In the following chapter the methodology chosen to approach the research question will be develop, we will explain our method of Data collection, Data analysis and the type of interview we chose. Then we will go further into the data collection approach, how did we conduct the interview, why we chose a certain type of marketing campaign. Finally we will expand on the ethic and validity aspect of our research.
Chosen research method
This study is aimed at understanding the underlying mental processes defining a consumer’s reaction to an advertisement and the possible variations after exposure to negative word-of-mouth regarding this specific advertisement. This is in order to deepen certain aspects of the research question, rather than to find a solution. For this reason, we can define this research as an exploratory research. This research design allows adaptability, which is necessary when gathering data with individuals. However, due to the generic use of qualitative information, there is no possibility to describe the study as a correct sample representing the wider population (Dudovskiy, 2018). With most exploratory researches, it is not uncommon for the author to change his/her direction or method over time when gathering information (Saunders & al, 2012).
Data collection approach
To understand the choice of approach we decided to turn ourselves to, we first need to explain the various methods available. Three options are available: deductive, inductive and abductive. As the goal of our study is to understand consumers’ reactions towards advertisement, we use qualitative research to use the human factor as our main tool. Humans interactions can be adaptive and responsive, develop in the case of unusual answers and discuss to verify accuracy on both sides (Merriam, 2002). In many cases, it is easier to use inductive reasoning when doing qualitative research because of the discrepancy in human behavior, to which theories are not always applicable.
Induction or the inductive approach is usually defined as drafting a theory, a principle based on observations (Edson, Buckle Henning & Sankaran, 2017). Most qualitative researches are turning to inductive approach due to the fact that the researchers are closer to the data, and analysis is open to interpretation, it will allow to go more in-depth into the subject studied. On the other hand, a deductive approach drafts data from previous researches in order to verify them or develop them. They often use quantitative methods to analyze data with the help of surveys or experiments, such as in mathematics or health care studies (Edson & al, 2017). Finally, an abductive approach is a mix of both whereby comparing results and possible causes, with the theory in the middle. In this case, we gather data from the respondents’ perspectives, and analyze it in order to try to build concepts and hypotheses so according to the literature, seeking to an inductive approach is more fitting to our research (Merriam, 2002). However, we first started to analyze the results and then looked for theories to interpret them, which is close to a deductive approach, but since we went further than these theories thanks to our data, we remain closer to an inductive approach.
Data analysis approach
When trying to understand a phenomenon using qualitative data, a few possibilities are available, the most common being: grounded theory, phenomenology, case study or basic interpretive qualitative study. While we are using specific ads, we cannot describe them as case studies, because the principle of a case study is where the data is drawn from, and in this case, our priority is the respondents and not the ads themselves. For this paper, the basic interpretive study was used for the following reason: we are trying to identify and understand how the interviewees make sense of a situation, and how that meaning is an instrument. The pattern of a basic interpretive study goes as follow: discussing a phenomenon, the insights of the involved parties, then interviews to gather data which was analyzed to create patterns, sometimes using literature as references (Merriam, 2002). After collecting the data, we followed five steps to analyze the information gathered throughout the interviews. The process was conducted accordingly (Yin, 2016):
Compilation: this is arranging the data collected (interviews and concepts) in a certain logical order, such as forming a database.
Disassembly: this procedure can be assimilated to “coding”, assigning recurrent labels or themes to the content and was done to gather the main mood of interviews.
Reassembly: this step consists of reorganization the compiled data by categories, according to the codes previously defined, which was done through sorting the respondents in a table.
Interpretation: this phase is done to create a narrative that becomes the base of the analysis as data is deepened and deciphered. This is where we linked the codes to the concepts.
Conclusion: the final stage is to draw conclusions from the study, related to the four anterior phases.
On the second step, after transcribing the interviews and reading them thoroughly, we decided on six codes that were recurrent and relevant in the responses, then expanded the concepts to fit for the analysis: relatedness, taboo, disapproval, understanding, neutral, methodic. Using a color code, we highlighted fragments of the answers in order to emphasize the significant segments according to the previously defined themes. This allowed us to categorize the respondents’ interpretations and analyze the interrelations between the various own reactions of one respondent, those relations being cross-narrated in the empirical findings.
