Enhancing sponsoring effectiveness through
mobile apps
University of Groningen Faculty of Economics and Business
Master thesis Marketing Management Strategic Innovation Management
Author: Jelmer de Vries
s1881353
Supervisor: dr. T.L.J. Broekhuizen
Second supervisor: dr. J.C. Hoekstra
Abstract
This study focuses on sponsoring via the mobile channel. It researches how mobile sponsoring, used in combination with traditional sponsoring, enhances total sponsoring effectiveness. A chain of effects is hypothesized which predicts that mobile sponsoring enhances sponsoring effectiveness through perceived interactivity of the app, which leads to engagement with the app, which leads to brand recall. The relationship between engagement and brand recall is expected to be moderated by high brand equity and high sponsor fit. The results show that interactivity is positively related with brand recall and not via engagement. Furthermore brand equity and sponsor fit have a direct positive relationship with brand recall, however, they do not have a moderating role. It is concluded that event organizers should develop an interactive app, which is used as platform for sponsors. The usage of this app should be stimulated for maximum results.
Keywords: Mobile sponsoring, mobile app sponsoring effectiveness, interactivity, engagement, brand equity, sponsor fit, brand recall
Acknowledgements
Table of content
1. Introduction ... 3
2. Literature ... 4
2.1 Sponsoring ... 4
2.2 Conceptual model ... 5
2.3 The relationship between Interactivity and Engagement ... 7
2.4 The relationship between Engagement and brand recall ... 9
2.5 Sponsor- event fit ... 10
2.6 Brand equity and brand recall ... 11
3. Methodolgy ...12
3.1 Case selection ... 12
3.2 Data collection and design ... 12
3.2.1 Sample description and average recall rates ... 13
3.3 measurements ... 14 3.3.1 Interactivity... 14 3.3.3 Brand recall ... 16 3.3.4 Sponsor fit ... 16 3.3.5 Brand Equity ... 17 3.4 Control variables ... 17 3.5 Correlation table. ... 18 3.6 Analysis ... 18 4. Results ...19
4.1 Independent sample t-test: recall rates for app-users versus non-app-users ... 19
4.2 Linear Regressions ... 19
4.4 Usage frequency test ... 22
4.5 Brand Equity and Sponsor fit ... 24
4.5.1. Direct effect ... 24
4.6 Overview of hypotheses... 25
5. Conclusion and discussion ...26
6. Managerial implications ...28
References ...31
1. Introduction
Sponsoring is increasingly used as a marketing communication tool. Investments in sponsorship have increased massively (Cornwell, Weeks, and Roy, 2005; Gwinner, Larson and Swanson, 2009; Jagre, Watson, Watson, 2001; Kim, Lin, and Sung., 2013; Roy and Cornwell, 2003; Tsiotsou and Alexandris, 2009). Sponsoring is a cost-effective marketing communication tool (Jagre, Watson, and Watson 2001). Event sponsoring is popular, because events offer an important platform where the company’s brands present themselves to a lot of consumers. Companies are only willing to sponsor when they think it is effective. Effectiveness of event sponsoring is a popular field of research (e.g. Cornwel et al., 2005; Lacey, Sneath, Finney, and Close 2007). In general, results show that sponsoring has a positive effect on brand recall (Barros, de Santos, and Chadwick, 2007; Nicholls, Roslow, and Dublish, 1999; Rowley and Williams, 2008). Most of these studies are performed in the offline (i.e. traditional) context, while almost all events reach out to their consumer through online and mobile channels. To test whether mobile sponsoring used in combination with traditional sponsoring would enhance sponsoring effectiveness, this research tries to extend previous research and focuses on a specific online channel: mobile sponsoring via apps. It can be very useful and interesting for companies to invest in sponsoring via this new communication channel.
One way of anticipating for the increasing trend of this new communications channel, is the development of event apps which are apps, especially made and used for a particular event. The number of smartphone application users grows increasingly (Kim et al., 2013). The mobile phone penetration (i.e. the proportion of consumers which own a mobile phone) by 2014 will be 90% (Gartner, 2010). Emarketer (2014) states that the smartphone penetration is growing as well. In 2014 the growth is 25%, which brings the total number of smartphones in use to 1.76 billion by the end of 2014. In addition, consumers spend more time using their mobile phone, up to 3.3 hours every day (Exact Target, 2014). When consumers use their mobile devices, 89% of the time they use mobile apps (Bosomworth, 2014). Research of ABI research (2014) reports the total number of app downloads by 2016 will reach nearly 44 billion. Next to that, consumers use their smartphone more to search the web, than they use their desktop computer (Business Insider, 2014). This all means that this new communication channel offers great possibilities and opportunities for mobile marketing, and thus mobile sponsoring. For instance investing sponsoring budget more effectively. However, little research is done concerning mobile sponsoring.
would be the case. With mobile sponsoring is meant commercial sponsoring in a non-branded event app. This will help marketing managers to determine where and how to invest their sponsoring budget in an optimal way. In this research, it is assumed that if consumers find the app more interactive, they find the app more engaging, which leads to a higher frequency of recalling the sponsored item. In addition, it is assumed that consumer engagement with the app stimulates explorative seeking behavior in the app. This leads to an increase in encountering the sponsors in the app, which in turn improves brand recall. Another way this relationship is explained, is by following the S-O-R framework (Mehrabian and Russel, 1974). This research states that an environment (the app) contain stimuli (interactivity), which leads to an internal state (engagement), which in turn leads to certain behavior. Interactivity and engagement are central concepts, in a way that interactivity is conceptualized as antecedent of engagement and engagement as antecedent of brand recall. Next to that, it is assumed that sponsor fit and brand equity both have a positive direct effect on brand recall and enhance the relationship between engagement and effectiveness.
Before this is discussed in more detail, it is important to define the terms. Mobile sponsoring refers in this research to sponsoring (displaying the brand) in a mobile application, which is not initiated by a brand itself. Furthermore, mobile sponsoring can be done through a smartphone application (from here: app). Smartphone apps are “end-user software applications that are designed for a cell phone operating system and which extend the phone’s capabilities by enabling users to perform particular tasks” (Purcell, Entner, and Henderson 2010, p. 2). The scope of this research aims at apps provided by event organizers, which have a utilitarian or hedonic function, and offer a platform for brands to sponsor the event.
The paper is structured as follows: Chapter 2 presents the conceptual model , followed by an explanation of the proposed relationships. After this, hypotheses are formed, which is followed by the method section, which explains the methodology. The results and conclusion are followed by the managerial implications. The limitations and future research is last part of this thesis.
2. Literature 2.1 Sponsoring
awareness and image (Grohs, Wagner, and Vsetecka, 2004; Pham, 1991). Sponsoring goals can be divided into two different categories: consumer psychological goals and consumer financial goals (Kim, 2010). The first is related to such goals as brand awareness and image as well as affection. The latter is mainly based on stock price changes (e.g. Tsjotsou and Lalountas, 2005). This thesis focuses only on brand awareness, which is measurable through brand recall.
