The effect of consumer brand experience and email colour on
email marketing effectiveness
Master ThesisMSc in Business Administration – Digital Business Supervisor: dhr. Gijs Overgoor MSc
Student: Karolina Rehakova – student number 11759887
Amsterdam 21/06/2017
Statement of Originality
This thesis was written by the student Karolina Rehakova who declares to take responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Acknowledgements
First, I would like to express my gratitude towards my supervisor Gijs Overgoor for the continuous support, his patience and his valuable feedback throughout the whole process. Without his guidance this study would not have been completed and improved. Furthermore, I owe my warmest gratitude to my partner Peter and my parents for providing me with support, advice and continuous encouragement in every chapter of my life. Finally, I would like to dedicate this thesis to my grandparents, who have always seen the best in me, and to my brother, to inspire him to never give up on his education, to challenge himself and to continue developing his creativity and intellect by pursuing graduate studies in the field of his interest.
Abstract
Nowadays’ era of consumer empowerment and information overload is challenging marketers to design effective online marketing communications. Despite the rise of social media, email marketing is still amongst the most popular marketing channels and is favoured for its customizability, measurability and superior returns. However, the average click-through-rate remains below 5%. This suggests that there is still a room to enhance email marketing effectiveness by improving its features and strategies. Therefore, this research investigates two yet unexplored variables, email colour and consumer brand experience level, and their effect on consumer’s propensity to click on a link within a marketing email. Based on previous research, it is assumed that email marketing newsletters performance will be higher for recipients with higher brand experience level and for recipients receiving emails containing cool colours. Using the data of an international consultancy service provider, a factorial treatment experiment was performed, followed by a binary regression analysis of almost 6000 responses of email newsletter recipients. The results have discovered a negative relationship between consumer brand experience and his propensity to click on a link in an email. For the second examined hypothesis, no significant correlation was identified between the colour of an email and the number of clicks generated. Although the hypotheses were not supported, this study provides novel insights into email marketing effectiveness that can enhance both managerial decisions and future research.
Key words: Email marketing effectiveness, consumer brand experience, email colour, call-to-action button colour, email click-to-open rate, email click-through-rate
INDEX 1. INTRODUCTION ... 8 2. LITERATURE REVIEW ... 12 2.1 BRAND EXPERIENCE ... 12 2.2 COLOUR ... 14 2.2.1 Structure of colour ... 14
2.2.2 Dual processing of colour ... 14
2.2.3 Colour in online marketing ... 16
2.3 EMAIL EFFECTIVENESS ... 17
2.3.1 Relevance of email marketing ... 18
2.3.2 Trusting beliefs in email marketing ... 18
2.4 CONCEPTUAL MODEL ... 20
3. METHODOLOGY ... 20
3.1 SAMPLE AND DATA COLLECTION ... 20
3.2 PROCEDURE ... 21
3.3 VARIABLES AND MEASURES ... 22
3.2.1 Dependent Variable ... 22 3.2.2 Independent Variables... 23 4. RESULTS ... 25 4.1 RECIPIENTS ... 25 4.2 DATA PREPARATION... 25 4.3 HYPOTHESIS TESTING ... 26 Hypothesis 1 ... 28 Hypothesis 2 ... 29
5. DISCUSSION... 29
5.1 GENERAL DISCUSSION ... 29
5.2 THEORETICAL IMPLICATIONS ... 33
5.3 MANAGERIAL IMPLICATIONS ... 33
5.4 LIMITATIONS AND FUTURE RESEARCH ... 34
6. REFERENCES ... 37
1. Introduction
In the era of consumer empowerment and information overload, there is a pressing need to understand the determinants of online marketing effectiveness (Hartemo, 2016; Edwards, 2016; Eppler & Mengis, 2004). Nowadays’ online marketing comes in a variety of forms accessed via different platforms including search engines, social media, websites or the long-established email inboxes. Interestingly, despite the rise of social media as the new high-reach interactive communication tool, email remains to be one of the most effective direct marketing channels with an average ROI of £38 for each £1 spent, and more than 17% higher average order rate than Facebook or Pinterest (Van Rijn, 2015; Aufreiter et al., 2014). Besides its exceptional returns, marketers favour email for its easy integration with corporate customer relationship management systems, and resulting customizability (Sharma & Seth, 2004), measurability (Jenkins, 2009) and ability to cultivate brand loyalty (Merisavo & Raulas, 2004). These information point to the fact that email is still a relevant player in the field of marketing communication and its success factors are worth investigating.
