung Busines Intelligence Centre, 2015; Zheng, 2018; Song, 2018). This research thus aimed ata target population of Chinese Welbo fr Wechat users who are actual or potential luxury fashion consumers, Doth mate and female, aged between 18 and 40 yeats old, and living in Shanghai and Chengdu, ‘A snowball sampling technique was used to reruit participants for focus group discussions (Sekaran & Hou, 2015). Information sheets and consent forms were sent to each potential participant to inform them about the details of focus group discussione, ask them to answer sereening questions to decide the most appropriate sample, and obtain ‘their consent for participation. There are to 10 participants in a focus group diseussion (Stewart, Shamdasani, & Rook, 2007). This research ‘conducted three focus groupe in Shanghai and Cheng, respectively, fora total of six groups (n =53). The appendix displays the demographic detals of all participants 6.1.4. Focus group discussion ac focus group discussion lasted for around two hours. Ast, two pilot focus groups were undertaken with univesicy students to test the focus group protocols effectiveness to identity letues. After necessary ‘modifications to the protocol, six foeus groupe were conducted i Chengdu and Shanghai. These group discussions took place in comfortable envionment, sich asthe meeting room of «hotel located inthe city centre and café on a university campus, making parteipants feel relaxed nnd active in their discussions. All these discussione were condiicted in Mandarin, the native language of participants and the focus group moderator, enabling the participants to express perceptions and opinions precisely and preventing miscommunication between participants with the moderator, The moderator zed the words uxtry brands’ and ‘bloggers’ instead of polnting out specific luxury brand ‘name or blogger during questioning, thereby preventing bias in pate pants" responses that can compromise the valldlty of focus group Findings. All focus group discussions were audiorecorded for subse quent data analysis.Taleo Face group gestions. ‘tino non) the oH, 03, R09) “amon ‘rade “Teno ed ad dit Beek altura rer an) toe Batnce ommerca- personal 2017; Kaine eta, 200) ‘ime nam Gomecmeaey Sou + ts eo ecco ire Sc 1 Whieh pe a hey 1 Wie mratve stein ‘eke yo ta ht eying the mete ay Foie oak rc wo bl or eae alee the hy prc ye? Ist ea teeny an rant feline hissy ed, Isyaur peso ofthe ele ise ae + Hfcbargd postive change + Dosouthink he gs pont ate commerce ot ‘commer enter eggs eae? neg ee eptateo ‘mecha pera tern i Hog Jounal fines Row 194 C021) 122-142 Table 6 (onimet) “ypetgection Theme = ‘Sponsorship dldosie( + ht doy dink ue vg Jen 2016) Ss esas? + Wie eave sates ding qesioas Sigs fe How cate tee ipower 0 Sle sate tegen + Hynt actaaeto se 6.2, Dota analysis ‘The qualitative data collected in focus groups includes the audio recordings of focus group discussions and participants’ handwritten responses. The audio recordings were transcribed verbatim. The tra scripts and participants’ handwritten responses, which were in Chinese, were direety coded and analysed without translation into English to Avoid any translation errors of loss of meaning. A hybrid approach of deductive and inductive thematic analysis was adopted for analysing ‘and recording themes within the rch, complex, and lengthy accounts of dota. This hybrid approach could ensure rigour within qualitative research using thematic analysis (Fereday & Muit-Cochane, 2006). A step-by-step process of data coding and analysing was adopted, a shown inFig. 2 ‘There are seven steps inthe data analysis proces, beginning with developing priori codes and themes according to literature review and the research framework: A. Thee narrative strategies 1, Brand attribute evalation Brand love inspiration « Selfidentty constuction 8, Breaking cultural bariers 1. Delivering cultural values >, Maintaining luxury brands images and positions . Balancing commercial-pergonal tensions D. Sponsorship disclosure ‘Those predetermined themes and codes also correspond to the ‘questions guiding focus group discussion, supporting thie applicability to the raw data. The second step involved initial reading the tanseri nd parte}pants" handwritten responses and summarising the key points to cteate intial code. In the thd step, both deductive and inductive coding was performed using NVivo data management software, The ‘wanseripts were imported into NVivo and matched with a priori codes nd themes andthe initial codes developed manually inthe second step. ‘Then the data were segmented, sorted, and reorganised. The analysis was guided, yet not limited, by a priori codes and themes. Inductive codes were aso assigned tothe data segments that present anew theme (Goyatzis, 1999). These data-driven codes were either different fom the predetermined codes and themes or extended the initial codes. A wide ange of new themes emerged from data and enriched findings to full the purposes of this study. Inthe fourth step, data was re-read to continually create additonal Induetive codes, ensuring all the codes were refined to accurately represent the meanings of data Inthe ith step, codes vere connected to Identify themes. Then each theme was assigned conese phase to describe ite meaning. The sixth step Involved grouping the identified themes under overatching themes defined with headings corresponding to the research questions. nthe seventh step, the identified themes were reported with quotations from data,Ima of ae Rewer 13 21) 122-142 1. Develop a priori codes and themes Literature review ~ 37a Research framework 2. Read texts to be familiar with the data and create initial codes 3. Perform deductive and inductive coding ee 4, Continue to create data-driven codes by re-reading data 5. Connect the codes and identify themes and assign a lable to describe the underlying meaning of each theme 6. Identify overarching or core themes and then cluster themes under head- ings that directly relate to the research questions. 