Anne Gebhardt & Jana Weerd
The Impact of User Comments
on Consumer Brand Attitude
An Empirical Study of Different Facebook
Preface
Statement of Originality
This document is written by Anne Gebhardt and Jana Weerd who declare to take full responsibility for the contents of this document.
We hereby confirm that we have written the accompanying thesis by
ourselves, without contributions from any sources other than those cited in the text and references.
This applies also to all graphics, tables, figures and images included in the thesis.
Authors’ Contribution
Acknowledgement
A major project like this is never the work of one person alone. The contributions of various people have made this project possible and we would like to extend our appreciation especially to the following:
First and foremost, we would like to express our deep gratitude to our thesis supervisor Bo Rundh. Not only has he taught us many important lessons throughout the courses of our two year program but also served as our patient and diligent supervisor for this final and most important project of our academic career so far. Thank you for your guidance and constructive recommendations during the planning and development of this dissertation.
Secondly, we would like to thank our fellow students for their feedback and encouragement throughout the thesis seminars and the writing process. Their input has been very beneficial for the continuous improvement of our thesis. Moreover we would like to thank all the participants, who took part in our survey. Their willingness to give their time and opinions so generously has been crucial for this project and is highly appreciated.
Finally, we would like to thank our families and friends. Without their continuous and loving support, pursuing a degree abroad would not have been possible. Thank you for your support and encouragement, for the love and the wisdom.
Abstract
Social Media has become an increasingly important part of people’s everyday life. With the rise of social networks, brands have discovered ways to use this phenomenon to their advantage. Providing a novel two-way communication between a brand and consumers, Social Media networks are a tremendously powerful tool to build strong and sustainable consumer-brand relationships. With Firm Generated Content (FGC) and User Generated Content (UGC) alike the interaction between these two entities is stronger than ever. While this is mostly perceived as very positive and beneficial for consumers brand attitudes and the overall brand equity, it also comes with possible pitfalls. While some firms fail to participate on Social Media in an attractive way, others are facing the increasing problem of negative user comments on their page. This thesis aims to critically evaluate the effects negative user comments could have on consumers’ brand attitudes, using the examples of H&M and Facebook. Throughout the literature review various topics have been assessed. Moreover a web-survey was conducted to gain direct consumer insights on the matter. The findings were statistically analyzed using the software SPSS. After the discussion of the findings, concluding recommendations can be found that shall benefit fellow marketers and brands when implementing Social Media into their communication strategy.
Keywords: Social Media, Brands, Brand Attitude, Brand Equity, Consumer-Brand Relationship, FGC, UGC, Digital Word-of-Mouth, Negative
Table of Content
1. Introduction ... 3 1.1. Problem Statement ... 4 1.2. Purpose ... 4 1.3. Research Question ... 4 1.4. Structure ... 4 2. Literature Review ... 6 2.1. Marketing ... 6 2.1.1. Marketing Communication ... 7 2.2. Social Media ... 82.2.1. Social Media Marketing ... 8
2.2.2. User Impact on Social Media ... 9
2.3. Brand ... 10
2.3.1. Brand Attitude ... 10
2.3.2. Brand Equity ... 11
2.3.3. Negative Brand Events ... 12
2.4. Consumer-Brand Relationship ... 12
2.5. Brands on Social Media ... 13
2.5.1. Firm Generated Content ... 14
2.5.2. User Generated Content and Digital Word of Mouth... 16
2.5.3. Negative Effects of Social Media ... 17
2.6. Overview of the Main Concepts ... 18
2.7. Hypotheses ... 20
3. Background of the Study ... 22
4. Research Methodology ... 25
4.1. Research Approach ... 25
4.2. Research Design, Strategy and Measurement ... 25
4.3. Data Collection Method and Sampling ... 26
4.4. Quality of the Study ... 30
5. Data Analysis ... 31
5.1. Survey Results ... 31
5.2. Preparation for Analysis ... 35
5.2.1. Variables ... 36
5.2.2. Statistical Tests ... 37
6. Discussion ... 44
6.1. Managerial Implications ... 45
6.2. Limitations and Bias ... 46
Table of Figures
Table 1 Demographic results……….30
Graph 1 Results v4 – Facebook usage………...31
Graph 2 Results v5 – Purchase frequency………..32
Graph 3 Results v1 – Opinion before……….33
Graph 4 Results v2 – Opinion after………34
Table 2 Coding of variable 1………..36
Table 3 Coding of variable 2………..36
Table 4 T-test Opinion before – Opinion after………...37
Table 5 T-test v1 – v2: p-value………..38
Table 6 ANOVA test Difference opinion – Facebook usage……….38
Table 7 ANOVA test v3 – v4: f-value/p-value………..38
Table 8 ANOVA test Difference opinion – Purchase frequency………...39
Table 9 ANOVA test v3 – v5: f-value/p-value………..39
Table 10 ANOVA test Difference opinion – Interest in H&M………..…40
Table 11 ANOVA test v3 – v6: f-value/p-value………....40
Table 12 ANOVA test Opinion before – Purchase frequency………...40
Table 13 ANOVA test v1 – v5: f-value/p-value………41
Table 14 ANOVA test Opinion after – Purchase frequency………..41
1. Introduction
With the rise of the digital era, consumer expectations on how a brand markets and communicates have changed drastically. Prior to - nowadays popular - social networks like Facebook, brands used channels such as TV or radio to inform their customers (Hoffman & Novak 1996). However, this way of communicating has now evolved into a two-way interaction with the consumer, facilitating them to have control over timing, channels and content (Mangold & Faulds 2009). While this freedom and the viral nature of social networks offer great opportunities, such as digital word-of-mouth (WOM), precicely this digital WOM can also be potential threat to a brand’s image and its brand equity (Killian & McManus 2015).
Countless research has been done on many topics surrounding Social Media, including how consumers engage with brand and even the impact of negative comments on a brand (e.g. Sexton 2015, Kumar et al. 2016). However the impact of negative user comments on other users, that are reading these comments and the potential impact on consumers’ brand attitude have not been considered yet. Even though this is an important part in understanding the dynamics of Social Media. Hence, this thesis is motivated by the detected research gap in the existing literature. Moreover a personal interest in fashion as well as Social Media, furthered the interest and motivation to approach this specific topic and led to the choice of brands, namely Facebook and H&M. Being Social Media enthusiasts themselves has benefitted the authors immensely during the course of this project, especially when it came to the data collection. This first chapter, which serves as an introduction to the topic, will provide the reader with the problem statement and also cover the research purpose, research questions and objectives set for this project. Furthermore this primary chapter provides a brief structural overview of the entire thesis, which shall guide the readers in the proceeding chapters.