A standard measure of email marketing effectiveness is the rate of recipients who click on hyperlinks provided in its content (Ellis-Chadwick & Doherty, 2012; Merisavo & Raulas, 2004; Lohtia, Donthu & Hershberger, 2003). While the click-thru rate (CTR) of emails has increased since the rise of opt-in and content personalization strategies (Statista, 2018a), the average CTR remains below 5% (IBM Marketing cloud, 2017). This suggests that there is still room for improvement in understanding the strategies and features that have the potential to increase email marketing effectiveness. One of the factors influencing consumers’ attitudes and behaviour is their level of brand experience. Park & Stoel (2005) found a significant positive effect of previous brand experience on purchase intention, and Campbell & Keller (2003) found it increases advertising impact. It could therefore be expected that higher levels of brand experience would also benefit the performance of online marketing communications. Yet, no
research has addressed the particular relationship between brand experience and propensity to click on hyperlinks within marketing emails. Another element whose effect on email marketing effectiveness is still rather unknown is its graphic design. While there is a number of online sources that offer advice on how to design effective emails, academic research on this topic is quite scarce. Only one study tested the effect of email colour on CTR, however, the email design used in this experiment is no longer compatible with current corporate email marketing trends and practices (Zviran et al., 2006). Hence, until this day, hardly any guidance was provided on how an effective email should look and which colours improve the likelihood of achieving the goals behind email marketing communications.
Understanding the best-practices in email marketing design and strategy is a pressing issue because, while the effectiveness and popularity of email is good news for marketers, it ultimately points toward an increase in promotional email generated and intensity of competition within consumers’ mailboxes (Naragon, 2017; Edwards, 2016). This highlights the need for companies to make deliberate data driven decisions about the content, design and scheduling of their email promotions. While past research provides a good picture of the best-practices in content and timing of email marketing (Edwards, 2016; Wilson et al., 2015; Hemersma, 2013; Ellis-Chadwick & Doherty, 2012; Eisinga, 2017; Merisavo & Raulas, 2004; Ansari & Mela, 2003; Marinova, Murphy & Massey, 2002), little is known about the impact of recipient’s brand experience level and email’s graphic design on email marketing effectiveness (Hsieh et al., 2018; Hartemo, 2016; Demangeot & Broderick, 2010; Labrecque et al., 2013).
Brand experience has long been a focus of marketing research, which continuously concludes that it plays a crucial role in building purchase intention (Park & Stoel, 2005), brand familiarity (Gefen, Karahanna & Straub, 2003; Gefen, 2000), satisfaction and ultimately brand trust (Sahin et al., 2011; Ha & Perks, 2005). A survey-based study by Ha & Perks (2005) revealed there is a direct positive effect of brand experience on brand trust, which in turn was
found to contribute to intention to comply with an appeal made in an email (Wilson et al., 2015). Although brand experience substantially benefits consumer attitudes towards brands in the online environment, its effect on their behavioural responses to online advertising remains unascertained. Given the ease of online advertising personalization, knowing the effect of one’s brand experience on his propensity to click on links in marketing emails could help companies develop more tailored marketing communications and evolve their online marketing strategies (Haugtvedt et al., 2005; Mittal & Kamakura, 2001). Hence, the relationship between brand experience and propensity to click on links in marketing emails is worthwhile examining.
One of the design elements with naturally arousing ability to activate potential consumers is colour (Shi, 2013). Colour was proven to affect the responsiveness to banner advertisements (Sokolik, Magee & Ivory, 2014) and social media posts (Constine, 2016; Woolaston, 2013; Sabate et al., 2014). Past studies established colour as an important variable in website design, affecting consumer perception of loading time (Gorn et al., 2004) and urge to make impulsive purchases (Parboteeah et al., 2009; Crowley, 1993; Bagchi & Cheema 2013). Furthermore, several studies show that website’s colour has the ability to increase its perceived trustworthiness (Karvonen, 2000; Fogg, Lee & Marshall, 2002; Fogg et al., 2003; Yang, Hu & Chen, 2005; Kim & Moon; 1998), which is the key determinant of purchase intention in online environment (Urban, Amyx & Lorenzon, 2009; Yoon, 2002; Schlosser, White & Lloyd, 2005). Considering the fact that colour clearly plays a role in consumer’s perceptions of companies in the online environment (Hsieh et al., 2018; Labrecque & Milne, 2012; Demangeot & Broderick, 2010; Crowley, 1993), it is surprising how little is known about its role in the context of email marketing. Chittenden & Rettie (2002) conducted one of the first studies implying that more colourful and attractive emails generate greater click-through rate, and Zviran, Te’eni, & Gross (2006) argue that colour has the ability to set the right mood for responding positively to a message or request. However, no specific guidelines have yet been provided in this area. Thus,
there is a room for investigation addressing the managerial need for understanding which colours are best utilized in email marketing campaigns.