The literature on the effects of sponsoring on brand recall, shows positive effects of sponsoring on brand recall. Most of these researches are, however, done in an offline context (Barros et al., 2007; Nicholls et al., 1999; Rowley and Williams, 2008). Little research has established the effects of sponsoring in online contexts. The results of this research seem to suggest that online sponsoring also works. However there is no research done on how online app sponsoring can enhance the offline sponsoring effectiveness. Often, at events, there is a combination of online and offline sponsoring. Literature shows both effectiveness of offline and online methods, therefore, it is the combination of both which determines the total sponsoring effectiveness. For this reason, this study focuses on how the use of mobile app sponsoring enhances the effectiveness of total sponsoring (i.e. the combination of mobile and traditional sponsoring). Because of these positive results in offline environments, it is expected that sponsoring in a mobile environment will be effective as well and thus will enhance sponsoring effectiveness. Next to that, it is expected that mobile sponsoring enhances sponsoring effectiveness because in-app sponsoring has the feature of interactivity. Mobile apps are unique in the way they are interactive, therefore this a distinctive element of this channel (Kim et al, 2013).
2.2 Conceptual model
Figure 1 shows the chain of effects concerning mobile sponsoring effectiveness. The advent of the mobile channel offers great opportunities for companies, for instance, to enhance total sponsoring effectiveness via mobile sponsoring. The key distinctive feature of this new communication channel is its interactivity (Kim et al, 2013). This paper follows the perceptual approach of interactivity (e.g. Broekhuizen and Hoffmann, 2012) which means that interactivity is in the eye of the beholder, and accordingly interactivity is referred to as perceived interactivity.
+ + +
+
+
construct that is widely discussed in literature (e.g. Ariely, 2000; Bretz and Schmidbauer, 1983; Broekhuizen and Hoffmann, 2012). It is the unique key feature of mobile apps and therefore it is the starting point of this chain of effects. In this study it is linked to engagement, which is linked to brand recall. This is fairly new in literature, thus the link with engagement is introduced is this study. Next, this study also looks at brand effects.
Interactivity can, via engagement be linked to positive marketing outcomes (brand recall) (Mollen and Wilson, 2010; Wang, 2011). Higher perceived interactivity is assumed to lead to higher engagement with the app. Engagement yields positive marketing results (Calder, Malthouse, and Schaedel, 2009; Mollen and Wilson, 2010; Van Doorn, Lemon, Mittal, Nass, Pick, Pirner, Verhoef, 2010), such as increased advertising recall (Wang, 2006). The positive relationship between engagement and brand recall can be explained, reasoning that when a consumer is engaged with the app, the usage frequency will increase, which leads to increased brand recall. Furthermore, engagement stimulates explorative seeking behavior and motivation to use the app to its full potential. Thus encountering the sponsors in the app page increases, which leads to a higher brand recall.
This study also focuses on the role of sponsor fit. There is disagreement whether the relationship between sponsor fit and brand recall is positive or negative; some research indicate that this relationship is positive (Simmons and Becker-Olsen, 2006; Walraven, Bijmolt, and Koning, 2014); other research states the opposite (Jagre et al., 2009; Olson and Thjømøe, 2009). Because of the mixed findings, it is interesting to research this in the mobile environment. Next to a direct effect on brand recall, it is also assumed that sponsor fit positively moderates the positive relationship between engagement and brand recall.
Furthermore, brand equity may impact brand recall rates (Keller, 2013). When a company has brand equity, there is also brand knowledge. When general brand knowledge is high, it is easier for a consumer to remember the brand, since it takes less cognitive processing.
Figure 1: conceptual model Perceived
Interactivity of the app
Engagement with the app
This study focuses on how mobile app can enhance the overall sponsoring effectiveness. Therefore, the overall hypothesis is as follows:
H1: Mobile sponsoring enhances sponsoring effectiveness. 2.3 The relationship between Interactivity and Engagement
Definition of interactivity. Definitions and conceptualizations of interactivity vary in wide range in literature. Roughly these conceptualizations can be divided in three categories: message-centered approaches, structural approaches and perceptual approaches (Bucy and Tao, 2007). When the app is interactive, it has hedonic and utilitarian benefits (Broekhuizen and Hoffmann, 2012). These benefits are perceived by individual users, because they come from individual experiences. Because of this, the perceptual approach of interactivity is used. Following the conceptualization of Yadav and Varadarajan (2005), interactivity is defined as the level a consumer experiences two-way communication, user control, responsiveness, when using an mobile app.
Two-way communication refers to degree that users are both senders and receiver of information (Burgoon, Buller and Foyd, 2001; Bretz, 1983). For example, users can send feedback via the app or evaluate activities via the app.
User control refers to the level of control users experience, in a way that users feel the freedom to use and navigate through the system as they like, when they use the app (Ariely, 2000; Broekhuizen and Hoffman, 2012; Downes and McMillan, 2000; Lynch and Ariely, 2000; Mundorf and Bryant, 2002). Users can navigate through specific content, and they are not obligated to use the app in a certain way.
Responsiveness is the speed of the response (Novak, Hoffmann, and Yung, 2000) and the degree of correspondence between the response and information applied for (Alba, Lynch, Weitz, Janiszewski, Lutz, Sawyer, Wood, 1997; Broekhuizen and Hoffmann, 2012; Lombard and Ditton, 1997; Lombard and Snyder-Duch, 2001). High responsiveness means that users can use and navigate through the app in the way they like and get the information they solicited for, at their own desired pace.
Two-way communication User control Responsiveness Alba et al. (1997) X
Ariely (2000); Lynch and Ariely (2000)
X Bretz and Schmidbauer
(1983)
Table 1. Overview of characteristics of interactivity
Definition of Engagement. Engagement is an important concept (Mollen and Wilson, 2010) with many different definitions (Calder et al., 2009). Engagement is a concept that covers different terms which are all mechanisms that can lead to competitive advantages (Mollen and Wilson, 2010). Calder et al. (2009), use the concept of experience when defining engagement. This study follows this concept, and uses the definition of Higgins (2006). Engagement comes from motivational experiences and consists of three elements: involvement, being occupied and interested in the item. This leads to the following definition: the mental state that comes from experiences, in which the consumer undergoes a cognitive process by feeling involved, occupied and interested in something. Thus, engagement is conceptualized as a mental state which comes from an experience. An important part of engagement is the mental state that is accompanied by cognitive processing (Mollen and Wilson, 2010). This study argues that, through experiencing the app, consumers can feel involved, occupied and interested. Following the elaboration likelihood model, users can via, higher engagement, more intensely process the content of the mobile app, and thereby better remember what they have read. In other words, when people are engaged with the app, they are likely to think harder on what they see and read in the app (Petty and Cacioppo, 1984).