7. Report themes and illustrate them with a few representative quotations from the data Fig. 2. steps inthe pres f dats analy 6.3. Reliab and vata Bougie, 2013), Dependability, which signifies consistency and stability ‘A main coder did the tasks of data in measurements, was ensured by using researeh instruments ‘and snalysis, and three dhroughout the six focus groups, including predefined questions for supervisors monitored the overall data analysis proces to guarantee the focus group discussion and fout technologies of photo elicitation, free validity and reliability of research results. Two concepts of depend: listing, raking, and choosing among alternatives, Consistency and ability and aul bility ate wed for testing telisbility (Sehavan < stability of research instruments were maintained, making the foeus‘group discussions replicable in the future. In addition, deductive cdi {a priott codes and themes derived fiom the research framework) and Inductive coding (codes and themes emerging from data) were wed to censure the reliability of the codes and themes identified in the data analysis proces (Crbtues & Miler, 1999), audleabilty wae ensured by recording focus group discussions and displaying a igue (2.2) about data coding and analysis to provide sulficient information to support revealing and understanding the daa clletion and analysis processes (Appleton, 1995), Validity was tested through three notions: descriptive validity, interpretive validity, and theoretial validity (Srewer, 2000), Deseip tive validity was gained by transribing recordings of six focus group liscussons verbatim to offer a thorough description of the data collec tion process. Interpretive validity was ensured by repeating what par ticipants had said to confirm their responses during focus group discussions and asking them to clarify or elaborate when a response was unclear or under further investigation. Furthermore, phrases used by partcjpants in diseusions were drecty used for inital coding in the second step of data analysis process, helping stay as close to the aetual accounts 3¢ possible and maximise interpretative validity (Wax!) 1986). Theoteical validity was gained by adopting a research fame. ‘ork to guide designing questions for focus discussion, analysing data, nd interpreting results, This research frameworks built upon the S-0-R Tamework and lteratare review concerning potential problems wih infuencee marketing. ‘Other measures of validity were ako adopted. Using open-ended ‘questions in focus group discussions gives participants the opport nities to respond to questions in their own words to maximise validity (@berbach & Reckman, 2002). Two pilot studies were carried out to ‘examine the predefined questions and tasks for discussion to ensure sufficient and high-quality data. Besides, quotations of participants responses are used t0 report research findings. In doing so, the argu ‘ments made in connection with the dataset ean be rendered believable and credible (Thorne, 2000). 7. Results 7.1. Perceptions of SMIs’ narrative straepes (RQH) Partlcjpants* perceptions of three SMIs’ narrative strategies were understood via using a fee listing task and a ranking task and the sequent discussions on each strategy. The results show that preeptions fof narrative strategies ranged over several subjects, including. the narrative strategy itself, SMls’ eWOM created using this aarative strategy, and brands or products recommended inthe SMIs' WOM. The most liked narrative strategy is self-identity construction, whilst the most disliked narrative strategy is brand love inspiration. Fs. © presents 4 summary of participants" perception of the three narrative strategies. Selt-dentity construction is used by SMls to link identities of brands and products with personal identities or pertonaity traits in eWOM, thereby demonstrating the performance of brands and products for se ‘expression and identity constructions (iuniz & 0 Guinn, 2001) SMs usually associate the identities of themselves or celebrities with produets for brands to build archetypes in their eWOM. More than half of the participants peteeived this narative stategy as helping them better Understand the identitles and positions of brands. However, several Parcleipansertcised this narrative strategy and showed a concern that ‘the Brand images and consumer groupe could be defined and limited by the identity of archetypes in SMIs" WOM. As one participant claimed ‘Consumer impressions of brands canbe byfienced and lite by dentes of bloggers" (SHS fp). Their purchase intention can be negatively influ fenced by the disliked endorsers, Therefore, this narrative strategy can encourage both positive and negative spillover effets of endorsers on consumers’ attitudes towards the endorsed brands (Carllat, @ Astous, & Christians, 2014). ‘Moreover, several participants stated that they want to imitate the Jounal fines Row 194 C021) 122-142 outfit, styles, and brand selections of admired bloggers after reading their messages created by using this strategy. Tis consumer desire to {imitate is suggested by Ki sind Kin (2019) as the underlying mechanism {or Sil serving as taste leaders and opinion leaders influencing con: sumers to purchase thelr endorsed products, services, or brands. Also, these messages encouraged three participants to create a self referencing association, which i about the behaviour of mentally relating their own identity to the product. Sef-referencing narratives can create a self product connection (Aull, Slavieh, & De Kerviler 2015) an leads to a persuasion process of matative transportation that evokes consumer immersion into the content (Fsclss, 2007). Besides, the other perceptions of thie narrative strategy are about being effective in driving traffic to messages, enhancing brand awareness and brand liking, and making messages persuasive, attractive, and educational Participants were motivated to read SM eWOM meszages and had @ deep impression of these messages, and their purchase intention could be enhanced, Brand love