1.1. Problem Statement
The last decades were shaped by the development of technology and especially the development of the internet. With the development of online media Marketing Communication is changing to a more customer-centric and customer-driven communication (Dahlen et al. 2010). With Social Media networks customers can communicate with each other and with businesses (Evans 2010). They can share their own experiences, problems, ideas and opinions with the whole world (Evans 2010). This brings different possibilities for companies to enhance customer communication and their experiences, but also threats (Evans 2010). Since information is shared with everyone online, negative comments about a brand spread fast and companies cannot control what is written about their brand (Evans 2010). Therefore it is crucial that companies understand how customers react to the information a brand publishes and especially to information other customers or competitors publish (Kumar et al. 2016).
1.2. Purpose
The purpose of this study is to examine if user comments on a brand’s Social Media page have an impact on customers’ brand attitude.
1.3. Research Question
The main research questions that can be derived from this purpose are:
Do negative comments have a negative effect on customers’ brand attitude?
Does the impact of negative comments vary among different customer/user groups?
1.4. Structure
2. Literature Review
This chapter will assess the relevant academic literature and shall serve as a framework for this thesis. Several topics and aspects will be introduced to the reader to ensure the overall comprehension of the topic. To begin with, the reader will be introduced to the overall concept of marketing and its subcategory communication. Proceeding to a specific type of communication, namely communication via Social Media, which will be one of the most relevant aspects for this thesis. Subsequently the terms Social Media marketing and the changed role of customers on Social Media will be explained. Followed by an elaboration on the term brand, brand attitude and brand equity. After that the effects of negative brand events are explained. The next part will focus on customer-brand relationships, followed by connecting the two main points, brands and Social Media in the paragraph ‘brands on Social Media’. Going further into detail on brands Social Media communication, the authors will take a look at firm generated content and user generated content and lastly arrive at the central point of this thesis, which is possible negative effects of Social Media. This is followed by an introduction to this thesis’ two main actors Facebook as a representative of a Social Media network and H&M as a popular brand on the named Social Media network. The chapter will be closed by the hypotheses that can be derived from the literature review.
2.1. Marketing
Marketing is defined by the American Marketing Association (2013) as ”the activity, set of institutions, and processes for creating, communicating, delivering, and exchanging offerings that have value for customers, clients, partners, and society at large.“
The purpose of marketing is to identify and meeting the needs of consumers and to create a demand for the company’s products (Kottler & Keller 2012). But marketing is much more than just selling a product, it builds a strong brand and a loyal customer base (Kottler & Keller 2012). It is an important part in the profit making of a company and contributes to a firm’s value (Kottler & Keller 2012).
changes are making marketing a complex job and companies need to keep up with the changes in customer needs and expectations (O’Driscoll & Murray 1998).
Marketing consists of many different parts, one of them is marketing communication.
2.1.1.Marketing Communication
The purpose of Marketing Communication is not primarily to sell products but rather to inform, persuade and remind consumers about the brand a company sells. The communication can be made directly or indirectly. Marketing Communication can be seen as the voice of a company and its brand that is used to establish a dialogue with consumers and build strong relationships between a brand and a customer. Through this way of communication customers can learn more about a product, e.g. how to use it and how it is produced. Marketing Communication links a brand to other people, places, experiences and feelings. By establishing the brand in the memory of consumers and creating a brand image, it contributes to the brand equity, which will be described later. Elements of the Marketing Communication Mix are Advertising, Sales Promotion, Events and publicity, Direct Marketing, Interactive Marketing, Word-of-mouth Marketing, Personal Selling. (Kotler & Keller 2012)
With the development in technology, the nature of Marketing Communication is changing (Dahlen et al. 2010). New media platforms are available and allow marketers create a more targeted communication (O’Driscoll & Murray 1998). With the new forms of advertising a more interactive communication is available, for example through mobile, wireless, internet or word-of-mouth (Kerr & Schultz 2010). Especially with the development of online media the Marketing Communication became more centric and customer-driven (Dahlen et al. 2010). Consumers are getting higher expectation on companies (O’Driscoll & Murray 1998).
2.2. Social Media
Since over 20 years Social Networking Sites (SNS) exist to connect people and let them exchange messages. Over the years more and more of these sites developed, with different purposes and enhanced possibilities. Some of these pages let you connect to old classmates, share music or videos with people or extend your business network by connecting with other professionals. The development of SNS underlines the shift the online community made. Instead of a focus around the same interest, communities now are people-centered and focus on the individual. Even though most SNS have different purposes and different content they all have one thing in common: they connect people and let them share media content and opinions. Communication and connections are the basis of SNS. All the different SNS are collected under the term Social Media. (Boyd & Ellison 2008)
2.2.1.Social Media Marketing
The development of Social Media does not only enhance the consumer experience but can also be helpful for businesses (Evans 2010). Customers can share their problems, ideas, and even potential competitive advantages online (Evans 2010). Companies need to understand the new technologies and engage with Social Media to survive in this age of empowered customers (Constantinides 2014). Social Media is strategically necessary and is becoming an important Marketing tool (Constantinides 2014). When it is implemented and used in the right way it creates new forms of customer values (Constantinides 2014).
Social Media is a very strong tool to build and maintain a relationship with the customer (Killian & McManus 2015). When a consumer goes to a brand’s Social Media page he or she is already interested in the brand and what it offers (Killian & McManus 2015). Consumers are starting a conversation when reaching out for help to a brand or company (Killian & McManus 2015). This can be the basis of a deep, long-term engagement (Killian & McManus 2015). Therefore Social Media pages should not necessarily used to sell products or services but rather to demonstrate a product to customers and include the customers in campaigns (Killian & McManus 2015). It is a very private platform used to create a friendly relationship with consumers; sales pitches are unwanted and can drive customers away (Killian & McManus 2015).
Social Media is holistic and wraps around the entire business, not only marketing and customer support can profit from it (Evans 2010). The customer is moving towards the company (Evans 2010). The role of the customer changed from being only a viewer to being a participant (Evans 2010).
It is important that Social Media is seen as an integrated part of the marketing strategy and is not seen as an isolated process (Constantinides 2014). It is the final step in the strategic effort to improve a product or service and meet the customer's needs (Constantinides 2014). There are different ways to use Social Media as a marketing tool, but the most important step is to create a Social Media presence and combine the possibilities (Constantinides 2014). Blogs can be used by employees, to share content about a company, as well as by so called influencers (opinion leaders, mostly from influential blogs) that serve as brand advocates and can share positive stories or reviews of a product or service (Constantinides 2014). A very important way of using Social Media is to include the customer, e.g. by offering customizable products or encourage them to engage and produce content about a brand (Constantinides 2014). Social Media campaigns can be integrated in traditional, offline campaigns, for example by encouraging the customers to share photos of a sponsored event (Killian & McManus 2015). That way customers engage with a brand and create online content that get shared with many different people (Killian & McManus 2015).