Relation between interactivity and engagement. Broekhuizen and Hoffmann (2012) state that “the level of perceived interactivity determines the extent to which consumers experience positive
1
Broekhuizen and Hoffmann also state that multimedia usage and fulfillment is an element of interactivity
Broekhuizen and Hoffmann (2012)1
X X X
Burgoon et al. (2001) Participation Downes and McMillan
(2000)
X X X
Johnson, Bruner, and Kumar (2006)
Reciprocity X
Mollen and Wilson (2010) X X X
Mundorf and Bryant (2002) Selectivity
Novak, Hoffman, and Yung (2000)
X
McMillan and Hwang (2002) X X X
Song and Zinkhan (2008) X X X
Wu and Wu (2006) X X X
Yadav and Varadarajan (2005)
intermediate states” (p. 31). These can be hedonic or utilitarian. Because these benefits, the consumer gets interested and involved in using the app and gets therefore more engaged (Broekhuizen and Hoffmann, 2012). Next, interactivity, results in a more positive attitude towards a product or website (Sicilia, Ruiz, Munuera, 2005). This positive attitude also leads to more engagement. Furthermore, because of perceived interactivity, people can have the feeling that they are in the online environment (also known as telepresence), which is an antecedent of engagement (Mollen and Wilson, 2010). Wang (2011), states that engagement is generated by interactivity. Following the Stimulus-Organism-Response (S-O-R) framework (Eroglu, Machleit, Davis 2003), it is assumed that an (online) environment, such as the app, caries stimuli, which have impact on consumers’ internal states (or organismic states), which result in a certain response (Mehrabian and Russel, 1974). This response can have positive marketing outcomes, such as brand recall. This research assumes that the event app is the environment that contains the stimuli, which causes the internal state of perceived engagement (Mollen and Wilson, 2010). Based on above, this study assumes that those who experience the app to be more interactive (in terms of two-way communication, user control, and responsiveness), experience a greater engagement with the app. This leads to the following hypothesis:
H2: A higher perceived interactivity leads to a higher engagement 2.4 The relationship between Engagement and brand recall
Engagement has been related to positive marketing outcomes, such as financial, reputational, regulatory, competitive, employee, and product consequences (van Doorn et al., 2010). Higher engagement leads to greater advertising recall (Wang, 2006). Also message involvement, message credibility, attitude towards the message and advertising, increases with higher engagement (Wang, 2006). Thus, engagement increases advertising effectiveness. Next, engagement is a necessary condition to increase brand equity, of which brand awareness is an important element (Mollen and Wilson, 2010). Furthermore, when consumers are engaged with a media vehicle, they are more responsive to the advertising (Calder et al., 2009).
recall. Based on above, it is expected that engagement lead to a greater recall. Hence, the following hypothesis is used:
H3: A higher engagement with the app leads to a higher brand recall. 2.5 Sponsor- event fit
The fit of the brand with the event can influence sponsor effectiveness (Becker-Olsen and Simmons, 2002; Cornwell et al., 2005; Olson and Thjømøe, 2009; 2011). Fit is often conceptualized as the level of how sensible it is that a particular brand, sponsors a particular object (Olson and Thjømøe, 2011). There are two different types of sponsor fit. The first type of fit is the level of relatedness between the consumers’ (i.e. visitors) demographics, interests, and lifestyle and the characteristics of the event. A sponsoring company is then targeting a specific type of customers. The second type of fit is between an event and how close a company is associated and related to the event (Jagre et al., 2001). In this research, the second type of fit is taken into account.
Literature shows that there is a relationship between sponsor fit and brand recall. Olson and Thjømøe (2011) state that fit (also known as congruence) is related to the ability of consumers to remember the sponsoring brand (Cornwell, Humphreys, Maguire, Weeks, Tellegen, 2006). For a sponsorship to increase brand equity, a high sponsorship fit is effective in the short (Simmons and Becker-Olsen, 2006) and long run (Walraven et al., 2014), because consumers increasingly remember the brand when the sponsor is linked multiple times with the object.
There is, however, also evidence that sponsor incongruence results in higher recall, so that recall is higher when sponsor fit is lower, but brand affect is higher (Olson and Thjømøe, 2009). Jagre et al. (2001) proposed that low sponsor fit with the object results in a higher recall rate than companies with a moderate or a high sponsor fit. An explanation for this is because companies with a low fit stand more out and thus cognitive processing is stimulated more, which leads to a more intense processing, which increases recall the company from the memory.
recognition was moderate at best. Respondents often made a type 2 error, (i.e. identifying companies as being sponsor when in fact they were not sponsor). The reason for this was that consumers where guessing that certain companies where sponsor because they could easily be associated with the event. This error is tested in this study. Nevertheless the majority of research shows a positive relationship between sponsor fit and brand recall, therefore the following hypothesis is formulated:
H4: Sponsor fit has a positive relationship with the brand recall of the sponsoring companies.
Next to the direct effect, a moderating effect of sponsor fit on the relationship between engagement and brand recall is also hypothesized. This study assumes that synergetic effects will occur, such that when consumers are engaged with the app, and confronted with a sponsor that scores high in sponsor fit, they will more likely remember this brand, as opposed to low fit brands. When sponsor fit is high, it takes less cognitive processing to link a company to the event and because of shared associations it is easier to remember a high fitting sponsor. Thus engagement will have a stronger effect on recall with high sponsor fit. Therefore it is hypothesized that:
H5: The positive relationship between engagement and recall is positively moderated by sponsor fit. 2.6 Brand equity and brand recall
Brand equity directly affects brand recall. Customer-based brand equity is: “the differential effect that brand knowledge has on consumer response to the marketing of that brand” (Keller 2013, p. 69). The idea is that customer-based brand equity resides in the minds of customers, their knowledge of the brand is what determines the brand equity, and consumers respond more strongly to the marketing efforts of high equity brands. Brand knowledge consists of two components: brand awareness and brand image.
When a brand has high brand equity, it is easier to remember the brand (Roy and Cornwell 2003). When brands have a high brand equity, consumers have high brand knowledge. This means that when consumers encounter the brand, they are familiar with the brand. This familiarity with the brand leads to an increase in brand recall, since it is easier to remember a brand the consumer is familiar with. This is due to the lower cognitive processing needed to remember a familiar brand. Therefore, brand equity is directly and positively related to brand recall.
impact is stronger for brands that they can easily remember. In other words, when a consumer encounters a familiar brand in the app it is easier to remember the brand than when the brand is unfamiliar. Thus, the following hypotheses are formulated:
H6: Brand equity has a positive relationship with the brand recall of the sponsoring companies H7: The positive relationship between engagement and recall is positively moderated by brand equity
3. Methodology 3.1 Case selection
In order to test whether mobile sponsoring leads to an increase in brand recall compared to as traditional sponsoring, a mobile app of the study association of the faculty of Economics and Business of the University of Groningen, is used. The event of interest is the EBF Recruitment Days. At this three-day recruitment event, 39 companies present themselves as potential employers for students. Students can meet the different companies during this event in a variety of different activities. For this event, a mobile app is developed and facilitated by the organizing study association. This app is free and contains the following features: the program of the event of each of three days, general information, a feedback button, a link to the Facebook page and Twitter page of the event organization, a map, information about the app (where users can give feedback about the app), a page where information about the companies can be found, and a feedback form of the event.
General feedback for the app is that it did not have a lot of relevance. The information which was stated in the app, was for the greater part also provided in a hardcopy booklet.