inspiration, which isthe most disliked SMIs' narsative strategy e ueed to demonstrate an intense fanship ot worehip of brands by SM of other reference groupe (Kiets & De Valek, 2010) Partie pants" perceptions ofthis strategy present a striking contrast, ranging from extremely postive to extremely negative, Several parteipants argued that this narrative steategy highlights people's enthusiasm to ‘ward brands, arousing thei interest in reading messages and curiosity bout the brands and why people are passionate about them. This narrative strategy was also perceived as provocative to arouse readers empathy. However, more than half ofthe participants argued that SMI messages chow a lack of objectivity and an absence of logical reasons or arguments. A few participants also claimed that these messages are exaggerated, brainwashing, and stereotyping people. ‘Several participants claimed that this atategy could make them comply with others’ opinions (Shai, Mali, Akram, «© Chalsabart, 2017). They stated that when making a produet selection, if they are ‘unsure about whieh is best, they might feel that following what the tajority of people have brought is a safe option. One participant described her view vividly: “think this svategy could easly lead con sumer 0 Jump onthe bandwagon. For exanple, the message described the roductas a must have item fr fashionable men. The, some men who want to be fashionable would say ‘wow, I want to buy it, want to Be ssh." (SHF-p). Alo, afew pasteipants mentioned that this strategy could evoke their interests in the promoted brands, promote their acceptance ofthese brands, and encourage them to purchase. ‘Moreover, participants suggested that this strategy may be effective to target the dependent consumers who conform to others’ brand att tudes and consumption behaviours, These consumers may intend to ‘obtain a sense of relational connection or gain the reward of group acceptance (Nloural, Laroche, Pans, 2005). I may not be useful for influencing independent consumers who refuse compliance behaviour for prefer logical arguments to emotional stimulation. Furthermore, they also suggested that this method could be making SMIs" eWOM to be attractive forthe consumers whose emotions are easly affected by other people's moods and emotions, or by the situation at that moment. ‘Thetefore, consumers’ cultural value and personalities can be modera tors to the effectiveness ofthe brand love inspiration strategy ‘Brand attribute evaluation i sed by SMI show their judgment on the performance of products from certain perspectives It was positively perceived by participants as making SMIs’ WOM contain efficient and detailed descriptions of products and logical arguments. With this strategy, SMIs" eWOM are practical for them to evaluate products and select alternatives, encouraging their message involvement and pur- case intentions. As one participant said: “This strategy could be helpful fr consumers who have particular purchase needs or potential purchase incention to understand the aroibutes of a produc. Bloggers can give useful suggestions to consumers rough their comprehensive but focused evaluation ofthe characteristics and performances of this product. These messages are similar product reviews given by loggers" (SH-fp). Conversely, a fewS tio etat Ima of ae Rewer 13 21) 122-142 Pecepton toward arate satay “Hope cng wa 6 messages “Ercan contr ta Percepton over tance or preats Consuming ca be Star and od * Deceased purchase {Pace imi onan eos ipesione ‘oy ig 3. Patcpnns perceptions ofthe SMS naeative ratesparticipants perceived the SMIs' eWOM created by using this narrative strategy as commetcalised. As one parteipant claimed: “T feel dae Dloger" brand tkng ls noe autente. These messages are Jus commercial ‘promotions. Ths, I havea low interest in reading them” (SHfp). Several partlejpants also expressed that the expliit promotional intent ofthese -messages was unpleasant, This strategy can lend to SMIs" eWOM raising consumers’ suspicion ofthe authenticity of SMls' brand liking and the possibility of SMls being involved ina commercial relationship withthe ‘recommended brands, Furthermore, the purpose of SMIs' WOM is to persuade consumers to accept the opinions and claims displayed in eWOM to stimulate in formation adoption (Uzunoglu & Mise! Kip, 2014), An exploration of| consumer perceived persuasive attempts and persuasive strength of ‘SMS narrative strategies could help understand the capability of narrative strategies for cultural adaptation and persuading Chinese luxury consumers Patiipants were asked about which narrative stra ‘egies have explicit persuasive intent. The most voted narrative strate sles were brand love inspiration and brand attribute evaluation. ‘Surprisingly, when patilpants were asked o select the most persuasive narrative stvategies in their mind, the most choten narrative strategy ‘was self entity construction. The ndings show that overtly persuasive Sls eWOM mescages, such as the mesiages created by using brand love inspiration and brand attribute evaluation, could evoke consumers resistance to persuasion. Thus, Mls should use these two strategies in a ‘way of diluting persuasive intents 7.2, Perceptions toward capabilities of SMIs" narrative rages to deal with the three sues 7.2.1. Break culure bariers (RQ2) ‘AC ist, a task of choosing among alternatives was used to stimulate participants’ perceptions concerning the attractive meanings of Western luxury brands, Their responses and following discuscion revealed that, for most participants, the prefered luxury brand meanings are indi vidualised meanings rather than social meanings. Tie finding is inter esting, as it draws a difference from previous studies’ findings, which indiated that Chinese luxury consumers have an intense desire for symbolle meanings for socialising and showing off (Zhan & He, 2012). However, nore participants in Chengdu reported a preference fr social meanings than their counterparts in Shanghai. Henee, the extent of cultural diversity in the city