Research Article: 2018 Vol: 22 Issue: 3
Sookyeong Hong, Hansei University
Jin-Ae Kang, East Carolina University
Glenn T Hubbard, East Carolina University
This experimental study compared story and non-story audio advertisements for a brand of an online start-up company, using three test groups of college students from a large university. Some participants reported more favourable responses to ads using stories than those hearing a control ad that was purely informational. There were also significant differences regarding participants’ desire to share information about the product by Word Of Mouth (WOM). Participants with high levels of involvement in the product were more responsive to the informational ad, whereas those with low involvement responded more favourably to a customer telling his story about using the product. Another finding showed that low-involvement participants responded more favourably to the purely informational ad regarding their desire to share information about the product by word of mouth. This exploratory study provides useful insights for future research pertaining to the use of storytelling in advertising and also in terms of involvement levels and central versus peripheral processing.
Storytelling, Advertising, Founder, Involvement and Customer.
As interest in brand storytelling grows, companies are engaging in more storytelling marketing in advertising or public campaigns, combining with social media. Brand storytelling is defined as an activity of a company constructing a story about the CEO, the value of the company, products, employees, customers and so on. Researchers comment storytelling enables a company to promote its brand better and build an effective communication channel with consumers. Companies include brand story sections on their websites more than ever before. Therefore, it becomes critical for a company to decide whose story should be created and which brand elements need to be highlighted to produce the most effective brand story. However, it is not simple to verify the effect of brand storytelling because storytelling involves diverse variables such as consumers’ narrative interpretation level, thought styles (paradigmatic or narrative) and story comprehension ability. Based on true stories and information with a true brand of a start-up company, this experimental study is designed to verify the effect of storytelling and test whose story-a founder/owner or a customer-might is more effective compared to the other one.
Scholars show a lot of disagreements and conflicts in defining a story and storytelling because many different areas such as literature, linguistics and psychology define the concepts in their own ways. This study, therefore, does not deal with contradictions in the discussion of story and storytelling but rather adopts fundamental notions from the perspective of communication and marketing.
A story is regarded as “chronological order of events,” (Abbott, 2000) and it consists of a beginning, middle, and end. A narrative is defined as “the representation of a real or fictitious event or series of events” (Genette and Levonas, 1976). Precisely, ‘narrative’ is more engaged in written form, and ‘story’ is more engaged in oral form. During the epoch when the oral statement had been predominant, people used ‘story' and ‘storytelling’ but when the text got the power, the term ‘narrative’ became popular. With the advent of the digital age, images and oral expressions are commonly used, which it brings forth the reappearance of ‘story’ and ‘storytelling'. In the field of marketing and communication, stories and narratives are often used as synonyms. Thus storytelling (narrating) is defined as the vivid description of ideas, beliefs, personal experiences and lifelessons through stories or narratives that evoke powerful emotions and insights (Serrat, 2010). Storytelling emphasizes the act of telling a story and the storyteller can deliver a story through media. In advertising, the storyteller often is a CEO, employee, consumer or model and the audience is someone who is listening (reading or watching) the story. Thus, the audience should be a potential consumer.
Bruner (1990) proposes four elements of story construction: agents engaged in actions to achieve goals, sequential order, canonical form and voice. If there is a story, there must be events and the narrative is a set of sequential events (Smith, 1981). The event in a story is essential because there must be something needing to be fixed. It is necessary to remove the abnormal factors and revert to normal. The protagonist (the main character) moves toward the goal and completes his/her mission in different time. With those elements, the storyteller (narrator) should have his/her perspective toward society and the world. These elements are connected to a plot to construct a story. The plot is designed to achieve a goal (intention, mission) of a story and gives a “direction or intent of meaning”. (Brooks, 1992) The plot infuses life into the story and adds verisimilitude to the story. The plot enables the author to “organize reality and make sense of the world unfolding within the narrative” (Holley and Colyar, 2012). Such elements of a story function as building blocks of constructing storytelling advertising. In this context, this study will create a founder’s story advertisement and a consumer’s story advertisement. There must be sequential events, a goal, and characters-and these factors are combined with the plot. Non-story advertising does not contain any of these narrative elements.