3.2 Data collection and design
Table 2. Descriptive statistics
† number of times a respondent used the app before and during the event
control variables, 3 items for socio-demographics), which means that the sample should at least consist of 95 respondents. In total 128 of respondents were asked to fill the survey, of which 126 completed the survey. The questionnaire was translated into Dutch, as the visitors of the event were also Dutch. The survey consisted of 19 questions, Only respondents who have used the app needed to fill in 19 questions. When users did not use the app, they needed to fill in 7 questions (12 questions were app use related). It was chosen to use single items, while maintaining the adequate measurements of constructs. This was important because respondents have less motivation to fill in a long survey. Next to that, longer surveys are filled in less accurate, thus it also enhances bias (MacKenzie and Podsakoff, 2012). In appendix A, the survey is showed. Before the survey started, respondents were explained to what the goal of the research was and respondents were assured that the survey is anonymous. This was important to reduce method bias to a minimum (MacKenzie and Podsakoff, 2012). When filled in, the researcher coded the questionnaire according to the day at which the visitors did attend the event.
3.2.1 Sample description and average recall rates
In total 128 respondents were asked to fill in the survey, of which 126 were valid. 53 respondents (42.1%) were app users. On average respondents were able to identify 7.48 companies (19.2% of total possible companies). The respondents were for the greater part students of the faculty of economics and business. Only 5.6% was student of another faculty. A majority of the respondents started their studies in 2009 (31%). A total of 90.5% of the respondents was a MSc. student. Most respondents (27.8%) follow the MSc Business Administration with the MSc. Marketing being the second largest study (19%). The third largest representative study was MSc. Finance (15.1%).
The sample represents the population (students of the Faculty Economics and Business and the University of Groningen) well. Of the population (M=65%, F = 35) MSc Business Administration is the largest study, followed by MSc Marketing and MSc Finance is third (Educational Administration FEB, 2015).
N respondents 126 (m=68.8%, f=31.2%)
N companies recalled mean 7.48
N app users 53
N Non app-users 73
Start study mode 2009
Start study median 2009
App usage frequency mean† 3.15
Table 3. Cronbach’s Alpha Interactivity. 3.3 measurements
3.3.1 Interactivity
Four items were used to measure interactivity. The four items represent the three constructs: two-way communication, user control, and responsiveness. These are the three elements in literature that explain interactivity (see table 1). In table 1 is stated that Broekhuizen and Hoffman (2012), Downes and McMillan (2000), McMillan and Hwang (2002), Mollen and Wilson (2010), Song and Zinkhan (2008), Wu (2006), and Yadav and Varadarajan, use these measures for interactivity. Therefore, these measures are also used in this thesis. The first measure is two-way communication, which is achieved when an app offers the possibility to dialog with the firm behind the app, for example to give feedback. (Johnson et al., 2006). User control refers to the choice consumers have to navigate through the app as they without interference of pop-ups consumers did not ask for (Ariely, 2000; Broekhuizen and Hoffman, 2012; Lynch and Ariely, 2000).) Responsiveness is “the speed of response in the interface and the degree of correspondence between the response and information solicited” (Broekhuizen and Hoffmann, 2012, p. 31). The same definition is used in this thesis. Cronbach’s alpha of these measures was .678.
A factor analysis was performed, to ensure every item belongs to the corresponding construct. Interactivity was the only reflective factor, thus only for interactivity a factor analysis was done. The KMO measure was .58 (p <.000), which is appropriate since the value needs to be between .5 and 1 (Malhotra, 2010). Thus, the interactivity scale consists of 1 factor. The communalities, which measure the amount of unique information that is captured in the factor, were all >.4, except for 1 item. However, because KMO was > .5, the sample was considered adequate and large enough to compensate for this.
Item’s Interactivity Cronbach’s Alpha if items deleted I have a lot of control when I using
the app
.630 I find the speed of the app high .535
The app reacts like I expect .626
The possibilities to send information via the app, are good in my opinion
Table 5. communalities and loadings of Interactivity (loadings need to be >.5) Table 4. Interactivity measures
Item Communalities Loading
I have a lot of control when using the app .533 .730
I find the speed of the app high .700 .837
The app reacts like a I expect .475 .690
The possibility to send information via the app are good in my opinion.
.375 .613
Since the Cronbach’s alpha is high (>.6) and the sample size was large enough, it can be stated that the construct is reliable. The communalities and loadings are of sufficient levels. The factor loadings must be >.5 for validity to be acceptable. Thus it was stated that the construct of interactivity is valid
(Hair, Anderson, & Tatham, 1987). 3.3.2. Engagement
Following So, King and Sparks, (2012), this study used a formative scale to measure engagement, based on 4 dimensions (see table 6). The correlation between these items is low, they are not interchangeable, and they form the construct. (Coltman, Devinney, Midgley, and Venaik, 2008; Jarvis, Mackenzie, and Podsakov, 2003). For engagement, the formative scale was based on a scale of So et al (2012). They developed a complete scale for measuring customer engagement for brands, which consists of 5 dimensions. These dimensions are adequate since the composite reliabilities of the dimensions were high (i.e. >.88) in their research. (Coltman et al., 2008). So et al. (2012) also include interactivity as a dimension for engagement. Since in this thesis interactivity is measured separately, it is left out of the engagement measurement. For every dimension only the items with the highest loading are used in this thesis to ensure high response speed. This way the domain of interest, engagement, is still represented.
Interactivity Measure
Two-way communication Perceived possibility of sending information as a user (Johnson et al., 2006)
User control Level of perceived control (Ariely, 2000; Lynch and Ariely, 2000)
Responsiveness Response time (Broekhuizen and Hoffman, 2012; Alba et al., 1997)
Table 6. Measures of engagement. Note: the definitions are all cited from So et al. (2012, p. 311), but
applied on app engagement, instead of brand engagement
Table 7. Sponsors from top-10 employer list.
All items are measured on a seven-point Likert scale, ranging from 1 (definitely not) to 7 (definitely). After the factor analysis, composite scores are constructed based on the unweighted scores of the items.
3.3.3 Brand recall
Brand recall is measured by asking consumers which and how many sponsors they could recall from the event. The measure used for determining the brand awareness is brand recall. This measure was retrieved form literature (Barros et al., 2007; Nicholls et al., 1999; Rowley and Williams, 2008). 3.3.4 Sponsor fit
Sponsor fit was defined as the degree to which the sponsoring company had fit with the event. As the event was open to every company, this study used the sponsor fit, based on goal of the target group of the event: students of the University of Groningen. The goal of recruitment event is for students and companies to meet each other. Students want to orient themselves where they might want to work after graduation. Because of that, companies which are popular for students, as future employees, did fit the event best. First Employers (2013) brings out a list every year which companies are seen as best employees by students. The companies which are in the top 10 of this list were assumed as companies that students expects to sponsor such events, because of the recruitment orientation of the event. Table 7 presents the companies which are in the top 10 of this list and which are sponsor at the event.
Company Ranking on list
Unilever #1 Shell #3 Heineken #6 Rabobank #7 Ahold #9 Engagement Measure
Identification The degree of a user’s perceived belongingness to the app. Enthusiasm The degree of excitement and interest that a user has in the app
Attention The degree of attentiveness, focus, and connection that a user has with the app Absorption The degree to which the user experiences a pleasant state of being fully
Table 8. Largest B2C sponsors 3.3.5 Brand Equity
Brand equity is the effect, brand knowledge has on a consumers response (Keller 2013). In this research it was assumed that the 10 biggest business-to-consumer companies are the companies that have the highest brand equity. Symbaloo (2014) offers an overview of the 50 biggest companies in the Netherlands. Of that the list, the business-to- business companies were filtered out, because consumers encounter business-to-business companies less. Therefore, they are assumed to have a lower brand equity than business-to-consumer companies. Table 8 shows which companies of this top 10 business to consumers companies, is present during the event. These companies are coded as having a high brand equity. Companies that are not on this list, are not considered to be high brand equity companies.