2.2.2. User Impact on Social Media
customer has changed (Evans 2010). They can share their own experiences with a product or a service with the world and thus leveraging the experience of others (Evans 2010). With the development of Social Media a new form of word-of-mouth (WOM) marketing was developed: digital word-of-mouth (Evans 2010). The difference to the traditional WOM is that customers can tell thousands of people about their experience in only one click, by sharing it on Social Media (Evans 2010). WOM has always been a driver of customer behavior and with digital media businesses anyone can read what other customers wrote (Fulgoni & Lipsman 2015). Social Media is linking people around the world without geographical restrictions (Fulgoni & Lipsman 2015). The high reach and attention digital WOM is getting it is made potentially more influential than traditional WOM (Fulgoni & Lipsman 2015). The digital WOM is integrated in the purchasing process of consumers (Evans 2010). They read about experiences of others, study reviews of products or services they plan on buying and consider ratings before making their final buying decision (Evans 2010). Other customers have nowadays more influence on the purchase intention of a single customer (Evans 2010).
2.3. Brand
Kottler and Keller (2016) define a brand as a name, term, symbol or design, or combination of them intended to identify the goods or services of one seller or a group of sellers and to differentiate from those of competitors. Companies are offering value propositions, sets of benefits, to consumers to satisfy their needs. By building and establishing a brand these value proposition will come from a known source. The aim of a brand is to create positive, favorable and unique associations with the brand name in people's mind and construct a brand image. It is a tool to create more trust in a company and its products. (Kottler & Keller 2016)
2.3.1.Brand Attitude
“Attitudes are mental states used by individuals to structure the way they perceive their environment and guide the way they respond to it. There is general acceptance that there are three related components that form an attitude: a cognitive or knowledge component, a liking or affective component, and an intentions or actions component. Each component provides a different insight into a person’s attitude.” Aaker & Day
Cognitive/knowledge component: a person’s information about an object, which
includes awareness about existence of object, beliefs about characteristics/attributes of object, judgement about relative importance of these attributes.
Affective/liking component: person’s overall feeling towards an object on a
like/dislike or favorable/unfavorable scale, or preferences.
Intention/action component: person’s expectation of future behavior towards an
object.
Aaker & Day (1990 p. 270-272)
2.3.2.Brand Equity
2.3.3.Negative Brand Events
Negative experiences can ruin a perfectly good day. Everyone has experienced this “phenomenon” that is called negativity effect. Negative information are considered as more important and more informative than equally positive information when forming an evaluate judgement and thus are overshadowing positive information (Ahluwalia 2002). In the study from 2002 Ahluwalia investigates the influence of the negativity effect on consumers. One result was that consumers who like a specific brand perceive positive and negative information about this brand as equally helpful, which means compared to consumers that are not emotionally involved with the brand, the negativity effect was weaker. However this only means that the positive information were influenced by the positive attitude of the consumer towards the brand, and the negative information were not impacted. Another important conclusion of the study is that consumers pay more attention to information about a familiar brand than about an unfamiliar one. (Ahluwalia 2002)
A more in-depth study about the influence of negative events on consumers’ brand attitude was made in 2015 by Yalin Li. Enterprise negative events are events occurring in enterprises that are linked to poor quality, bad service, inefficient management and cause customers to experience negative thoughts about this enterprise (Wang & Yang 2011, quoted by Li 2015). These events always result in a harm of the company (Li 2015). Especially with the open communication via the internet and the growing media industry it is a big threat to businesses (Li 2015). The study was conducted with different customer groups, on with people highly engaged with a brand and one with low engagement (Li 2015). The high engagement group felt more betrayed by negative events than the low engagement group (Li 2015). Moreover it showed that the more severe a brand event the higher the negative effect on brand attitude (Li 2015). The study makes clear that especially loyal customers are vulnerable to negative exposure and companies need to take actions to counteract negative events (Li 2015).
2.4. Consumer-Brand Relationship
category, by linking the brand name or logo to a specific product. They are building associations with the brand and get more aware of it, which means it is important for a company to create positive associations for consumers. The Experiencing stage occurs when consumers come in contact with the brands, e.g. by seeing brand logos, advertising or on Social Media. The consumer gets affected by the brands and creates positive or negative feelings towards it. In this stage the consumer is not just a recipient of information anymore, but starts to socially engage, e.g. via Social Media. In the Integrating stage all information are put together and weighted to combine an overall brand concept. Consumers create a relationship, similar to a friendship, with a brand that involves trust. A next step would be the Signifying stage where the brand becomes an internal cue. Consumers identify with the brand and use it to express themselves. In the Connecting stage the attitude towards a brand is formed. The brand attitude shows the degree of favor or liking of a brand. The bond between brand and consumers grows stronger and consumers get emotionally attached. With a strong relationship with the brand, groups of like-minded consumers form a brand community. The members of such communities feel emotionally involved and connected with the group. Social networking sites and especially brand pages on Facebook can be seen as brand communities. (Schmitt 2012) In fact Social Media interactions have a positive influence on the quality of brand-relationships (Hudson et al. 2015). Consumers get more attached to the brand on Social Media, and the interactive communication gives more satisfaction to consumers. Companies need to use these interactions to understand and respond better to consumers’ needs and thus create higher brand loyalty and profits. Consumers that are more active in interacting with brands on Social Media have a stronger brand-relationship. (Hudson et al. 2015)
2.5. Brands on Social Media
brand and consumers (Hudson et al. 2014, Saboo et al. 2014). Social Media sites can enhance the reputation and the image of a well-established brand (Hudson et al. 2014).
Besides that Kumar et al. (2016) argue that the virtual presence of other brand sympathizers on the Social Media page and the interaction between fans can reinforce a favorable brand attitude among customers. Moreover, Gensler et al. (2013) agrees by saying that unlike other types of media, a firm’s communication through Social Media networks could be part of their equity-building efforts, aiming to manage brands and nurture customer relationships.
It is important to align the communication through Social Media with all other Marketing Communication and with the brand personality, to stay authentic and build trust (Killian & McManus 2015). However the brand messages send should still be customized for the individual audience of the different channels (Killian & McManus 2015).
There are many different statistics of Facebook usage both concerning customers and firms. One example is the 2014 Social Media Industry Report (Stelzner 2014) that stated that 97% of firms are using some form of Social Media (VanMeter et al. 2015).