Recently, storytelling has been actively used not only in the field of literature and entertainment but also in education, art, policy, marketing and service. Although it is unknown what the real effect of brand storytelling is, researchers have investigated consumers’ behaviour under the assumption that consumers are reasonable (Fishbein and Ajzen, 1977). In general, a consumer’s decision is the result of the sum of all information gathered. The theory of reasoned action assumes consumers make logical and rational choices, so the theory predicts consumers’ future behaviour grounded in their previous attitude and intention. However, this theory is not always applied to actual purchasing behaviour because consumers often engage in impulsive or unreasonable consumption behaviour. They do not always follow the predicted pattern. When people make a purchase decision, they assume various situations of using the product or service and imagine the consequences of using it (Adaval and Wyer, 1998). Consumers are often captivated by emotion or specific memory. For example, even if they do not need a product, they may buy it when it reminds them of their childhood, parents, friends, pets and so on. Narratives (stories) can explain this unexpected and irrational consumer behaviour.
Storytelling contributes to connecting consumers’ inner experiences to their outside interpretation of the message. With this methodology, storytelling also can be “a tool of persuasion” in advertising, marketing and communication (Delgado-Ballester and Fernández-Sabiote, 2016). Stories about brand origins have a positive potential impact on customers (Schimitt et al., 2009), and storytelling makes a brand more convincing (Escalas, 2004). Padgett and Allen (1997) analyzed narrative processing during service encounters and found that it increased understanding. They concluded that narrative advertising might be effective for the service brand. Escalas (2006) found that narrative self-referencing gives a favourable evaluation of the advertised product. Previous studies show that storytelling affects consumers' evaluation of the brand image, product and service.
In the world of WOM (Word-Of-Mouth), the consumer’s storytelling shows a significant effect on other consumers. Delgadillo and Escalas (2004) examined consumers WOM communications and found that they often take a narrative form. WOM story may influence consumers, and this can affect their attitudes toward the brand. Moore (2011) also highlights the importance and the influence of WOM content.
Especially, the consumer-based story is considered more powerful than another subject story. Hong and Cho (2016) found that respondents who were exposed to a non-story ad showed the highest negative emotion, but the lowest negative emotion in consumer-based storytelling ad. Also, for the word of mouth intention, consumer story was more effective than non-storytelling or corporate storytelling ad. Urban et al. (2017) assert that a consumer-based story influences people to consider purchasing products. In the consumer-based story, consumers see themselves in the story and feel highly engaged with the brand. Consumer’s project their own story into the consumer-based story appeared in the advertisement and connected their story with the character inside of it. The consumer-based story gives familiarity, authenticity and a positive impression on people.
Involvement is an important consideration in any advertising study. The extent to which participants are interested in the product being advertised before being introduced to the message, as well as the extent to which such a predisposition affects the variables being tested, can add to the explanatory potential of a study like this. Specifically, a key question of this study is whether the presence or lack of narrative storytelling affects a person’s receptivity to a product or message, which led us to the Elaboration Likelihood Model (hereafter, ELM; Petty and Cacioppo, 1983). According to the ELM, a person more highly involved in the product should be more tuned in to the central message, whereas peripheral cues might more influence a lowerinvolvement participant. A complicating factor is a determination of precisely which aspects of the commercials constitute central and which constitute peripheral cues. Intuitively, a story with the main character and narrative structure might appear less likely to be perceived as purely informational (central) than an ad with an announcer simply listing information about a product. However, it is also possible that announce listing information about a product might seem more credible to low-involvement participants than a professional broadcaster playing the role of a customer or founder of a company. Therefore, we formulated a research question assessing whether involvement level affected the effectiveness of story-based advertising but did not form specific or directional hypotheses.
Hypothesis 1: Attitude toward the ad will differ from a Customer’s story ad, founder’s story ad, and the non-story ad.
Hypothesis 2: Attitude toward the story of the ad will differ from a Customer’s story ad, founder’s story ad, and the non-story ad.
Hypothesis 3: Favourability toward the ad will differ from a customer’s story ad, founder’s story ad, and the non-story ad.
Hypothesis 4: Word-Of-Mouth (WOM) intention will differ from a customer’s story ad, founder’s story ad, and the non-story ad.