of residence influences Chinese consumer adoption of cultural values and leads to differences in Chinese Iuxsry consumption behaviours. able 7 displays a summary ofthe perceptions fof the two types of luxury brand meanings Social meaning refers to the conventionally shared symbolic mesa ing of luxury brands by social groups, such as Weibo ot WeChat users brand communities, or cultural groups such as hip-hop fashion fol lowers, Social meaning is perceived as helping to veduce the uncertainty consumers may feel towards purchasing a product. One respondent highlighted the capability of uch meaning for product ditfereniation: eis more helpful for diferentiaing thls bag from others" (CDS-fp). She also mentioned that these meanings were helpful for her to reference Public evaluations on products or brands when making a purchase de cision: “Taking ths dress as an example what are the public evaluations on its performance? Is ithe one want to own? Or whether my evaluation of consistent with mass opinions?" (CDS). Another participant indicated his belief in the reliability of publi preferences for products: “Afr: ‘mass opinions are more important and reliable than inavidual eetings. fel like I won't go wrong if I buy this product” (SH-mp). There ate several Potential reasons for partielpants taking the meanings derived from social groupe into consideration when evaluating products, From 2 cultural perspective, China's collectivstic cultures promote consumere to structure their judgments in the ligt of group norms (Zhan & He, 2012). From a consumer behaviour perspective, consumers are likely to accept symbolic meanings given by reference groups o satisfy theirneed {or peyehological acocition with these groupe (Eecalas & Bettaan, Jounal fines Row 194 C021) 122-142 ‘Table? Participants’ perceptions of sical and indvidalised mennings attached to usury rans ‘Socal meanings of sary Brands ears of prcpins Pectin Poe Fpl opts» Dining prod om septs” afresh ‘tse 4 esas a ftw pe Peeetions ‘rept tier coomners {parat formeking 5 Need red thee nig 0 Negative Romanlact emrans dott nid fing at Spies ben apa 2005), Alo, this parteipants’ perception can be linked with the cl lective hoxury consumption behaviour that signifies consumers pur chasing certain luxury products since these are very popular among the masses, regardless oftheir beliefs (Shik tal, 2017). By contrast, one participant had an intense feeling of disgust and antagonistic sentiments toward social meanings of luxury brands, She stated that.“ don’ ike to give labeler products or people. You desribed this productos Being the hotest sling tom in China’, and tus, Fdon'ewant to buy icany more” (CDI-p). Tis participant may have an independent cvientation and is unlikely to conform to group norms (Sia eta, 2017). Similarly, one respondent expressed her feeling of dislike to ‘wards the meanings that encourage conformity behaviours: “ aemally ‘think ha his ype of meanings not ad, fr example, niche and chi’ and the item that ean Be matched with anying™ 1 feel that these terms are acceptable. But I don't like other tee ike ‘the hotest zoe online’. Hah, my pereption ts based on the words and phrases they used to interpret the ‘meanings’ (CD2,p). Therefore, Mls should carefull interpret the social ‘meanings as mass accepted meanings to signify the feature, benefits, land distinctiveness of ‘brands rather than persuading. consumer conformity, Unlike social meanings derived from social groups, individualised meanings are about subjective meanings that a luxury brand holds for a particular individual. All the three SMls° narrative strategies could help to interpret the individualised luxury brand meanings to consumers. (One participant indicated her interest was aroused by the meanings derived from other consumers. Alzo, one participant expressed her desire for individualised meanings about products, regarding these ‘meanings ae those“ most wantto read and lean about" (CDI -p). Another respondent claimed that: “I relly need to read the personalised meanings siven by others to decde whether oF not to buy chs product” (CDI fp). Be sides, several partleipants chowght tha individualised meanings might bye lest referential for themselves rather than having strong feelings of| slike toward these meanings. One patiipant elucidated that she was indifferent to these meanings: “le & merely about hie own opinions and personalized information” (SH2fp). Another participant also showed her concern about this by saying: “These words are about the fashion bloggers‘own felings and evaluations, which may apply t himself ur not ro me (stip. “The subsequent partlipants'decusson is about pereeptions toward narrative ststegies’ capabilities to convey selfexpressive value and social value of huxuty brands to them. To target Chinese consumes who tive significance to self related values of luxury goods, SMis should tailor their eWOM to encourage personal connection with the recom: mended brand. Accordingly, participants were asked to select the narrative strategies that stimulate the connection between their sel concept and brand images. The mest chosen narrative strategies were slidentity construction, This narative strategy highlights brands oF products! contibutions 10 identity construction and self-defining behaviour, facilitating participants mentally picturing themselves rela tive to the recommended product in SMls’ messages. Then participants ‘were asked to choose the SMle' narrative strategies that help signal Achievement and socal statue and affirm community membership. The ‘most preferred narrative strategies are brand love inspiration and sel deni construction. ‘To sueceed in influencer marketing eampaigns in China's luxury market ite necessary to convey the globally standardised brand iden- tigy and postions to consumers and maintain consistency of brand image in SMIs" eWOM. Thus, the second part of participants’ discussion pre: sents two main themes regarding participants’ pereeptions toward the capabilities of narrative strateges to