2.5.1.Firm Generated Content
According to Kumar et al. (2016) the concept of social customer relationship management (social CRM) is becoming increasingly important. Kumar et al. (2016) also add that for social CRM to be effective, firms must understand how customers react to the firm generated content (FGC) and whether particular groups of customers might benefit more from the firm’s social engagement. Characteristics that could influence a customer’s perceptiveness to the FGC are customer-firm relationship, customer’s technology savviness and customer’s propensity to use social networking sites regularly, because these could indicate a person’s motivation and ability to engage with a firm online (Kumar et al. 2016).
Kumar et al. (2016) define these three characteristic as follows: Length of relationship, tech savviness and Social Media proneness.
Length of relationship: Based on academic literature in the field of branding it has
marketing messages of well-known brands show more frequent as well as more positive responses among consumers. Overall it is said that a long term relationship with a brand also results in greater satisfaction among the customers (Palmatier et al. 2006). In return, more satisfied customers feel a higher degree of loyalty towards a brand (Ranaweera and Prabhu 2003), which makes them more likely to be positively affected by FGC (Kumar et al. 2016, VanMeter et al. 2015).
Tech savviness: Nowadays firms are always finding new technologies to make
their business more appealing to tech-savvy customers. Hence these customers are expected to be more accessible, more receptive for engagement in form of FGC and ultimately also more interactive with brands on their Social Media platforms. Therefore they are able to extract greater benefits from the Social Media engagement, which leads them to respond better to FGC. (Kumar et al. 2016)
Social network proneness: With the rise of Social Media and its ability to digitally
connect consumers, more time is spent on online interactions with other customers in order to exchange information, learn about products and evaluation of brands and products (Kumar et al. 2016, Hudson et al. 2015, Saboo et al. 2014). Kumar et al. (2016, p.11) call these consumers “social network-prone”, while VanMeter et al. (2015, p. 72) refer to them as “individuals with a strong attachment to social media”. According to Naylor et al. (2012) the virtual presence of brand fans on Social Media platforms can positively affect a potential customer's purchasing behavior. Hudson et al. (2014) also argue that the likelihood to purchase and recommend a brand or its products increases after the consumer follows that brand on Social Media. Thus Kumar et al. (2016) state that customers who are highly prone to using Social Media will value the ability to connect with like-minded customers and share their experiences to a greater extend and will ultimately also be more receptive to Social Media engagement with a brand. Lastly, in accordance to what has been said before, Schulze et al. (2014) explain regular Social Media users will consider the opinions of other people with similar views to a greater extent and in turn also show a better response to FGC and also user generated content (UGC) in the Social Media network.
2.5.2.User Generated Content and Digital Word of Mouth
2.5.3.Negative Effects of Social Media
One of the side effects of Social Media is that the customer starts to compare companies and their activities online (Evans 2010). That way customers build expectations based on, e.g. how other businesses handle problems (Evans 2010). News travel fast in the social web and a fast, but consistent reaction of the company is important to keep the customers’ approval (Evans 2010). Especially customer complaints have to be handled timely and properly (Evans 2010). Customer complaints can give an insight of what bothers the customer and it is important to understand what lies behind the complaint (Evans 2010). Thus Social Media really is a two-sided coin - the same influencer that leads many customers to a site, can also weaken your brand equity and profitability when expressing dissatisfaction (VanMeter et al. 2015).
Social Media has a big impact on a customer’s opinion about a brand (Evans 2010). People are gathering online and sharing information, knowledge or their opinions about a product, brand or service. This communication can be considered more sincere than the marketing communication of a company and shows more truly what a brand is. Companies cannot control the brand image that results from the online communication, but they can still influence it. This is the reason a strong Social Media strategy is crucial to stay competitive. A popular content published by a brand might be shared by people among their friends and increase their brand loyalty. (Evans 2010)
response can increase the attrition rate of customers by 15 % (Sexton 2015). One of the problems of online networks is that bad news spread faster than good news, because they get shared more often (Sexton 2015) and as proved with the negativity effect: negative information is valued more than positive. Another problem is that the open forum of the internet and the anonymity can lead to cyberbullying of businesses by customers (Sexton 2015). A single opinion can have a lot of power online, or as Corstjens & Umblijs (2012, p. 447) put it: “An unhappy customer can topple a government”. Companies need to monitor comments in real time and ensure a quick and effective handling of complaints, but most importantly move them out of the spotlight (Sexton 2015). A good strategy is to bring the customer on your side and make him part of your marketing team (Sexton 2015).
2.6. Overview of the Main Concepts
Concept Definition/Purpose References Marketing - Identify and meet needs of
consumers - Generate demand
- Build a strong brand and loyal customer base
Kotler & Keller 2012
Marketing Communication
- Tool to build a strong brand - Establish brand in consumers
memories and create a brand image - Contributes to brand equity - Highly influenced by development
of Social Media
Kotler & Keller 2012 Dahlen et al. 2010
Social Media marketing
- Customer-centric communication - Customer is moving towards the
brand
- Companies get an insight into consumers opinions
- Creates connection between customer and brand
- Used to build a strong consumer-brand relationship
Evans 2010 Hudson et al. 2014
Digital WOM - Customers share opinions about products online
- Influence the purchase decision of other customers
- Influence the brand attitude of other customers
Fulgoni & Lipsman 2015
Brand - name, term, symbol or design, or combination of them intended to identify the goods or services of one seller or a group of sellers and to differentiate from those of competitors
- create consumer trust into products/services
Kotler & Keller 2016
Brand Attitude - information about a brand, which includes awareness about existence of brand, beliefs about
characteristics/attributes of brand, judgement about relative importance of these attributes.
- customers feeling towards a brand on a like/dislike or
favorable/unfavorable scale, or preferences
Aaker & Day 1990
Brand Equity - “Marketing effects uniquely attributable to the brands” - creates value for the customers as
well as for the company that sells the product or service
- consists of a set of brand assets that are linked to the brand (brand loyalty, name awareness, perceived quality and brand associations)
Keller 1993 Aaker 1991
Consumer-brand relationship
- important part in creating brand equity
- consumers come in contact with brand on Social Media build associations
- influences brand attitude - consumers get attached to brand,
esp. when connecting on Social Media
2.7. Hypotheses
The literature review showed the opportunities and threats of Social Media for companies. As a result it became an integral part of a company’s communication strategy.
Homburg et al. (2015) highlights that many scholars such as Chen and Xie (2008), Godes and Mayzlin (2009) as well as Rishika et al. (2013) have stressed the important role of consumer characteristics and the resulting differing reactions to a brand’s Social Media activities. Mangold, Miller and Brockway (1999) found that consumers with a high feeling of commitment towards a brand are more likely to spread a positive message but are also likely to be vehement if disappointed. The study of Li (2015) showed similar results to the extent that customers, who are highly engaged with a brand are more affected by negative brand events than customers with low engagement, because they felt more betrayed. From this study the first hypothesis will be derived:
H1 Customers with a high engagement in H&M are stronger negatively influenced in their opinion about H&M by negative user comments than customers with a low engagement.