Hypothesis 5: Information seeking intention from the corporate’s website will differ from a customer’s story ad, founder’s story ad, and the non-story ad.
Hypothesis 6: The product purchase intention will differ from the storyteller type of the ad.
Research Question: How does involvement level moderate the impact of attitude toward the ad and contents of the ad, favourability towards ads, WOM, information seeking intention and product purchase intention?
A convenience sample of 120 college students in the United States participated in this study, and 73 female (60.8%) and 47 (39.2%) male students participated. The participants were recruited at several undergraduate-level communication courses at a large south-eastern university in the US. Although extra credits were given to those who completed the questionnaire, one quality control question was inserted to screen out inattentive participants. The experiment was set up through an online survey platform. The research participants were randomly assigned to one of the three groups in the research design. The median age of the participants was 20, ranging from 18 to 23.
A pre-test was conducted with undergraduate students (n=70) and graduate students (n=11) to check manipulation of the stimuli and the questionnaire. The administration of the pretest took approximately 8 mins. Changes were then made in the questionnaire to correct inappropriate wording and ambiguous terms.
The manipulation check enabled us to make sure the study participants perceived who the storyteller of the study as we intended. The participants answered whether the ad they listened to delivered a founder’s story, a customer’s story or just information about the product. Pre-test results showed that approximately 72% of the pre-test participants correctly answered the question as manipulated (χ2=81.64, df=4, p<0.000).
This experimental study used a post-test-only control design with two treatments groups (the storyteller types of the ad: a founder’s story and a customer’s story) and one control group (a traditional informative ad). We chose to create an advertisement for a real-world company rather than a fictitious one to increase the external validity of the study. We chose Casper, a mattress company, based on its real campaign on podcasts associated with the National Public Radio website. Casper, a relatively new company, has rapidly grown up since its foundation in 2014. It has raised $70 million in venture capital and hit $100 million in cumulative sales in two years (Welch, 2016). Since Casper sold its mattresses only online until they earned investment from Target (Delrey, 2017), they rely on various types of advertising and Word-Of-Mouth to increase the visibility of the brand. We created audio advertisements, which could be utilized in a radio or podcast ad.
We based our choice of a product category for the study on an assessment of what our participating university students might realistically purchase in the not-too-distant future if they have not already. We also attempted to choose a product category requiring a purchasing decision high enough in involvement that participants might tune in mentally to get information but not so high as to make it unrealistic that they would be influenced by one playing of a radio ad. Therefore, the stimuli we created were three different versions of Casper advertising. The control group stimulus has only information of the Casper mattress such as technologies and materials. Information about the product was found on the company’s website (casper.com) and from the already existing advertising. The script of founder’s story advertising was written based upon the interview of Neil Parikh, one of the co-founders of Casper in Inc.com (Welch, 2016). The customer’s story was found on the customer review list on the company’s website. All stories were based on the facts, and the real names of the founder and customer were used in the ads.
A male professional narrator recorded the three stimuli and the same background music was used. We instructed the narrator to read all three scripts in as similar a tone and style as possible while maintaining believability that he was a professional announcer for the non-story version, a young man relatable to college students for the customer story and the founder/owner of a company for the owner story. Each ad has the word, “Casper” four times. The word count of each ad was 53 for the information-based ad, 72 for founder’s story and 78 for customer’s story. However, the running time of all three ads is all identical with the use of the same length of background music, so that we controlled a compounding factor that the different length of the ads would cause.
Attitude toward the ad was measured with six items, asking if the ad was: (a) trustworthy, (b) persuasive, (c) pleasant, (d) interesting, (e) positive and (f) upbeat7 (Cronbach’s α=0.740, M=3.70, SD=0.57, range=1-5).
Attitude toward the contents of the ad was measured with eight-item, asking if the contents of the ad are: (a) interesting, (b) creative, (c) impressive, (d) different, (e) touching, (f) meaningful, (g) believable and (h) convincing (Cronbach’s α=0.895, M=2.88, SD=0.77, range=1-5).
favourability toward the ad adopted five items from previous studies (Hubbard, 2010), and we changed the wording of two items to reflect the context of our study as follows: (a) The ad was good, (b) it was boring (reversed), (c) this is a kind of ad that I respond to, (d) it was unpleasant (reversed) and (e) I would like to hear more of this (Cronbach’s α=0.758, M=3.09, SD=0.67).