Company Ranking on list
Shell #1 ING #2 Unilever #3 Ahold #4 KPN #9 3.4 Control variables
Visiting of workshops of brands. Consumers visit various activities of different companies. The exposure of these companies to these students was naturally greater than of non-visited companies. Including or excluding the visited companies in the measurement of brand recall could make a difference. For that reason, it was tested whether the difference in measurement of brand recall would make a difference. It turned out the way of measuring brand recall (i.e. including or excluding the visited companies), did not influence the results. The results concerning brand recall, are measured as brand recall excluding visited companies, unless stated differently.
Visiting day. The event takes three days. So it was expected that respondents on the first day had a lower brand recall than on the third day.
Brand sensitivity. Next, it may be the case there is a difference between consumers how sensitive they are for brands (Beaudoin and Lachance, 2009). This may have in influence on recognition and recall on brands. For that reason, consumers are asked how they rate themselves as being sensitive to brands.
Table 9. Correlation table. *P <0.1, ** P<.05
3.5 Correlation table
The next table shows the correlations between brand recall, interactivity, engagement, and usage frequency. Usage frequency is the only control variable shown, since this is the only control variable which is significantly correlated with brand recall. It is remarkable that engagement has no significant correlation with brand recall. Also the significant relations of usage frequency are notable. In the results section, this will be discussed in more detail.
Brand recall Interactivity Engagement Usage
frequency Brand Recall Interactivity .286** Engagement -.058 .268* Usage Frequency .297** .256* .072 3.6 Analysis
This study used SPSS 20 for all its statistic tests.
Independent sample t-test: First an independent sample t-test to test the difference in the number of correctly recalled brands between app users and non-app users.
Regression analysis: A linear regression was done between interactivity and engagement, followed by a linear regression between engagement and brand recall. Next, a mediation test was done, to check whether engagement acts as a full mediator, a partial mediator, or does not act as a mediator. The basis for the mediation analysis was Baron and Kenny (1986). After this, a Sobel test was conducted to assess the significance of the indirect effect. The last step was to use a Bootstrap method increase reliability and also verify the outcome of the Baron and Kenny test and the Sobel test2.
Brand Equity and Sponsor Fit. The direct relationship between brand equity and sponsor fit was measured using crosstabs. The crosstabs showed whether the brand recall was higher for high brand equity and sponsor fit companies. This was measured in percentages, since the total of non-brand equity and non-sponsor fit companies differ from the total of brand equity and sponsor fit companies.
Moderating relationships. The relationship between engagement and brand recall is expected to be
2
Table 10. Companies correctly identified; users vs. non-users
moderated by Sponsor Fit (SF) an Brand Equity (BE). A multiple regression is not possibly since the effects of SF ad BE are measured at a different level, than the effect of engagement. SF and BE are measured at brand level, while consumer is measured at consumer level. Therefore, crosstabs were used to show whether a high level of engagement or a low level of engagement interacts with high or low brand equity or sponsor fit.
4. Results
4.1 Independent sample t-test: recall rates for app-users versus non-app-users
The independent sample t-test reveals that app users, indeed, recall more brands, which confirms H1 (p<.0001). The app users identified on average 10.04 (25.7%) of the companies. For non-app users, this was 5.73 (14.7%). Of the 53 app users, 2 respondents made a type 2 error. Of the 73 non- app users, 4 persons made a type 2 error (i.e. they identified a company which did not sponsor the event).
Average correctly identified Percentage correctly identified
App users 10.04 25.7%
Non-app users 5.73 14.7%
4.2 Linear Regressions
To test H2, a linear regression was performed between interactivity and engagement. Model 1 tests the relationship between interactivity and engagement without control variables. Model 2 tests the relationship between engagement and brand recall. In model 3, usage frequency was added to model 2, because this control variable is the only which is significant (p<.1). In model 4, the other control variables are added to model 3.
Table 11 shows that none of the models show significant results on a 5% significance level. Model 1 shows the results on H2. On a 5% significance level, this hypothesis is rejected. However, on a 10% level significance level, H2 is accepted.
Model 2, model 3, and model 4 show the results for H3, which that a higher engagement with the app leads to a higher brand recall. The outcomes of these models are remarkable. Mainly because none of the models is significant and H3 is rejected in every model.
Brand recall
Interactivity
Engagement
A
B
C’
C
Brand recall
Interactivity
Table 11. Results regressions engagement and brand recall Standard errors are in parentheses * P <0.1, ** P <0.05
Figure 2. Overview relationships mediator analysis
The values are 7.2%, 3%, 5.9%, and 3.8% respectively. This values are very low, which means that the model fit is poor. Therefore, the interpretation of this model requires some caution.
4.3 Mediator analysis
As the conceptual model presented in figure 1 shows, there is a hierarchy of effects between the concepts. To gain more insights about the order of causality and to test the total effect on brand
Model (1) Model (2) Model (3) Model (4) Independent variables Interactivity .451* (.227) - - - Engagement - -.150 (.360) -.206 (.347) -.136 (.355) Controls
Number of times used the app
- - .902*
(.461)
.904* (.465) Number of days attend the
event
- - - -.696
(.673)
I think brands are interesting - - - -.282
(.391) I used the app often in
comparison with the average users of the RD
- - - -.101 (.482) Constant 1.411 10.605 8.611 10.976 R-square .072 .003 .095 .130 Adjusted R-square .054 -0.16 .059 .038 F Observations 53 53 53 53
recall, a mediating analysis was conducted. As with the linear regressions, the mediator analysis is done using multiple models to show the effects of adding control variables. Model 5 is without control variables, to model 6 is frequency of usage is added, and model 7 includes all control variables. Figure 2 shows the different relationships which are tested.
For a mediator effect, 4 conditions must hold. The first condition is that relation A needs to be significant. In the table is shown that this condition is significant at a 10% significance level. The second condition is that relation B is significant. In the second row is shown that this relationship not significant and thus the second condition is not met. Because of this, there is no mediator effect (neither a full mediator nor a partial mediator). Also relation C’ (relation B was also incorporated in the test) is significant, which suggests that the direct is significant and the mediation effect is insignificant. To verify this outcome, a Sobel test is conducted. The Sobel test is not significant, which means that the indirect effect of interactivity on brand recall, insignificant is. Thus, the Sobel test confirms Baron and Kenny. Bootstrapping is a method where the original sample is taken as population and of that population sample with replacement in taken. Than AB is calculated from that sample. The estimates from the Bootstrap is done 5000 times, which makes the sample size large and thus reliable (Preacher and Hayes, 2008). This is also not significant, which confirms that there is not a mediation effect.
It is remarkable that the direct effect between interactivity and brand recall is significant. In model 5, both relations are significant on a 5% significance level.