H2 Customers with a low engagement in H&M are less negatively influenced in their opinion about H&M by negative user comments than customers with a high engagement.
H3 Frequent users of Facebook are stronger negatively influenced in their opinion about H&M by negative user comments than infrequent users.
3. Background of the Study
Founded in 2004 in the United States by Mark Zuckerberg and co-founders Dustin Moskovitz, Chris Hughes and Eduardo Saverin Facebook has become one of the most important Social Media platforms to date. What started as a college network platform quickly developed into one of the largest networks worldwide, reaching 1 million users within the first year, 12 million people in 2006, and already 360 million people in 2009. With its mission “to make the world more open and connected” (Facebook 2016) and with - on average - over 1 billion active users daily (December 2015), Facebook has become an important tool in the daily life of people but also an increasingly powerful marketing tool.
On their own Facebook page they state that it is their mission is “to give people the power to share and make the world more open and connected. People use Facebook to stay connected with friends and family, to discover what’s going on in the world, and to share and express what matters to them.” (Facebook 2016). This is precisely the reason why Facebook has become an important marketing tool. It enables communication and interaction between a brand and its customers.
Facebook offers multiple ways for companies to connect with customers and build their brands. Companies can set up a page on Facebook to present their brand or business, by sharing information (e.g. opening hours, store locations, etc.) and campaigns with an audience (Facebook 2016). These pages offer the possibility for customers to connect with a brand and communicate directly with the company (Facebook 2016). By sharing photos, stories or campaigns and encourage customers to share their own experiences a brand will get closer to the customers and connects with them. The pages also include a messaging service, where customers can directly contact the company, or more specifically the administrators of the page.
H&M
“From a single women’s wear shop in Västerås, Sweden, to six different brands and 3,900 stores all around the world” (H&M, 2016). The evolution of the H&M Group is a real success story within the global fashion industry. With six independent brands: H&M, Monki, Weekday, Cheap Monday and & other Stories, the H&M Group is one of Sweden’s posterchildren for international success. On the Forbes list of the most powerful brands globally H&M made #33 and is the 2nd most powerful retail brand, behind Walmart (Forbes 2015). It is a globally known brand that is active in 61 markets worldwide with 210 billion SEK sales in 2015 (H&M Full-year report 2015).
Within the scope of this thesis their main and founding brand H&M is of particular interest, as it is highly popular on Social Media – accumulating 26.721.431 follower (April 2016) on Facebook alone.
H&M (2016) describes their business concept on their webpage as “Fashion and quality at the best price in a sustainable way”. The concept of H&M is to offer “inspiring fashion with unbeatable value for money”. A wide range of collections are offered to suit women, men, teenager, and children (H&M Business Concept 2016). The brand is present in 61 markets worldwide and offers online shopping in 30 countries (H&M, 2016). Not only is it the biggest brand within the H&M Group but also the most diverse one, because the width and variety of their collections will cater to the needs of nearly everyone. But this does not mean the brand has finished its expansion. In 2016 roughly 400 new stores are being planned, mostly in existing markets, but H&M also strives to explore three new markets – New Zealand, Cyprus and Puerto Rico. With ongoing success and solid and continuous growth H&M has acquired a strong financial position that enables the brand to seize new business opportunities, as the recently launched H&M Beauty division or the investments into distance selling, especially shopping via smartphones
guidelines for the company and its suppliers concerning business ethics, health and work safety, sustainability, fair wage and high quality services.
Even though the described topics are all important in the H&M business practice many consumers criticize the company’s handling of these topics (The H&M Way 2016). The Facebook page of the brand contains many negative comments concerning bad product quality, unfriendly customer service and especially the hazardous working conditions in the factories that supply H&M. H&M has been criticized multiple times in public. For example in 2010 when 21 workers died in a fire in one of the Bangladesh factories that supplies H&M (Hickmann 2010). Other negative exposure on Social Media and newspapers include accusations of using too thin models, which promotes a negative body image, using too tanned models, which lead to many comments about cancer risk of tanning (The Local 2012), and a scarf design that resembled a sacred Jewish prayer shawl (Lieber 2016).
H&M’s Social Media strategy is global, even though they use specific pages for different countries in the native language. H&M uses Facebook to present their products in an appealing way, share campaigns with hashtags to encourage participation (e.g. #HMLovesCoachella) and to promote their blog (H&M ONLINE 2016). H&M is one of the leading fashion brands on Social Media. The standard Facebook page (https://www.facebook.com/hm/) will serve as a reference throughout the study.
4. Research Methodology
The following chapter of the thesis outlines the research methods applied to meet the formulated research aims and objectives stated in first chapter. The systematic collection and interpretation of data will serve the purpose of arriving at a result, which provides novel insights for fellow academic researchers. The authors of this thesis justify this study with a detected research gap in the existing literature regarding the effects of negative user comments on customers brand attitude. Although much research has been carried out on the basis of Social Media marketing and its effectiveness, impacts and also pitfalls, none of them focus in particular on negative user comments and their effect on customers brand attitude.
4.1. Research Approach
For every research project one of the two existing research approaches is applicable. The inductive approach concerns itself with developing a theory resulting from the collection of data, whereas the deductive approach identifies a hypothesis based on already existing knowledge of a field, which is thereupon empirically examined. The findings of the empirical study are then compared to the theory and initial hypothesis, which subsequently will lead to a rejection or acceptance of the presumptions (Bryman & Bell 2011). The authors of this thesis use a deductive research approach.
4.2. Research Design, Strategy and Measurement
In order for a cohesive research design, one must first of all distinguish between a single quantitative, qualitative or multiple methods research design. Secondly the nature of the research design must be clarified. Generally speaking, there are three different types of research: exploratory, descriptive and explanatory (Saunders et al 2012).
that one must understand a phenomenon before one can try to explain it. Those particular kinds of research are called descripto-explanatory studies (Saunders et al. 2012).
The research strategy describes the detailed approach of data collection and analysis. It consists of different research techniques that can be facilitated in order to answer the research question and achieve the predefined objectives. These techniques can for example be experiments, surveys and case studies (Saunders et al. 2012).
The authors of the thesis have opted for a single quantitative research design with an exploratory nature. To provide a holistic view on this particular topic of interest, the authors have reviewed and evaluated relevant, existing literature. Concerning the research strategy, quantitative data was collected through an online self-completion questionnaire, which will help to answer the research question and verify or reject the hypotheses.