In order to assess the veracity of these theories and whether they apply to this study, we will use qualitative interviews, conducted individually to collect in-depth information on the subject. According to the results expected, three main different types of interviews can be used: structured, semi-structured, unstructured. A structured interview will follow the same specific questions, in the same order, leaving few to no room for flexibility. Oppositely, unstructured interviews have no guidelines and results in a more spontaneous exchange. For this research, the choice leaned towards semi-structured interviews with the help of a questioning guide: mostly the same questions were asked to the interlocutors, some of them being added or omitted according to the answers, in order to orientate the discussion (Ryan, Coughlan & Cronin, 2009). Interviews were done through self-selecting sampling, by communicating a request for cases on social media, then collecting the answers the public will voluntarily give (Saunders et al., 2016). Many factors can lead to disparities in the answers, apart from the characteristics that we discussed in the conceptual framework. The respondents may not interpret the questions the same way, give honest answers or the differences may come because of a variation in the relationship between the interviewers and the interviewees (Gomm, 2004).
Now that we have explained the methodology approach from the literature, we will further detail the fieldwork technique used for gathering information during the research.
Interviews were conducted face-to-face by the three researchers; the first taking the role of moderator to make sure the interview stayed on topic, the second to push the conversation further allowing interviewees to elaborate on their answers and the third in charge of taking notes to highlight the verbal and nonverbal cues from the interviewee. This format enabled free speech from the interviewee and an in-depth conversation between both parties. Prior the interview, interviewees were asked to sign a consent form (Appendix 1) that allowed researchers to record and use the information in the context of this study. The length of the interviewees varied from 10 to 23 minutes, all recorded on the same recording device. They were all conducted in English to ensure maximum homogeneity.
The interviews were articulated in three sections: introduction, display of the first campaign and online reaction, display of online reactions to the second campaign and then exhibit of the second campaign. The introduction section allowed the researchers to set interviewees into the topic’s mindset. Its purpose was to get the interviewees to talk about themselves and own experiences around the subject.
Prior the display of the first case, researchers presented the brand to the interviewee in order to assess awareness and current perception. Interviewees were then shown the first campaign and were asked to comment the campaign. Then researchers presented a sample of online reactions to the interviewee. Researchers assessed from the interviewee response his/her’s perception of both the brand and online reactions. The second case was conducted in the reverse order. After introducing the brand, researchers presented first online reaction without the interviewees being aware of the campaign itself. Interviewees were asked to comment these reactions hence give their assumptions on the actual campaign. The second campaign was displayed afterward, letting interviewees compare their expectations built on online reactions to the actual campaign. To sum up, the interview gave researchers knowledge on the following subjects, in this order:
Personal information (gender, nationality, social media use…)
Relationship with social media and advertisements
Knowledge of the brands (brand awareness, self-perception, consumption behavior…)
Exposition to advertisements and reactions online, first the ad than the reactions for the first case, than oppositely the reactions before the ad (pictures)
Opinion from the customer after seeing the ads and the reactions (change of mind, personal feelings…)
A more detailed interview guide (Appendix 2) was used during the research, in spite of the fact that it is not always in the same manner, depending on the respondent’s answers (i.e. pushing further some questions or lighten others).
In order to maintain the anonymity of the interviewees, we will designate them by a letter. Only their age, gender, nationality and program of studies are publicly displayed.
The two contents chosen are different in their form and subject. The first advertisement is publicity from the brand Protein World (Figure 2), a UK based company selling protein nutritional complements to promote a healthy, fit lifestyle. The ad was displayed in the London tube in 2015 as a visual poster splattered across the wall. The six reactions to this campaign we picked were chosen by notoriety or influence (that is, the ones that had the most reactions or relevance to the subject) (Figure 4). The second is more of a marketing campaign, a statement than an advertisement itself. It is part of a Budlight campaign (Figure 3) in 2015 based on the hashtag #UpForWhatever that displayed several different catchphrases on their bottles, such as Coca-Cola did with names or locations. In this case, the four reactions were chosen for their strong interpretation and negativity towards the campaign (Figure 5).