maintain consistent international Tuxury brand images. Parteipants were asked about which narrative strategies could be perceived as capable of representing the image and positon of luxury brands in bloggers’ messages. The most mentioned narrative strategy was self-identity constuction. Then the moderator ‘asked participants about which narrative strategies could change th Impressions of Western luxury brand’ images instead of direct asking ‘them about perceptions of narrative strategies that could maintain consistency in luxury brands’ images. Some paticipants claimed that brand Jove inspiration and brand actibute evaluation could exert ‘negative changes in thelr impressions of luxury brands’ images, whereas salFidentity construction positively changed their impressions and Improved their understanding about luxury brands’ images. Thus, ac cording to participants’ responses, self-identity construction ean help ‘maintain consstney in luxury brands" images. ‘To summarise, the findings suggest that self.dentty construction i the most compelling narrative strategy for breaking cultural batiers in ‘SM eWOM Participants" perceptions show that this narrative strategy has the repertoire to deal with the issues used as measures for the capability to break cultural barriers, including interpreting the preferred luxury brand meanings, conveying social values of luxury brands, stimulating linking self-concept with brand images, portraying luxury brands’ identities and positions, and maintaining consistency in luxury brands’ images. Therefore, selPidentity construction can connect ‘Western luxury brand identities to the cultural values and ideology of ‘Chinese consumers. This capability supports the role of SMls as eultral Intermediaries to transform the meanings of luxury brands into socal and culturally adjusted meanings for the consumers in a particular ‘market (Delisle & Parmentier, 2016). ‘rablee Pats perceptions of commercial elements Sls WOM. Jounal fines Row 194 C021) 122-142 7.2.2. Resolve commercatpersonal tnson in SMls' WOM (RQS) ‘Photo elieltation was adopted to evoke paytielpants' thoughts and perceptions abour the value ofthree SMls'narative strategies to smooth the tension in commercial and personal content in Shs" eWOM. The fst step exploted participants’ perceptions ofthe commercial elements in bloggers’ messages, including prices, giveaways, discount codes, competitions, and purchase links. Their discussions ranged over several subjects, such asthe issues being questioned, SMls, €WOM, brands, and products, A summaty of participants’ perceptions about this issue is shovin in Table Participant’ perceptions reveal that commercial elements would not lead to consumers’ negative attude toward eWOM or negatively affect consumers’ willingness to purchase the recommended products. Inter tingly, most of them stated that commercial elements are acceptable Dbecauze they vant to know the information concerning the purchace of luxury brand, auch ae price, discounts, purchase links, or store loca- tions. These pieces of information can help them make price compati- sons and sequential purchasing behavious.SMIs embedding commereal lemente in eWOM could save their tne to search for silat informa- tion, One respondent indicated that getting uted tobeing surrounded by advertising in daly life is a reason for being comfortable with com mercial content. As he stated: "Because we normally recelvet00 many ‘advertisements onthe phone, Thus fel tha this sort of content accep able” (SH3.mp), ‘Another patticipant conveyed her positive attitude toward con ‘mercial information in bloggers’ message, even though she might not pay attention toi. She sad: “This (messages containing commercial ee ment) is acceptable because I read bloggers’ posts merely for obtaining i {formation about products. {probably would not click the hypertink provided by him. But, T would not have negative feelings about the product recon ‘mended nhs poet even if find the post hasbeen commercialized” (CD2-f) ‘These postive perceptions contrast to Wester studies, which suggest that commercial content in SMls' eWOM could reduce the persuasive effect of messages on consumer brand atitudes and purchase behaviour (Halvorsen ota, 2013; Pl & Sandstsm, 2013; Boerman eal, 2017). However, these pareeptions are in lie with McKinsey & Company's (2017) descriptions of Chinese luxury consumption behaviour, whieh Include price sensitivity and awaveness of price gaps in diferent mar kets, and needing reassurance about luxury products quality, then: ticity and price. Alo, several paticipants claimed that the commereal lements in SMIs' WOM may make them feel that these messages are created genuinely and honestly and therefore worthy of belie. Thus, commercial elements may somewhat enhance the credibility of SMI: wom, ‘There are stil somewhat negative participant perceptions about Sis and products, including decreased expectations on products and 2 negative impression of SMIs. One participant mentioned that his expectation ofthe recommended product was downgraded after being ware ofthe commercial content: "When I read a blogser' message, if ‘hink the bxroduced producti okay and then I realize dat ii an adver ‘isement, my expectation of ic decreases” (SHS-mp). These negative per ceptions show that Chinese consumers high and positive expectations nore of prep ecg of prep Neg pecetons Peas oad tran prs 1 negate reson 2 ed tow {ign oi seed by adver iy le 1 heli tego he + Espciton the prodit ected 1 Gaertn el psn of Te Hoge {ein god ingreson of he nse‘on products could be lowered after realising the commercial content in ‘Sls’ eWOM. They may doubt that SMls" positive eWOM about the produets ls ereated due to brands sponsorship. Next, patiipants discussed how to recognise commercial content Many partlelpants stated that sometimes they could not realise whether the price and purchase links shown in bloggers’ mestagesaze commer cialised contents, or derived from bloggers’ personal experienees and the desire to share useful information