Word-of-Mouth intention used three items as follows: (a) I would like to talk to people about the product, (b) I would recommend the mattress to my friends and (c) I would like to tell the story of the ad to people around me (Cronbach’s α=0.785, M=2.39; SD=0.87, range=1-5).
Information seeking intention included five items such as: I would like to: (a) read about the company on the company’s website, (b) read about the product details on the company’s website, (c) read other customer reviews on the company’s website, (d) consult with the online customer representative on the company’s website and (e) search the brand on the internet (Cronbach’s α=0.905, M=2.83, SD=1.03, range=1-5).
Purchase intention adopted 5 statements such as: (a) I want to purchase the mattress in the ad, (b) I will consider purchasing the mattress in the ad sometime in next couple of years, (c) I prefer the mattress from the ad over other brands, (d) I will remember the mattress from the ad if I need to buy one in the future, and (e) I would recognize the brand if I saw an online ad for it (α=0.796, M=2.80, SD=0.80, range=1-5).
Involvement was measured with three items as follows: (a) I will need to buy a new mattress within the next one year, (b) I have been thinking recently of buying a new mattress, and (c) I am on the market for a new mattress (Cronbach’s α=0.875, M=1.81, SD=0.789, range=1-4). Participants were divided into low and high involvement group, based on the mean value.
We conducted one-way ANOVA to test hypotheses using SPSS 24. LSD and Bonferroni test were added as Post Hoc Tests to examine the differences among the three conditions. The ANOVA test for the Hypothesis 1 was not significant regarding the summated variables of the attitude toward the ad. However, the itemized ANOVA test showed that participants who listened to the customer’s story ad perceived the ad as upbeat (M=4.26, df=2, F=2.517, p=0.085<0.10), compared to the founder’s story exposure group (M=3.91) and non-story ad exposure group (M=3.79). Therefore, the results partially supported Hypothesis 1.
The ANOVA test for the Hypothesis 2 was also not significant regarding the aggregated variable of the attitude toward advertising content. However, the itemized ANOVA test on each of the eight items revealed that participants who listened to the customer’s story ad (M=3.00, df=2, F=5.132, p=0.007<0.01) were more likely to think the content of the ad was meaningful, compared to the non-story exposure group (M=2.29) and those who listened to the founder’s story (M=2.80).
Participants who were exposed to the founder’s story replied that the content of the ad was different (M=2.91, df=2, F=2.817, p=0.064<0.10), compared to the customer story listeners (M=2.3). The p-value was, however, slightly higher than 0.05. The results partially supported Hypothesis 2.
Hypothesis 3 examined the effect of storytelling ad formats on respondents’ favourability toward the ad. Neither the summated variable of favourability nor itemized variables showed a statistically significant difference among the ad formats. Therefore, Hypothesis 3 was not supported. We did not find significant ANOVA test results in Hypothesis 4 and Hypothesis 5, which tested the effect of storytelling ad types on Word-of-Mouth (WOM) intentions and information seeking intention, respectively.
The ANOVA test for the Hypothesis 6 did not show significant differences in the summated variable of purchase intention. However, research participants who were exposed to the founder’s story ad were more likely to respond that they would remember the product from the ad (M=3.03), compared to those who listened to the customer’s story ad (M=2.56, p=0.084<0.10). The p-value was slightly higher than 0.05, though. Therefore, Hypothesis 6 was partially supported.
Lastly, we examined how involvement played a role in the effect of storytelling ad formats on the summated outcome variables in our RQ with a two-way MANOVA tested. The results of the Between-Subjects Effects test showed that involvement level and storytelling ad types have some interaction effect on attitude toward the ad, WOM intention, and information seeking behaviours.
Participants with low involvement were more likely to think the advertising was positive in the founder’s story (M=4.33, df=2, F =4.21, p=0.017<0.05, η2=0.07) compared to the other two types of the ad (Mcustomer story ad=4.29, Mnon-story ad=4.21). When the involvement was high, participants were more likely to be favourable to the customer’s story (M=4.42) ads or the nonstory ad (M=3.43), compared to the founder’s story ad (M=3.43).