In pursuance of the findings by Chittenden & Rettie (2002) and Zviran et al. (2006) this study aims to fill the gap in current literature by analysing the link between the colour used in email marketing and the engagement generated in form of click-to-open rate (CTOR). Adopting key findings from website design literature, this paper will investigate which colours lead to the highest CTOR for email marketing messages. In addition to colour, the effect of recipient’s brand experience on his propensity to click on a link in a marketing email will be examined. Such information could be of great relevance for marketers seeking to understand how to increase consumer engagement in form of clicks on the links provided within email marketing messages.
For these reasons, following research question is formulated:
What is the effect of different levels of recipient’s brand experience, and colour used in email design, on the click-to-open rate of marketing newsletters?
In order to answer this question a 4 (colour scheme: red, orange, blue and green) x 5 (brand experience) factorial treatment experiment was performed, using data from an international consultancy service provider. This study follows the rationale of Mochon et al. (2017) identifying multivariate testing as an effective, and important, way of understanding and maximizing consumer response to firm-initiated promotions.
Finding the answer to this question has several theoretical and managerial contributions. First, this study contributes to academic literature by shedding light on the impact of recipients’ experience with the sender brand on his responsiveness to its email marketing communications. Second, it investigates to what extent does colour affect CTOR in nowadays’ email marketing.
In addition, this study ascertains whether discoveries related to e-commerce and web-design effectiveness apply to email marketing, which is a similar online environment with pull-marketing nature (Thompson & Wassmuth, 1999). Accordingly, this study contributes to managerial practice by providing insight into the influence of recipients’ level of brand experience and email colour on brand’s email marketing effectiveness. Its results will advance managers’ ability to make predictions of the effect of emails’ design and target audience on performance, hence make better informed decisions and plans.
This research will be structured as follows: First, the relevant literature on the concept of brand experience will be introduced. Second, the concept and psychology of colour will be explained, followed by its effects on consumer attitudes and behaviours in marketing setting. Further, the role of brand trust in generating consumer response to email marketing will be discussed. Lastly, the theoretical framework will be presented.
2. Literature review
To shed light on the empirical work that has addressed the relationships investigated in this research, this chapter provides an overview of the relevant literature touching upon the concepts of brand experience and colour hues with respect to their effect on human perceptions and behaviours. First, the concept of brand experience is introduced and its relationship to other brand-related constructs is investigated. Subsequently, colour and its effect on human psychology, followed by its role in online marketing is described. Finally, recently conducted research on email effectiveness is reviewed and a theoretical framework is presented.
2.1 Brand experience
Experience has been described as a degree of familiarity with a certain subject area, obtained through a certain type of exposure (Braunsberger & Munch, 1998). Brakus, Schmitt & Zarantonello (2009) claim brand experience comes from internal as well as behavioural
consumer responses evoked by brand-related stimuli that are part of a brand’s design, identity, communications and environment. Thus, brand experience is formed by accumulation of relevant brand information throughout ad exposure, information search or product usage (Ha & Perks, 2005; Baker et al., 1986).
Brand experience has long been of great interest for researchers and marketers because it is closely related to a number of desirable brand-related constructs. Ha & Perks (2005) investigated the role of brand experience in building brand satisfaction and brand trust, concluding a significant direct effect on each. This is in line with Gefen’s (2000) results of an empirical experiment on commerce, which show that people with greater experience with e-vendor have generally higher trust in e-e-vendor. Furthermore, brand experience is the key building block of brand familiarity (Ha & Perks, 2005); a construct described as an understanding based on previous interactions and experiences (Gefen, 2000; Gefen, Karahanna & Straub, 2003), developed over time with accumulation of brand-related direct or indirect experiences (Park & Stoel, 2005; Alba & Hutchinson, 1987), or accumulation of time that a consumer spent with processing information about the brand (Baker et al., 1986). There is a strong, consistent evidence that sufficient level of brand familiarity mediates brand preference in case of increased exposure to the brand (Obermiller, 1985; Stang, 1975; Moreland & Zajonc, 1977), that it contributes to trust in a brand (Gefen, Karahanna & Straub, 2003) and increases the impact of online advertising (Campbell & Keller, 2003).