with readers. This finding i= similar to the argument in Western studies that it could be dificult for consumers to distinguish commercial content from personal content in bloggers’ messages (Boe:nan etal, 2017). The ways used by partici: ‘pants to eeparate the commercial content fom personal content are looking for personal experiences depicted in bloggers’ messages and recognising the waiting style used by bloggers. The descriptions about personal experiences with products and the sentences writen in @ colloquial style would be considered as personal content. In the third step, participants were encouraged to discuss thelr per ceptions about the narrative stateges that could smooth commer clal-personal tensions in bloggers’ eWOM. They mentioned that brand attribute evaluation has such potential, This finding i interes ‘arratve strategy was negatively perceived as potentisly commercia ised. According to thet diseusions on other relevant topics such as sponsorship diselooure, this narrative stiategy may make commereil clements in bloggers’ messages reasonable. Thus, they may have a tolerant atitude towards commercial content and read it to full 3 purpose of information seeking, thereby having positive perceptions of ‘commercial information, a elucidated in the preceding section. 7.23. Mitigate negative perceptions abour sponsorship disclosure (RQ) Participants were informed about Chinas regulation on revealing 3 ald partnership between bloggers and the sponsoring brand, and ‘questioned their perceptions of such disclosure in SMls' eWOM. Con- ‘uaty to the findings from prior studies, most participants considered sponsorship disclosure acceptable, except forfour participants who had ‘egative feelings about i. Tels peteeptions are displayed in Table 9 ‘Some participants stated that i was acceptable and understandable for them that bloggers receive materials in exchange for promoting produetson their blogs, eepecally those bloggers who make a living rom blogging, As one participant elucidated: “I hink iis acceptable. The Income of a Blogger comes from buldng a large fon Base, catching great atenion, and collaborating with brands" (SHS fo). Several participants mentioned that sponsorship disclosure could evoke their rational ‘thinking. One participant sald: fhe indicates the paid parmership with ‘other pares in his message, Iwill ue rational inking when reading these messages" (SH3-f). Another participant supported this claim by saying "Because his mesage provides he information Ince, dclosing sponsorship ‘makes me read thie meszage rationally, rather than blindly following hie recommendation”. For other participants, sponsorship disclosure made ‘Ss’ messages more truthful as they honestly reveal the sponsorship, so ‘they are not trying to mislead or cheat them. However, four patiipants had negative peseeptions relating to sponsorship’ disclosure, eWOM, and SMIs. Por one participant, ‘rable Pts perceptions of sponsrsip dalosre i Ss WOM. Jounal fines Row 194 C021) 122-142 sponsorship disclosure was non-acceptable, and it could have a poten- tilly negative effect on produet purchase. As he stated: “lam Indiferent to this mesiage Because Ii just an advertisement... mean, you may think ‘har disclosing sponsorship i good and transparent. However it may have @ negative effet wien you read the message o help you o make a purchase decision” (SHI-mp), Fr the rest of them, revealing a paid partnership ‘would make them have negative impressions of these bloggers, and they ‘would be less wiling to read the sponsored message after recognising it asadvertising, When researchers asked them about the reasons for their negative perceptions, one participant responded: “As followers normally expect that bloggers can provide some useful information derived from thir ‘wn viewpoint, rather than commercialized infarmation” (SH). Participants’ perceptions reveal that sponsorship disclose in SMe WOM seems to be acceptable to Chinese consumers. This finding is diferent from consumers’ negative perceptions of sponsorship dislo sure found in Korea (Hwang & Jeong, 2016) and the Netherlands (Goornan et al, 2017) Participants argued that sponsorship disclosure could make SMls' WOM more authentle and provoke rational thoughts about the message content. Also, even if some consumers’ negative peteeptions about sponsorship disclosure are evident, yet no negative perceptions toward sponsors, brands, ot produets introduced in SMIs sponsored eWOM were found, Participants were then asked to choose the narrative strategies capable of making sponsorship disclosure in bloggers’ eWOM more acceptable. The most seleted narrative strategies were brand attribute evaluation and self-identity construction. This finding challenges the suggestion by Auliezct tl. 2020) that the optimal solution for Sisto manage authenticity of content is to express their passion for the endorsed brand when their commercial partnership with this brand is ‘wansparent in their eWOM The findings of present research indicate that brand love ineplration, which is used by SM to express worship ‘and passion forthe recommended brands, may not be ueeul for making sponsorship transparency acceptable to consumer. @. Discussion Brands nd SMI should be wel prepared to embrace the challenges and opportunities brought about by inflencer marketing. SMls can in fluence consumers by distributing eWOM and develop and maintain close relationships with consumers (Pak & Bosch, 2019), However, there are potential risks of losing their eWOM effects, such as over commercialised content and sponsorship disclosure in their eWOM (Gialvorven etal, 2013; Boermas etal, 2017). To provide insights into this area, this research explored the value of three SMIs’ narrative strategies, brand attribute evaluation, brand love inepration, ad self identity construction, to overcome potential issues with influencer ‘marketing ‘Amatix chechlist (iy. is given to illustrate consumer perceptions ofthe three narrative strategies and their eapablities for dealing with the