The model fit is determined by using the R-square for model 5 and the adjusted R-square for model 6 and 7. The fit of the models are 10.2%, 10.4%, and 10.8% respectively. This means that in all models a little over 10% of the variance is explained, which is very low. Therefore, the interpretation of the requires some caution.
Model (5) Model (6) Model (7)
Interactivity on engagement (Relation A) .451* (.227) .449* (.237) .445* (.250) Engagement on Brand Recall (Relation B) -.375 (.358) -.377 (.351) -.329 (.353) Direct Effect (Relation C) 1.240** (.581) .974 (.589) 1.212* (.604) Indirect interactivity on Brand Recall (Relation C’) 1.783** (.634) 1.404** (.576) 1.521** (.599)
Lower Confidence interval -.857 -.826 -.912
Table 12. Mediation analyses Standard errors are in parentheses Bootstrap level: 5000 times * P <0.1, ** P <0.05
Figure 3. Overview models 8,9,10,11
R-square .102 .156 .211 Adjusted R-square .066 .104 .108 AB relationship (A*B) -.169 -.169 -.146 Sobel test (AB relationship) -.169 -.169 -.146 Bootstrap -.169 (.208) -.169 (.217) -.146 (.232)
4.4 Usage frequency test
Table 13. Mediation analyses Standard errors are in parentheses Bootstrap level: 5000 times
Analyses are done using procedures of A.F. Hayes * P <0.1, ** P <0.05
Model (8) Model (9) Model (10) Model (11)
Engagement on Usage Frequency (Relation A) .080 (.156) .010 (.110) Interactivity on Usage Frequency (Relation A) .480* (.064) .147 (.193) Usage Frequency on Brand Recall (Relation B) .700** (.312) .904* (.465) Usage Frequency on Brand Recall (Relation B) .554* (.314) .803* (.450) Total Effect (Relation C) -.150 (.360) -.127 (.365) Total Effect (Relation C) 1.240** (.581) 1.330** (.614) Engagement on Brand Recall (Relation C’) -.206 (.347) -.136 (.355) Interactivity on Brand Recall (Relation C’) .971 (.590) 1.212* (.604) Lower Confidence interval -.130 -.131 -.009 -.155 Upper Confidence interval .464 .266 .968 .754 R-square .095 .130 .136 .196 Adjusted R-square .059 .038 .101 .111 AB relationship (A*B) .056 .009 .266 .118 Sobel test (AB relationship) .056 (.122) .009 .266* (.221) .118 Bootstrap .056 (.139) .009 (098) .266 (.234) .118 (.239)
Table 13 shows that, in model 8 and 9, usage frequency is no mediator. Also the indirect effect is not significant. Next, as expected because of the positive correlations, the relationship between usage frequency and brand recall is in both models strong and significant.
Table 15. Crosstabs Brand Equity and sponsor fit recognition Number of visited companies included in brand recall Table 14. Crosstabs Brand Equity and sponsor fit recognition. Number of visited companies excluded from brand recall
* P <0.1, ** P <0.05, *** P <0.01
4.5 Brand Equity and Sponsor fit
In the conceptual model is hypothesized that brand equity and sponsor fit have direct effect on brand recall, as well as a moderating effect between engagement and brand recall. These constructs are measured on brand level instead on consumer level, therefore other analyses are required. 4.5.1. Direct effect
The direct effect are measured using crosstabs. In the tables the number of recalled companies and the percentages of the maximum possible recall are stated. In table 14 brand recall is calculated by the number of companies that respondents were able to identify, subtracted by the number of companies that respondents have encountered. It is assumed that the companies a respondent encountered (i.e. actively interacted with), are “too easy” to identify and therefore the effect of the app on recall would be biased. However, since there are only 5 companies with high BE and 5 companies with high SF, the number of visited high BE or high SF companies has major influence on the results and therefore this might also be biased. For that reason, recall is also calculated, including the visited companies (table 15).
In table 14 and 15 is shown that the percentage of high BE and high SF companies identified, is significantly higher than companies with low BE and SF (see underlined values). This suggests that high BE and high SF companies are more often recalled than other companies. These results provide sufficient support for hypotheses 4 and 6.
Recall (excluding visited companies) Low BE High BE Identified 436 (24%) 96 (36%) Not identified 1366 (76%) 169 (64%) Chi-square = 17.496*** Low SF High SF Identified 452 (25%) 80 (30%) Not identified 1350 (75% 185 (70%) Chi-square = 3.151*
Recall (including visited companies) Low BE High BE
Table 16 . Crosstabs Engagement level and BE and SF companies recall. Number of visited companies excluded from brand recall.
* P <0.1, ** P <0.05, *** P <0.01
Table 17. Crosstabs Engagement level and brand equity companies recall. Number of visited companies included in brand recall.
* P <0.1, ** P <0.05, *** P <0.01
For the interaction effects, crosstabs are used as well. Again, the number of companies recalled is calculated in at the same way as for the direct effects. Tables 16 and 17 state the brand recall for a high level or a low level of engagement, in combination with high and low BE and SF companies. Thus, the tables show the interaction between interactivity and BE or SF, on brand recall. Both tables show there is not a significant interaction effect. It was expected that a combination between high engagement and high BE, or between engagement and high SF, would lead to the highest recall (i.e. highest percentage). None of the models were significant, thus H5 and H7 are rejected. Table 4.6 provides an overview of the hypotheses and their corresponding results.
Low Brand Equity (Correctly identified)
High brand Equity (Correctly identified)
Engagement Low 232 (23.53%) 55 (37.93%)
Engagement High 204 (25.00%) 41 (34.17%)
Chi-square = .527
Low Sponsor Fit (Correctly identified) High Sponsor Fit (Correctly identified)
Engagement Low 242 (24.54%) 45 (31.03%)
Engagement High 210 (25.74%) 35 (29.17%)
Chi-square = .201
Low Brand Equity (Correctly identified)
High brand Equity (Correctly identified)
Engagement Low 326 (33.06%) 66 (45.52%)
Engagement High 298 (36.52%) 61 (50.83%)
Chi-square = .003
Low Sponsor Fit (Correctly identified)
High Sponsor Fit (Correctly identified)
Engagement Low 333 (33.77%) 59 (40.69%)
Engagement High 298 (36.52%) 61 (50.83%)
Table 18. Overview of hypotheses A = accepted, R = rejected * Only at 10% significance level
4.6 Overview of hypotheses
Hypotheses
H1: In-app sponsoring enhances sponsoring effectiveness. A
H2: A higher perceived interactivity leads to a higher engagement. R H3: A higher engagement with the app leads to a higher recall. R H4: Sponsor fit has a positive relationship with the brand recall of the
sponsoring companies.
A H5: The positive relationship between engagement and recall is
positively moderated by sponsor fit.
R H6: Brand equity has a positive relationship with the brand recall of
the sponsoring companies.
A H7: The positive relationship between engagement and recall is
positively moderated by brand equity.