The decision for the Protein World campaign originates from the important vague of reactions at the time of the controversy, not only on social networks but in the traditional media as well. Here, the ad is shown first in light of the fact that the reactions were intimately related to the advertisement itself, only based on the image itself and its explicitness. All negative reactions were triggered by the exact message the company tried to convey. As women in advertisements are becoming a widely discussed and controversial subject, going back in time and analyzing the movement is interesting as 50% of the worldwide population is concerned by this “issue”.
As for the Budlight’s campaign, the decision was taken in light of the fact that oppositely from the first campaign, the reactions are taking wide and “violent” interpretations, which are not directly linked to the brand message but rather due to a critical context, triggering all kinds of taboo and room for personal interpretation. Here, the subject of alcohol, consent, and responsibility are highlighted by the detractors but the brand did announce their intent was not in the same optic.
The reason we chose to display Twitter responses over other social media’s reactions is due to high connectivity and liberty of speech that the platform offers. Is it not uncommon nowadays for brands to have a social network account, however they generally do not perform specific and adapted marketing strategies, rather applying the same on all platforms. The issue with this approach is that not all social networks are the same, Twitter being probably the riskiest of all. The blue bird company thrives on an original dialogic culture, established on high communication and information sharing, not only instantaneous but also advanced by a high reachability to a wide audience. Most users are known for their libertarian and anti-ideology attitude, their attraction for controversy and the exploitation of content for parody purpose. Therefore, a company will arise in conversations, whether directly involved or not, creating a lack of control over the information and content spread about the brand (Weller, 2014).
Ethics and validity
Ensuring quality and validity of the study is important, not only to ensure reliability in the results but also a correct and ethical conduct of research. There is hardly a way to define a “good” study but certain stances can be adopted to guaranty rigorous and trust-worthy conclusions. In order to address those concerns, we will develop about the validity (internal and external), reliability and transparency (Merriam, 2002).
Internal validity can be defined as the congruence between the research and reality: is the study true to reality, true to objectives? Reality is usually an idea of interpretation, relative to individuals; and in qualitative studies, it is usually the authors’ interpretation of the respondents’ own interpretation of the subject. One of the main tools to ensure internal validity is the use of triangulation: the correlation between multiple investigators, theories, sources and methods to confirm the outcome. In our case, the triangulation is made with the help of multiple researchers, sources, and theories (Merriam, 2002). The peer review strategy is also used all along the process of writing the paper, by the implication and correction of other students in seminars.
External validity, also called “generalizability”, is whether the study’s results are applicable to other situations or greater samples. Due to the small and random sampling associated with qualitative studies, the question of generalization is hardly worth considering, especially as the point is to get in-depth insight and not a general truth. One solution to still add external validity to a research is thanks to case-to-case transfer, in which the reader asks thyself if the situation would be applicable to thy own situation. To enhance this situation, a rich description and a great variety of the sample is primordial (Merriam, 2002).
Table of content
1.2 Problem Discussion
1.4 Research Question
2 Literature Review
2.1 Consumer’s perception through advertisement
2.2 Rise of Vox Populi: Word-of-mouth through Internet.
2.3 The spread of negative word-of-mouth and its responses
3 Conceptual Framework
3.1 Self Perception
3.2 Responses to advertisement stimulus
4.1 Chosen research method .
4.2 Data collection
4.3 Ethics and validity
5 Empirical Data
5.2 Protein World
6.1 Coding via concepts
6.3 Analysis summary
8.1 External determinants of the phenomenon
8.2 Outcomes on influence
8.3 Interconnection within phenomenon
8.4 Limitations and further research
GET THE COMPLETE PROJECT
Effective or offensive advertising? An exploratory study on negative Word-of-Mouth and consumers’ perception