Lastly, the time horizon of a research project has to be considered. A study can either be longitudinal, meaning lasting over an extended period of time or cross-sectional, which is performed once and delivers a sort of snapshot of time (Saunders et al. 2012). Thus this project can be defined as a cross-sectional research, resulting in one-time, quantitative results.
Aside from the research design, the measurement of the study has also be considered beforehand. A very important point, especially for the questionnaire, are the scales of the questions. According to Aaker & Day (1990) there are various types of scales. The most commonly known ones would be: nominal, ordinal, interval and ratio scale (Aaker & Day 1990). But there are also itemized-category scale, comparative scale, rank order scales and also very common Likert scales (Aaker & Day 1990). It is crucial to consider the scales of the questionnaire items needed in order to conduct certain statistical tests.
4.3. Data Collection Method and Sampling
design was chosen, since it is the appropriate design when examine the relationship between different variables (Bryman & Bell 2011). The data are collected with a web survey, a version of a self-completion questionnaire. These types of surveys have to be completed by the participants themselves without the supervision or interference of an interviewer (Bryman & Bell 2011). A big advantage of web surveys is that they save time and costs and deliver more results in a shorter time (Aaker & Day 1990; Bryman & Bell 2011). Furthermore the answers are more reliable since participants have to answer directly without any intermediary person (Aaker & Day 1990). It can also be more convenient for the participants, because they can fill in the survey in their own pace and in privacy, which means the answers might be more thoughtful (Aaker & Day 1990). An important advantage for this study is the anonymity offered by this method (Aaker & Day 1990). Talking about negative comments and opinions about a brand can be a sensitive topic for some. However because of the missing interviewer, the questions need to be designed very carefully and easy to understand by the participants (Aaker & Day 1990).
The questionnaire consisted of 15 questions, which can be divided into seven parts. Please refer to Appendix 9.1 for the complete questionnaire. The first part of the survey (Question 1&2) was designed to gain an overview of the respondents’ general Facebook usage. The second part (Questions 3-6) examined the respondents’ engagement with brands on Social Media. The third part (Question 7&8) focused on the respondents’ buying behavior. The fourth part (Question 9&10) examined personal opinions and interest of the respondents. The fifth part (Question 11) was a small experiment showing respondents negative user comments found on H&M’s Facebook page and then asking for their opinion once again. Part six (Question 12) aimed to gain insight into the respondent's perception of severity of different negative events that could happen with a particular brand. The seventh and last part (Question 13-15) focused on collecting demographic data from the respondents.
To make the survey easier for participants all the questions were closed-end questions. The answers of these questions are already given in forms of lists to choose from, rating scales, etc. (Aaker & Day 1990). This saves time for respondents and makes it easier to answer questions (Aaker & Day 1990). The advantage for the researcher is that the answers are directly comparable, which is important for large scale surveys (Aaker & Day 1990).
Moreover the wording of the questionnaire in a neutral way is highly important as the wording should not be leading or loaded and imply the desired answer (Aaker & Day 1990). Simple vocabulary was used to make it understandable to most people without tremendous effort. Vague wording was avoided and the instructions were formulated as detailed as possible. However, the survey of this study showed one error in the question design. The layout and instruction of one question (Question 12) was difficult to understand for some participants, which lead to incomplete answers for the particular question. Besides this no bigger challenges were discovered in the survey design. In general it has to be considered that close-end questions offer only a limited amount of response category which can make people insecure about their answers and lead to a distortion of responses (Aaker & Day 1990). On the other hand providing alternative response categories might suggest options that the participant would not have consider by himself/herself (Aaker & Day 1990). Therefore the authors tried to find a suitable amount of response option (5 to 7 options) that would display a wide enough variety and therefore cover most of the participants’ opinions without being confused by too many options. Another potential problem could be the order of the response options, as people tend to choose the middle if unsure (Aaker & Day 1990). However, in order to sustain a logical flow of the response options, the author decided to keep the middle option neutral. Another point that has to be carefully considered is having a “i do not know” option. In this particular case, the authors decided against a “i do not know” because the questions were not difficult to answer and neither compromising in a way that could have led people to not wanting to disclose their answer. Also this avoids having a large amount of “i do not know” answers because of laziness.
There are various types of probability sampling: simple random, stratified random, systematic or cluster sampling (Saunders et al. 2012).
In this particular case the process of the sampling can be defined in two ways, depending on the choice of literature.
The sample for this study consists of people who are familiar with Facebook and H&M. Since Facebook is the biggest online social network and H&M is a well-known brand, these criteria apply to a big amount of people.
When based on Bryman and Bell (2011) it can be assumed that there is no distinct sample frame for this study, because of the vast amount of people that qualify for this study. Hence without a distinct sample frame it is nearly impossible to get a probability sample (Bryman & Bell 2011). For web surveys it is suggested to distribute it online by posting the URL on relevant web pages (Bryman & Bell 2011). Within the scope of this study the web survey was posted on Facebook in different Facebook groups that were regarded relevant (e.g. “Mode/Fashion Trade Sverige”). However the responses coming from these groups were limited and it became clear that people are more likely to answer web surveys, when they are in a way related to the sender (e.g. same group of friends, studying at the same university). Therefore the survey was distributed through private Facebook pages and through itslearning, the learning platform of Karlstad University. This way of distributing can be seen as a convenient sample, where the survey is given to people one has access to (Bryman & Bell 2011). A convenient sample usually has a better response rate, but cannot be generalized so easily (Bryman & Bell 2011). Even though the recipients of the survey are not entirely unknown there is still a great number of unknown representatives, since it is unclear who answers the survey and who did not (Bryman & Bell 2011).
4.4. Quality of the Study
5. Data Analysis
The following chapter analyzes the quantitative findings of the data collection via the online questionnaire. The data analysis was carried out with the software package SPSS PASW 23, used for statistical analyzes.
5.1. Survey Results
Over a period of 27 days 557 answers could be collected. 56,9% of the participants are female and 43,1% male. Almost half of the participants are between 16 and 25 years old, a third is between 26 and 35 and only about 20 % over 35. The age distribution resulted from the distribution channel of the survey, consisting of friends and fellow students. The age groups over 35 were not easily accessible for the authors. The majority of responders are students (75%), followed by full-time workers (18,3%) and part-time workers (4.5%), only a small amount of participants are in the category unemployed (1,6%) and stay-at-home parent (0,9%). Again, this distribution is a result of the distribution of the survey, as a major channel was the university’s network itslearning. However the demographics of the survey are suitable for this study and reflect an important target group of H&M. H&M states on their website that their products are for everyone, young or old, male or female. The low priced products however make it very accessible for students. When looking at the H&M Facebook page (https://www.facebook.com/hm/) the target group seems to be females in their twenties. For one the models used in the posts and campaigns are mainly in their mid-twenties and female. Furthermore the campaigns are attractive for a younger audience, e.g. the Coachella campaign in April. The audience of Coachella, a music festival in the US, is mostly between 18 and 35 years old (Nielsen 2015).