R
5. Conclusion and discussion
sources of motivation to elaborate was personal relevance of the information (Petty and Cacioppo, 1984; Shernoff and Csikszentmihalyi,2003). The information in the app had no personal relevance to the users and therefore they were not motivated to elaborate on the information. The result of this was an insignificant relationship with brand recall. Furthermore, engagement has overlap with the concept of flow. Flow means that someone is in an ecstatic state, which is the optimal state a person can be in. According to Csikszentmihalyi (2004)there are 7 conditions which appear when a person is in a flow. These are: completely involved in what we are doing, a sense of ecstasy, great inner clarity, knowing what needs to be done, and how well we are doing, knowing that the activity is doable, a sense of serenity, timelessness, intrinsic motivation. These conditions show there is a lot overlap with engagement. To get in a state of flow or engagement is dependent on the amount of challenge a person experiences on that moment and amount of skills they feel they have at that moment (Shernoff, Csikszentmihalyi,2003). Both the challenge and the skills needed to use the app were not high. Therefore, users did not engagement with the app. A last explanation is that it was tested it only one app. Because of that it was not possible to measure different levels of engagement with different apps. This, in combination with low engagement with the app, led to low variation in the engagement levels, which made that the analyses, involving engagement, insignificant.
better. Lastly, past research (Yang et al., 2006) shows a positive relationship between interactivity and brand recognition in video games. It is assumed this is also applicable to this research. It turns out that not engagement, but interactivity and usage frequency are the important constructs when it comes to brand recall.
It was furthermore expected that brand equity (e.g. Keller, 2013; Roy, and Cornwell, 2003) and sponsor fit (e.g. Olson and Thjømøe, 2011), would lead to higher brand recall. As expected, high brand equity companies and high sponsor fit companies were significantly recalled more than low brand equity companies and low sponsor fit companies. For brand equity this can be explained that when consumers have brand knowledge, it is easier to remember the brand and therefore it increases brand recall. Companies who a high sponsor fit are more remember than no sponsor fit companies because it takes less cognitive processing to remember those brand (Jargre et al., 2001), since it is more logical that those brand are present at the event.
Finally, it was expected that there would be interaction effects between engagement and high brand equity and high sponsor fit companies, on brand recall. These results were not significant, meaning that no interaction effect exists. The reason for this might be the same as for all the insignificant relations with engagement: the variation in the level of engagement was low, therefore no significant relationship could be measured.
6. Managerial implications
Another way is to make the app a vital part of the event. You have to use the app in order to have the same experience as non-app users.
7. Limitations and directions for future research
A limitation of the research is that interactivity is measured as perceived interactivity of the consumer. The key features of the app that determine the actual interactivity are used as measure. Another limitation is that in this research the constructs of interactivity and engagement are tested, with the use of single item scales. This was done because of the response speed of the survey. Using multi-item scales for measuring the constructs of interactivity and engagement, can lead to different levels of interactivity and engagement. That way validity and reliability can be enhanced. Further, the app was not very relevant to use, because the added value was low. A consequence of this was that the measured values of engagement had a tendency to be centered around the mean. Direction for future research is that different apps with different levels of interactivity of the app should be used. That way different level of interactivity and engagement can be tested. Then interactivity should be measured using features of the app and interactivity and engagement should be measured using multi-item scales.
A limitation concerning the moderator analysis research is that there is overlap between the sponsor fit companies and brand equity companies. Both categories contained of 5 companies, of which three companies were in both categories. In future research, there should be no overlap in brand equity and sponsor fit companies.
Furthermore, a limitation is that endogeneity might be possible. In this thesis the difference between app users and non-app users was measured. The respondents who downloaded the app, did choose to download the app (i.e. they were self-selected). It might be the case that the people which downloaded the app, have certain traits, which the non-app users does not have. Therefore it is the case of omitted selection. The question is what the outcome would have been if the non-app users did download the app. Because this is unclear, there might be an omitted cause (other than app usage) which led to a higher brand recall for app users. In future research endogeneity needs to be taken in to account. It could be solved if respondents are randomly assigned to use the app (Antonakis, 2011).
a device and brand recall. This twist can have important implications. In future research the combination of interactivity, usage frequency and sponsoring effectiveness should be a focus.
Lastly, a limitation was that the sample was rather small and, although it covered the population well, it can not necessarily be generalized every event in every situation. The sample (and population) only consisted of students and therefore the results can be different for consumers, other than students.
References
ABIresearch (2014). 44 Billion Mobile App Downloads by 2016. Accessed on October 14, 2014, from https://www.abiresearch.com/press/44-billion-mobile-app-downloads-by-2016.
Alba, J., Lynch, J., Weitz, B., Janiszewski, C., Lutz, R., Sawyer, A., & Wood, S. (1997). Interactive home shopping: consumer, retailer, and manufacturer incentives to participate in electronic
marketplaces. The Journal of Marketing, 38-53.
Antonakis, J. (2011). Endogeneity: An Inconvenient truth. [Movie]. Accessed on January, 5, 2015, from: https://www.youtube.com/watch?v=dLuTjoYmfXs.
Ariely, D. (2000). Controlling the information flow: Effects on consumers’ decision making and preferences. Journal of Consumer Research, 27(2), 233-248.
Baron, R. M., & Kenny, D. A. (1986). The moderator–mediator variable distinction in social
psychological research: Conceptual, strategic, and statistical considerations. Journal of personality and social psychology, 51(6), 1173.
Barros, C. P., Barros, C. D., Santos, A., & Chadwick, S. (2007). Sponsorship brand recall at the Euro 2004 soccer tournament. Sport Marketing Quarterly,16(3), 161-170.
Beaudoin, P., & Lachance, M. J. (2006). Determinants of adolescents' brand sensitivity to clothing. Family and Consumer Sciences Research Journal,34(4), 312-331.
Bosomworth, D. (2014). Statistics on mobile usage and adoption to inform your mobile marketing strategy. Accessed on September 28, 2014, from:
http://www.smartinsights.com/mobile-marketing/mobile-marketing-analytics/mobile-marketing-statistics.
Bretz, R., & Schmidbauer, M. (1983). Media for interactive communication. Beverly Hills, CA: Sage. Brodie, R. J., Ilic, A., Juric, B., & Hollebeek, L. (2013). Consumer engagement in a virtual brand community: An exploratory analysis. Journal of Business Research, 66(1), 105-114.
Broekhuizen, T., & Hoffmann, A. (2012). Interactivity Perceptions and Online Newspaper Preference. Journal of Interactive Advertising, 12(2), 29-43.
Bucy, E. P., & Tao, C. C. (2007). The mediated moderation model of interactivity. Media Psychology, 9(3), 647-672.
Burgoon, J. K., Buller, D. B., & Floyd, K. (2001). Does participation affect deception success?. Human Communication Research, 27(4), 503-534.
Calder, B. J., Malthouse, E. C., & Schaedel, U. (2009). An experimental study of the relationship between online engagement and advertising effectiveness. Journal of Interactive Marketing, 23(4), 321-331.
Cornwell, T. B., Weeks, C. S., & Roy, D. P. (2005). Sponsorship-linked marketing: opening the black box. Journal of Advertising, 34(2), 21-42.
Coltman, T., Devinney, T. M., Midgley, D. F., & Venaik, S. (2008). Formative versus reflective measurement models: Two applications of formative measurement. Journal of Business Research, 61(12), 1250-1262.
Csikszentmihalyi, M. (2004). Flow, the secret to happiness. [Movie]. Accessed on January, 9, 2015, from http://www.ted.com/talks/mihaly_csikszentmihalyi_on_flow?language=nl.