Issues of breaking cultural bares, releasing commeclal- personal tension, and reducing the negative influence of sponsorship disclosure. ‘The results reveal chat selCidentiy construction could break cultural in eneprs of pepon Seon oF Perens deci Negntive rcs {Listed tard Se + Aceptble se Sesame + Specendip rasa of Scone «Reid net oe ponte meng 2 Att meson f ported ops‘The most liked SMls' narrative strategies ‘The most disiked SMis' narrative strategies The Sis narative strategies with expt ee seeps mn ei ces bey tie rm as ee ‘0 maintain consistency in Capabity | ey in ry Ima of ae Rewer 13 21) 122-142 — seen ‘owardsporaordie isco lg. 4 A mates checklist of participants perceptions ofthe value of SMa" mratve sree. Darsiers to interpret consumer preferred luxury brand meanings, convey consumer desired values of luxury brands, portmy luxury brands’ ldentities and positions, and maintain consistency in luxury brands’ Images. Moreover, brand attsbute evaluation may help balance com: ‘mercial-personal tensions. It ean make commercial elements in SMls* ‘WOM reasonable to consumers and evoke consumer tolerant atitude towards commercial content, thereby making consumers read SMls* ‘eWOM to seek information about purchasing luxury brands. Finally, ‘negative peteeptions about sponsorship disclosure could be dilated by Stimulus ‘a rare od srategy poe ‘sing brand atribute evaluation and self identity constuction. "ig. 5 shows a conceptual model to ilustrate the potential factors that moderate SMIs' narrative strategies’ contributions to resolving Influencer marketing Isues. Cultural values ean be a significant factor affecting the effectiveness of SMle' narrative strategy. Thie study lok Into China's luxury market context where the consumers hold deep- rooted collectiviet cultural values and diversified modern Western cul- ‘tual values (Zan & He, 2012). As displayed by the stiking contrasted consumers” perceptions toward the narrative strategy of brand love Response Coed (Sonsumer ose’ shawl Petey ig 5, Potent fctrs in uence the value of SMS native statis to overcome ses wit inuenece makeingInspiration, consumers with diferent cultural values can have extremely positive or extremely negative perceptions toward this suategy. Con: fumer perception ofthis strategy also highlights the role of consumer personality in inluencing the stategy'selfectivenes, as displayed in Fig. 5. i the group discussions, consumers analysed the personalities making people be attracted to SMls’ eWOM created using brand love Ima of ae Rewer 13 21) 122-142 Inspiration. ‘Another influential factor is consumer expectation. According t0 ‘consumer perceptions, when they hold high and postive expectations of| the recommended product aad SMls, their expectations ean be dow: saded by the commereal element and sponsotship disclosure in SMI ‘eWOM. These dereased expectations may result in consumer scepticina / Using ive Strategies for Developing \ Influencer Marketing Campaigns (io elon toe wat ara cotoad Paid infuancer Eamod Influencer Marketing oni tareting ea fe (Sxoengeuz sgh styling) Media singing) et uarhting Objects fet uarhting OD cee ase ate TSR ‘and Selection <- Strategies > for Storygiving oo Aras aca caters ices teniones ere ae at rere teen anaes Seon Se operas Cony oi vo oneceeteerac arta ens Noratin unt at conten ceson ex Neate, Sti stan (os tty erence) Nara agen canbe epee oes puaaaicalve (og uteaneyarorcar) ‘Cuttural barriers “Maintain consistency in ee errant emerson) a og ntaney crore) —— perry ny Tred TEE aan prc ce nee ¥ aT SaaS etme y =o ae ' Measure Fosults ig 6, Concept model of wing marae states to culate infencermacketing.ofthe authenticity of SMIs' postive eWOM and a reduced intention to read SMIs" WOM. Besides, consumers’ motivation to read SMls' WOM and marketing environment can also affect thelr perceptions of om: mercial elements and sponsorship disclosure. The commercial elements In SMIs* eWOM, sueh as priee, discounts, and purchase links, ean be positively perceived by consumers when they read SMIs' eWOM for ‘brand information acquisition. A marketing environment ofthe preva lence of commercial advertising in consumers’ daly lives can make consumers tolerant and comfortable with commercial content ‘The underlying mechanism of persuasion needs to be considered ‘when bamessng the value of SMIs" narrative strategies for breaking cultural barviers. This study found that the most liked selfidentity construction strategy can persuade purchase intention by evoking @ ‘mechanism of consumer desire to imitate ( & Kin, 2019) in contrast, ‘the most disliked brand love inspiration strategy war eitcized by con ‘sumers as projecting an intense persuasive intent ro stimulate consumer ‘conformity. Consumers demonstrated a desire for SMIs’ logical argu ‘meat on thelr Brand love. Similarly, for SMs interpreting social mean: Ings to them, consumers showed thelr preference to the meanings presenting the features, benefits, and distinctiveness of brands rather than the meanings encouraging conformity. Also, la terms of pereeption toward sponsorship disclosure, consumets mentioned that their tational thinking could be simulated by sponsorship disclosure. According! ‘Sis should adopt logical reasoning in arguments to creat theit eWOM ‘when thelr commercial partnership with brands i revealed ‘Based on the findings of this researc, a conceptual model (Pi. 6) created ¢o demonstrate how to se SM narrative strategies for cut: vating two types of influencer marketing: paid influencer marketing and ‘eamed influencer marketing. MIs’ narrative strategies could be used for the most significant steps in the preparation stage of these two ‘marketing campaigns. ald influencer marketing, which i a high-level of marketer: controlled campaigns, alms to collaborate with SMIs to ereate and spread content about a brand or new product to shape consumers a titudes and influence their Behaviours (Uno Mise Kip, 2014), ‘Aer seting marketing objectives, selecting SMs and producing crea tive content could be essential