Student
Working full-time Working part-time Currently unemployed Stay at home parent Retired 415 (74.6%) 102 (18.3%) 25 (4.5%) 9 (1.6%) 5 (0.9%) 0 (0.0%) Table 1 Demographic results
Most questions were marked as mandatory and had to be answered. Exceptions are question 4, which was skipped if anyone answered “no” in the previous question and question 8, which was skipped if the previous answer was “never”. Question 12, which asked about the impact of different events on brand opinions, was changed to a voluntary question since the layout was misleading and some participants got stuck in the questionnaire. Some of the mandatory questions are missing one or two answers, these incomplete data sets were deleted.
As mentioned before the first part of the questionnaire examines the Facebook usage of the participants. 43,1% of all participants stated to use Facebook less than 7 hours per week. About a third (33,8%) said they use the social network between 7 and 14 hours each week and 14,7% use it between 15 and 21 hours. Less than 10% (8,4%) are using Facebook over than 21 hours per week. The majority of respondents (55,2%) are following 0 to 5 brands on Facebook. 21,6% said they are following 6 to 10 brands, 10,8% follow 11 to 15 brands, only 4% are following 16 to 20 brands and 8,5% of the participants are following more than 20 brands on Facebook.
In the second part of the questionnaire we learn more about the participants’ engagement with brands on Facebook. Most of them (68,8%) stated they are not engaging with the content brands share on Facebook and a third of the participants (31,2%) does engage with brand content. When asked about the nature of the engagement, almost all participants are liking posts (96,6%), nearly half of them also comments on posts (46%) and about a third is sharing the content with friends (36,4%). Just a small minority of 8,5% is following the H&M brand page on Facebook, the rest (91,5%) does not follow the page. Since most participants are not following the H&M page, it is no surprise that 86,5% are never engaging with the content of H&M. Nearly 10% answered they engage once a year with the posts, 3,4% once a month and only one person answered the he/she engage once per week. None of the participants is engaging with H&M posts on a daily basis. When asked about their purchasing behavior one quarter of the responds are buying at H&M once every sixth month, a bit less than a quarter purchases once every three month and about 20% once every year. 8,1% of all participants is buying once every month and 20,3% stated they never buy at H&M. From those who are buying, almost 80% are buying in store, 8,5% online and 12,1% are equally buying online and in store.
Graph 2 Results v5 - Purchase frequency
negative to positive. Only 7% answered that they have a negative opinion about H&M, whereas 12% answered they have a positive opinion. 17,4% of the participants have a somewhat negative opinion and 30,7% have a somewhat positive opinion. About a third stated they have neither a positive nor a negative opinion about H&M. From the results it can be seen that the opinion was more positive than negative. When asked about their interest in the brand 21,7% answered they are not at all interested in H&M, 16,7% are somewhat uninterested and almost a third of the participants are neither interested nor uninterested. On the other side 27,3% are somewhat interested and 5% are very interested in H&M. After that the participants were shown the negative comments and asked about their opinion once more. This time on a seven point scale, in case someone with an already negative or positive opinion have now a very negative or very positive opinion. 9,4% stated that their opinion now is very negative, 15,6% have a negative opinion and 23,7% a somewhat negative opinion. Over a third of the participants (35,1%) have neither a positive nor a negative opinion about H&M after seeing negative comments. On the positive side, 9,9% stated they have a somewhat positive opinion, 4,9% have a positive one and even 1,4% answered they have a very positive opinion.
Graph 4 Results v2 - opinion after
The last question was supposed to help understanding the impact of negative events on people’s opinions about a brand compared to other events. Unfortunately not everyone answered this question correctly. From the answers gotten for this question the following ranking can be created (1 - most impact, 5 - least impact):
1. Bad product experience 2. Bad customer service 3. Brand scandal
4. Negative press
5. Negative user comments on Social Media
37,6% answered that negative comments have the least impact on their opinion compared to the other events and only 10% said they have the most impact.
5.2. Preparation for Analysis
5.2.1.Variables
From the hypotheses postulated in the first part we can extract different variables. First of all different groups of people should be analyzed: highly engaged users, lowly engaged users as well as frequent purchasers and infrequent purchasers. The focus of the study is the opinion of these groups and how it is influenced by negative comments.
The first variable will be the opinion before reading the mean comments and the second variable will be the opinion after reading the comments. To clarify the differences between the opinions, and thus see how it changed, a third variable is calculated based on the first two.
v1 - opinion before (dependent variable) v2 - opinion after (dependent variable)
v3 - difference between opinion before and after (v3 = v2 - v1) The different groups are presented by the variable for Facebook usage and the variable for purchase frequency.
v4 - Facebook usage (independent variable) v5 - Purchase frequency (independent variable)
All of the variables are ordinal variables. Their categories can be ranked in a specific order, but the distance between each category is not the same (Bryman & Bell 2012).
Variable one, two and three are quantitative variables. This means they are measured as a number. Even though the answers are not numeric, they can be written as a number. For example when question 14 was answered with “somewhat negative” it is coded as -1 and the answer “positive” with (+)2. The other two variables are categorical variables. Their categories have no numeric meaning.
How positive or negative is your opinion of H&M? negative Somewhat negative Neither negative nor positive Somewhat positive positive -2 -1 0 1 2
Table 2 Coding of variable 1
After reading these comments, how positive or negative is your opinion about H&M now? Very negative negative Somewhat negative Neither negative nor positive Somewhat positive positive Very positive -3 -2 -1 0 1 2 3
Table 3 Coding of variable 2
Variable 3 shows the change of the opinion and is therefore the difference between v1 and v2. If the mean of v3 equals -0,77 it means that the opinion changed by -0,77 points, and in that case has gotten more negative (as indicated by the negative value).
5.2.2.Statistical Tests
The t-test is one of the most basic tests in statistical analysis. It is used to analyze significant differences between the means of two groups (Mertler & Vannatta 2002). The t-test is applied when the independent variables have only two answer categories (e.g. yes/no, male/female,...) (Mertler & Vannatta 2002). The one-way analysis of variance, also called ANOVA, has the same purpose as the t-test, but it is applied when the independent variable has more than two categories as answers. Since the ANOVA test only identifies differences in group and not which of the groups are different a post hoc test is needed (Mertler & Vannatta 2002).