Downes, E. J., & McMillan, S. J. (2000). Defining interactivity a qualitative identification of key dimensions. new media & society, 2(2), 157-179.
Emarketer (2014). Smartphone users worldwide will total 1,75 billion in 2014. Accessed on
September 18, 2014, from
http://www.emarketer.com/Article/Smartphone-Users-Worldwide-Will-Total-175-Billion-2014/1010536.
Eroglu, S. A., Machleit, K. A., & Davis, L. M. (2001). Atmospheric qualities of online retailing: a conceptual model and implications. Journal of Business research, 54(2), 177-184.
Exact Target (2014). 2014 Mobile Behavior Rapport. Accessed on October 13, 2014, from
http://www.exacttarget.com/sites/exacttarget/files/deliverables/etmc-2014mobilebehaviorreport.pdf.
Gartner (2010). Newsroom, Gartner Highlights Key Predictions for IT Organizations and Users in 2010 and Beyond. Accessed on October 13, 2014, from:
http://www.gartner.com/newsroom/id/1278413.
Grohs, R., Wagner, U., & Vsetecka, S. (2004). Assessing the effectiveness of sport sponsorships–an empirical examination. Schmalenbach Business Review, 56(2), 119-138.
Gwinner, K. P., Larson, B. V., & Swanson, S. R. (2009). Image transfer in corporate event sponsorship: assessing the impact of team identification and event-sponsor fit. International Journal of
Management and Marketing Research, 2(1), 1-15.
Hair, J. F., Anderson, R. E., & Tatham, R. L. (1987). Multivariate data analysis 2nd ed. New York: MacMillian.
Higgins, E. T. (2006). Value from hedonic experience and engagement. Psychological review, 113(3), 439.
Jagre, E., Watson, J. J., & Watson, J. G. (2001). Sponsorship and congruity theory: a theoretical framework for explaining consumer attitude and recall of event sponsorship. Advances in Consumer Research, 28, 439-445.
Johar, G. V. Michel Tuan Pham og Kirk L. Wakefield (2006). How Event Sponsors Are Really Identified: A (Baseball) Field Analysis. Journal of Advertising Research vol. 46 (juni), 183-198.
Johnson, G. J., Bruner II, G. C., & Kumar, A. (2006). Interactivity and its facets revisited: Theory and empirical test. Journal of Advertising, 35(4), 35-52.
Keller, K.L. (2013) Strategic Brand Management – Building, Measuring, and Managing Brand Equity. Harlow: Pearson Education Inc.
Kim, J. W. (2010). The worth of sport event sponsorship: an event study. Journal of Management and Marketing Research, 5(1), 1-14.
Kim, E., Lin, J. S., & Sung, Y. (2013). To app or not to app: Engaging consumers via branded mobile apps. Journal of Interactive Advertising, 13(1), 53-65.
Lacey, R., Sneath, J. Z., Finney, Z. R., & Close, A. G. (2007). The impact of repeat attendance on event sponsorship effects. Journal of Marketing Communications, 13(4), 243-255.
Lombard, M., & Ditton, T. (1997). At the heart of it all: The concept of presence. Journal of Computer‐ Mediated Communication, 3(2), 0-0.
Lombard, M., & Snyder-Duch, J. (2001). Interactive advertising and presence: a framework. Journal of Interactive Advertising, 1(2), 56-65.
Lynch Jr, J. G., & Ariely, D. (2000). Wine online: Search costs affect competition on price, quality, and distribution. Marketing Science, 19(1), 83-103.
MacKenzie, S. B., & Podsakoff, P. M. (2012). Common method bias in marketing: causes, mechanisms, and procedural remedies. Journal of Retailing, 88(4), 542-555.
Malhotra, M. K., & Grover, V. (1998). An assessment of survey research in POM: from constructs to theory. Journal of operations management, 16(4), 407-425.
Malhotra, N.K. (2010). Marketing Research. New Jersey: Pearson Prentice Hall.
McMillan, S. J., & Hwang, J. S. (2002). Measures of perceived interactivity: An exploration of the role of direction of communication, user control, and time in shaping perceptions of interactivity. Journal of Advertising, 29-42.
Mehrabian, A., & Russell, J. A. (1974). An approach to environmental psychology. Cambridge: MIT Press.
Memorygroup BV (2013). De populairste werkgevers om je carriere te starten. Accessed on November 15, 2014, from http://www.firstemployers.nl/first-employers-2013/doelgroepen. Mollen, A., & Wilson, H. (2010). Engagement, telepresence and interactivity in online consumer experience: Reconciling scholastic and managerial perspectives. Journal of business research, 63(9), 919-925.
Nicolas Carlson (2014) Why Google is slowing. Accessed on October 21, 2014, from: http://www.businessinsider.com/why-google-is-slowing-2014-10.
Novak, T. P., Hoffman, D. L., & Yung, Y. F. (2000). Measuring the customer experience in online environments: A structural modeling approach. Marketing science, 19(1), 22-42.
Olson, E. L., & Thjømøe, H. M. (2009). Sponsorship effect metric: assessing the financial value of sponsoring by comparisons to television advertising.Journal of the Academy of Marketing Science, 37(4), 504-515.
Olson, E. L., & Thjømøe, H. M. (2011). Explaining and articulating the fit construct in sponsorship. Journal of Advertising, 40(1), 57-70.
Cacioppo, J. T., & Petty, R. E. (1984). The elaboration likelihood model of persuasion. Advances in Consumer Research, 11(1), 673-675.
Pham, M.T. (1991) The evaluation of sponsorship effectiveness: a model and some methodological considerations. Gestion 2000. 4, P. 47-63.
Preacher, K. J., & Hayes, A. F. (2008). Asymptotic and resampling strategies for assessing and comparing indirect effects in multiple mediator models.Behavior research methods, 40(3), 879-891. Purcell, K., Entner, R., & Henderson, N. (2010). The rise of apps culture. Pew Internet & American Life Project, 14.
Roy, D. P., & Cornwell, T. B. (2003). Brand equity’s influence on responses to event sponsorships. Journal of Product & Brand Management, 12(6), 377-393.
Rowley, J., & Williams, C. (2008). The impact of brand sponsorship of music festivals. Marketing Intelligence & Planning, 26(7), 781-792.
Sicilia, M., Ruiz, S., & Munuera, J. L. (2005). Effects of interactivity in a web site: the moderating effect of need for cognition. Journal of Advertising, 34(3), 31-44.
Simmons, C. J., & Becker-Olsen, K. L. (2006). Achieving marketing objectives through social sponsorships. Journal of Marketing, 70(4), 154-169.
So, K. K. F., King, C., & Sparks, B. (2012). Customer engagement with tourism brands: Scale development and validation. Journal of Hospitality & Tourism Research, 38(4): 304-329.
Song, J. H., & Zinkhan, G. M. (2008). Determinants of perceived web site interactivity. Journal of Marketing, 72(2), 99-113.
Symbaloo (2014). 50 grootste Nederlandse bedrijven. Accessed on November 15, 2014, from
http://www.symbaloo.com/mix/4465-50-grootste-Nederlandse-bedrijven.