Seemingly contradictory results are shown in the WOM intention. When the involvement is low, participants’ WOM intention was higher on the purely informational advertising (M= 2.10, df=2, F=2.76, p=0.068<0.10, η2=0.046), compared to the other two types of the ads (Mcustomer’s story ad=1.72, Mfounder’s story=1.96). Those highly involved in the product purchase situation were more likely to tell the story of the founder’s story ad (M=2.86), rather than the customer’s story (M=2.37) or the one of information (M=2.23).
We also found an interaction effect between the advertising story types and information seeking behaviours (F=0.017, df=2, p=0.016<0.05, η2=0.984). In both high-involvement and low-involvement group, the informative ad is more likely to have the participants seek information about the brand (Mhigh-involvement/informative ad=3.22, Mlow-involvement/informative ad=2.78). The founder’s story (Mhigh-involvement/founder’s story=3.17, Mlow-involvement/founder’s story=2.68) is more effective than the customer’s story (Mhigh-involvement/customer story=2.94, Mlow-involvement/customer story=2.53). Overall, high involvement group showed the higher information seeking level.
The supplemental Tests of the Between-Subjects Effects demonstrated that involvement showed a main effect on outcomes such as WOM, eWOM and information-seeking variables. Participants who are considering mattress purchase in a couple of years (high-involvement group) were more likely to talk about the grand online and offline and more likely to search information about the brand and the company on the Internet.
The results of this study lend some support to the idea that ads using storytelling are more engaging than purely informational ads. This study also demonstrates that people are skeptical of direct advertising pitches delivered by those purely motivated to sell a product-people like announcers delivering non-story informational messages or the founders of companies. Advertisers have long employed testimonials by “regular folks” as an alternative to hard-sell ads, and these results suggest this is an effective approach, at least when targeting low-involvement audience members. However, it also appears that low-involvement participants responded more favourably to the purely informational ad regarding their desire to share information about the product by Word Of Mouth (WOM). While on the surface this appears contradictory to other results, there is a possible explanation based on the Elaboration Likelihood Model (Petty and Caccioppo, 1983). According to the ELM, low-involvement participants would be more tuned in to peripheral cues than high involvement participants. One such peripheral cue might be that a professional narrator sounds more credible reading facts about a product than performing the “role” of the founder or customer of a given company. High-involvement participants would focus on the information delivered-even in story ads-and be less attentive to the appropriateness of the voice or delivery style to a given message. One other note: The actual company used in this study employs informational non-story ads in its real campaigns (NPR Podcasts, n.d.), which was part of our motivation for choosing Casper. Without claiming to know Casper’s corporate strategy or the internal workings of its advertising methodologies, some of this study’s results suggest possible benefits to story-oriented ads, whereas the WOM result in low-involvement participants lends support to their current approach. Another interesting finding from this study is in the distinctions between which constructs were most impacted by which ad types. It was clear from the beginning that the effectiveness of ads should be operationalized in multiple ways. In a sense, this was an exploratory study, because we included multiple constructs in our instrument and, therefore, have a variety of results. For instance, the founder’s story was more effective when tested by participants’ recall of information, whereas the customer’s story engendered a more positive attitude about the ad. Similarly, the founder’s story appeared to catch people’s interest more effectively than the others, but the customer’s story triggered more desire to find out additional information about the product. Naturally, purchase intention would seem to be the ultimate manifestation of an ad’s effectiveness, and this study did not yield results on that construct, which is hardly surprising. However, the fact that a mattress is an expensive item with great impact on a person’s daily life makes it an inherently high-involvement purchase. Therefore, it would be unlikely for someone to fully form an intention to purchase such a product after merely hearing a 25-s audio commercial. This is why it is necessary for companies to evaluate the effectiveness of ad campaigns over longer periods of time and with actual sales numbers as part of the dataset.
Hubbard, G.T. (2010). Putting radio localism to the test: An experimental study of listener responses to locality of origination and ownership. Journal of Broadcasting & Electronic Media, 54(3), 407-424.