for the initial stage of sald intlueneer ‘marketing eampaign (Kovinets etal, 2070; Aucezet etal, 2020). The study suggests that narrative strategies may help in these two steps. First, marketers may use narrative suategies ae standards for content and structure in their analysis and selection of SW, thereby choosing the one whose content is the best fit for the marketing campaign and brand image. For instance, matketers may consider thatthe messages created by using self.identity construction are appropriate for a ‘campaign. Thus, they can select the SM who ate proficient at using thie narrative strategy. Marketers may also use narrative strategies as structures to measure SMIs’ content to choose the ane who can create the content that i ideal for their campaign. ‘Second, marketers can collaborate with the selected SMle to use narrative strategies as guidelines for ereating eWOM. In this way, narrative strategies can alko help deal with the potential Issues sur. rounding the marketing campaign, such as cultural baries, commer clal-petsonal tension in eWOM, and sponsorship disclosure. For ‘example, the findings from this study indicate that seidentity con: struction may be useful for breaking cultural arsers, Brand attibute ‘evaluation may help balance commercial-personal tension, Negative perceptions of sponsorship dislosure can be mitigated by brand att bute evaluation and elf identity construction, Narrative strategies ean aso be used to create common theme for developing earned influencer marketing inthe process of story giving, thereby targeting and attracting the Shs who do not receive compen: sation to promote brands. Story-iving changes the narrative from the brand's perspective to that of consumers, offering brands the oppor nity wo eo-create brand images with consumers by soliciting them to share their narratives and personal brand stories on socal media (Wlighes et al, 2016). A buza could be generated because of the high Jounal fines Row 194 C021) 122-142 credibility of consumer generated content, making story giving more engaging and compelling (Hewett, Rand, Rast, & van Heetde, 2016). Although It is impossible to contol the interaction and conversations anvong consumers, stry-giving ensues that marketers da not have to Telinguish control of consumers’ information creation completly ‘Marketers ean exeate a common theme for consumer stories (iee8 (a, 2010). This is the step in which narrative strategies could be ‘sell ‘Narrative strategies can be referenced and adopted as guidelines for selecting and defining the theme for consumer-generated stories. For instance, marketers can adopt brand love inspiration to exeat the topic, ‘char sharing personal experience of purchasing a particular brand t0 express fanship in a meaningful way. With a well-defined theme to encourage buzz on socal media, consumers and SMIs may be attracted by the story giving activity and proactively start conversations about i Moreover, the purposes of conducting story-giving activities may be ‘ising consumer attention toa brand an co-crention of brand meanings ‘without duting the established brand image in the process (Flushes ral, 20/6). Accordingly, this esearch suggests that the common theme defined for a story giving activity should signify the social values of a brand, stimulate selbrand association, represent the brand's ideaticy And postion, and maintain consistency inthe brand image, Narrative strategies could help achieve these goals, as evidenced inthe research resulls ‘To summarise, this model shows the capacity and approaches of narrative strategies to benefit ifluencer marketing on social media [Narrative strategies could be wseful for selecting SMe and creating content inthe development of paid influencer marketing campaigns, and ‘resting a common theme foreamed influence marketing inthe procest of story-giving. Specific types of narrative strategies displayed in the ‘model were recommended based on the findings ofthis study, which indicates Chinese consumer’ positive perceptions toward SMIs' narra- tive strategies for creating eWOM about hixury brands, They can be changed when thie model is ueed for informing influencer matketing conducted to promote other product categories or in different eultural 9. Implications, limitations, and future research 9.1. Theoreticlinpizaions ‘This study conuibutes to influences marketing literature (Tylon, 2020) by shedding ight on resolving potential inluencer marketing problems. First, this study theorises the value of SMis* narative stra ‘egies to overcome content creation isis in infuencer marketing, Prior literature has acknowledged that lack of control over content creation, ‘commercialised content, and sponsorship disclosure in SMls eWOM can lead to negative consumer responses (nls & Borchers, 2019; De Cicea eral, 2020; Audteze etal, 2020). However, the approach to cope with ‘these challenges to diminish potential negative consumer responses is ‘underexplored. Tis research fils this gap and finds that SMIs' narrative strategies can help break cultural bartiere balance commercial-personal tension, and mitigate negative consumer perceptions of sponcorship disclosure in SMIs' WOM. By discussing research findings, conceptual ‘model iy. 6) is developed to ilustrae the ways to use SMIs'narative strategies in developing two types of influencer marketing: paid infu fencer marketing and earned influences marketing. ‘Second, this research is one ofthe frst attempts to take an interna tional perspective to explore overcoming infeneer matketing estes to target foreign markets, This research answers the call for further research on the methods to develop influencer marketing to overcome challenges brought by cultura differences (Dwivedi et al, 2020). Based fon the literature review, this esearch suggests four factors to examine the capability of SMis' narrative strategies to cope with eultural differ fences: a). interpret local consumers’ prefered brond meanings, b) jer local consumers’ desired values of brands, ). portray an