5.3. Analysis
At first it is tested if there is a statistical difference between v1 and v2 by
comparing the means of the two variables. By using a t-test it is determined if there is a significant difference in the means of the two variables. See Appendix 9.2 for the results in SPSS. When looking at the mean of both variables the mean of the opinion before reading negative comments was higher (0,24) than for the opinion after (-0,59). Participants had a more positive opinion about H&M before reading the comments and got a more negative opinion after reading negative comments. This difference is significant (p < 0,05), which means there is a significant difference in the opinion of people before and after reading negative comments on Facebook.
Mean N
v1 How positive or negative is
your opinion of H&M? ,24 556
v2 After reading these comments, how positive or negative is your opinion about H&M now? - How positive or negative is your opinion of H&M?
-,59 556
Paired Samples Test Paired Differences t df Sig. (2-tailed) 95% Confidence Interval of the Difference Upper Pair 1 How positive or negative is your
opinion of H&M? - After reading these comments, how positive or negative is your opinion about H&M now? - How positive or negative is your opinion of H&M?
,923 17,017 555 ,000
Table 5 t-test v1 - v2: p-value (p < 0,05)
In the next step the independent variable v4 and the dependent variable v3 are
tested for significant differences. Since the independent variable has four answer categories the ANOVA test is applied. The mean represents the differences between the opinion before and after reading negative comments. For example the opinion of the group that uses Facebook under 7 hours per week changed by -0,77 and thus has gotten more negative. All groups changed their opinion to a more negative one. However the difference is not significant (p > 0,05). The post-hoc test shows that there is no significant difference between any of the groups.
v4 Facebook usage N Mean (v3)
under 7 239 -,7699
7 - 14 h 188 -,8617
15 - 21 h 82 -,8171
over 21 h 47 -1,0000
Total 556 -,8273
Table 6 ANOVA test Difference opinion - Facebook usage
Difference (v3)
Sum of Squares df Mean Square F Sig.
Between Groups 2,421 3 ,807 ,613 ,607
Within Groups 727,003 552 1,317
Total 729,424 555
The ANOVA test is also used when testing v3 and v5 for significant differences.
Just like in the group for Facebook usage, the opinion of the different purchase groups about H&M became more negative. The ANOVA and post-hoc test show that this difference is not significant (p > 0,05), neither in the groups in general nor between specific groups. It cannot be generalized that people who frequently purchase H&M clothes are differently impacted by negative opinions that people who infrequently buy at H&M.
v5 Purchase frequency N Mean (v3)
never 112 -,6161
once a year 127 -,7717
once every six month 140 -,8214
once every three month 132 -1,0076
every month 45 -1,0000
Total 556 -,8273
Table 8 ANOVA test Difference opinion - Purchase frequency
Difference (v3)
Sum of Squares df Mean Square F Sig.
Between Groups 11,027 4 2,757 2,114 ,078
Within Groups 718,397 551 1,304
Total 729,424 555
Table 9 ANOVA test v3 - v5: f-value/p-value (p > 0,05)
An additional variable was tested to its significant difference with variable v3: v6
- interest in H&M. This variable is not a part of the hypothesis, but can be tested
V6 Interest in H&M N Mean (v3)
not interested at all 121 -,6612
somewhat uninterested 92 -,8043
neither interested nor
uninterested 163 -,7914
somewhat interested 152 -,9671
very interested 28 -1,0714
Total 556 -,8273
Table 10 ANOVA test Difference opinion - Interest in H&M
Difference (v3)
Sum of Squares df Mean Square F Sig.
Between Groups 8,238 4 2,060 1,574 ,180
Within Groups 721,186 551 1,309
Total 729,424 555
Table 11 ANOVA test v3 - v6: f-value/p-value (p > 0,05)
To see if there was a difference at all between groups in their opinion about H&M, variable v1 and v5 as well as v2 and v5 were also tested with an ANOVA
test.
Before reading the negative comments it is clearly visible that the opinion about H&M is getting more negative the less a participant buys at H&M. This is not a surprise since people with a negative opinion are unlikely to buy products from H&M and people with a positive opinion buy frequently at H&M. These differences are significant (p < 0,05) and the post-hoc test shows that the differences between all single groups are significant as well.
V5 Purchase Frequency N Mean (v1)
never 112 -,58
once a year 127 -,05
once every six month 140 ,29
once every three month 132 ,80
every month 45 1,24
Total 556 ,24
How positive or negative is your opinion of H&M? (v1)
Sum of Squares df Mean Square F Sig.
Between Groups 173,380 4 43,345 49,066 ,000
Within Groups 486,755 551 ,883
Total 660,135 555
Table 13 ANOVA test v1 - v5: f-value/p-value (p < 0,05)
When asked about their opinion after reading negative comments it is still visible that people who purchase every month keep a positive opinion about H&M, even though it might be less positive than before. There are significant differences in this group. The post-hoc test however shows that this significance exists only between the following groups:
Never - once every six month Never - once every three month Never - every month
Once a year - once every three month Once a year - every month
Once every six month - every month
V5 Purchase frequency N Mean (v1)
never 112 -1,20
once a year 127 -,82
once every six month 140 -,54
once every three month 132 -,20
every month 45 ,24
Total 556 -,59
Table 14 ANOVA test Opinion after - Purchase frequency
After reading these comments, how positive or negative is your opinion about H&M now? - How positive or negative is your opinion of H&M? (v2)
Sum of Squares df Mean Square F Sig.
Between Groups 99,199 4 24,800 15,369 ,000
Total 988,322 555
Table 15 ANOVA test v2 - v5: f-value/p-value (p < 0,05)
With the help of the statistical test we can decide whether the hypothesis will be rejected or accepted. The first two hypotheses, that customers with a high engagement with H&M are stronger negatively influenced and customers with a low engagement are less negatively influenced by negative comments, was tested with the comparison of v3 and v5. The purchase frequency indicates a low
or high engagement in the brand, since people who do not buy frequently have less contact with the brand than people who are frequent buyers. The result of the test did not show a significant different between the engagement groups, which means hypotheses one and two will be rejected. There is no significant evidence that high engagement customers are stronger negatively influenced than low engagement customers. Furthermore there is no significant evidence that low engagers are less negatively influenced than high engagers.
The third and fourth hypotheses, claiming that frequent Facebook users are stronger negatively influenced by negative comments and infrequent users are less negatively influences, was tested with v3 and v4. A significant relationship
between the Facebook user groups could not be proven. People using Facebook more frequently are not influenced in a stronger way than people using Facebook more infrequently. This means hypotheses three and four will be rejected as well.
It could still be proven that a significant difference between people’s opinion about H&M before reading negative comments was more positive than after reading the comments. Thus negative Facebook comments do have an impact on the opinion about a brand. The impact does not differ between different groups of purchasers, Facebook usage or even interest in H&M.