While past research provides a good understanding of the effect of brand experience on consumer attitudes, the research on its effect on consumer behaviour is rather limited. Yet, research has shown that consumer brand attitudes serve as predictors of behavioural outcomes such as information search about the brand (Lee et al., 2015), behavioural brand loyalty (Romaniuk & Nenycz-Thiel, 2013), or purchase loyalty (Chaudhuri & Holbrook, 2001; Smith & Swinyard, 1983). In light of these findings, this research aims to investigate the effect of
positive attitudes generated by brand experience on consumer’s behavioural reactions to online advertising; clicks.
2.2 Colour
2.2.1 Structure of colour
Colour theorists widely agree on three independent properties of colour: hue, saturation, and brightness (Thompson, Palacios & Varela, 1992). In order to understand the potential marketing implications of colour, we first provide a definition of each. For better visualization, Appendix A shows the basic hue combinations, brightness levels and saturation levels based on the Munsell (1912) colour system. First, hue corresponds to the wavelength of the light, representing a colour’s value of red, yellow, blue and green pigment (Valdez & Mehrabian, 1994; Zviran et al. 2006). Based on hue, colours may be divided into warm, cool or neutral (Crowley, 1993). Long wavelength colours such as red, orange and yellow, represent the warm group. Cool colours including green, blue and violet, comprise of short wavelength. Neutral colour white is reflecting a combination of all wavelengths, while neutral colour black reflects zero wavelength as it absorbs all visible light. Second, saturation is the intensity or vividness of a colour. High-saturation colours are rich and striking whereas low-saturation colours are greyer and dull (Valdez & Mehrabian, 1994; Hsieh et al. 2018). Third, brightness is the black-to-white quality of colour representing its degree lightness or darkness (Valdez & Mehrabian, 1994; Gorn et al., 1997). It is a continuous dimension where low-brightness colours appear “darkish” due to black mixed into its pigment (Hsieh et al. 2018), and high-brightness colours appear “whitish” and pastel-like in appearance (Gorn et al., 1997).
2.2.2 Dual processing of colour
Colour has the ability to affect people’s emotions, associations and subsequent behaviour (Hoadley, 1990; Valdez & Mehrabian, 1994; Hsieh et al., 2018). The results of vast research from the department of psychology infer that human beings are affected by colour by
two distinct mechanisms; physiological and psychological (Sokolik, Magee & Ivory, 2014; Crowley, 1993). This inherently leads to dual processing of colour.
First, the physiological arousal dimension is dependent on the wavelength of light, which is affecting the level of consumer’s brain activation (Shi, 2013). Holding a U-shaped relationship, most arousal and attention are elicited by hues on the extremes of the wavelength spectra, i.e. violet and red (Figure 1 by Crowley, 1993; Bagchi & Cheema 2013; Magee, 2012).
Second, besides creating an automatic and natural neurobiological response, colours can also trigger certain feelings and moods (Valdez & Mehrabian, 1994). This leads us to the psychological evaluative dimension of colour processing, where colours induce the feeling of pleasure (Crowley, 1993), and contribute to one’s attitude based on his individual and cultural associations (Shi, 2013). As illustrated in Figure 1, this relationship is positively linear so that the cool colours are consistently preferred over warm colours (Crowley, 1993). This is underscored by shared associations established from early stages of life, when human brain associates red with dangers and mistakes while blue, the colour of sky and ocean, is linked with openness and freedom (Elliot et al., 2007; Kaya & Epps, 2004).
As a result, we are facing an interesting paradox, when subjects are physically attracted to warm colours, yet feel that cool-colour environments are generally more pleasant (Shi, 2013; Silver & McCulley, 1988; Bagchi & Cheema, 2013; Magee, 2012; Bellizzi & Hite, 1992). The way this phenomenon is reflected in the field of marketing is presented in the following section.
Figure 1: Two-dimensions of response to colour by Crowley
2.2.3 Colour in online marketing
Due to the low cost and ease of colour adjustment in digital environments, colour has become an important and effective tool of shaping consumer perceptions in online advertising (Labrecque & Milne, 2012). Yet, marketers must be aware of the interplay of evaluative and arousal dimensions of colour when designing their marketing communication materials. They should always consider the emotions and associations they aim to invoke when selecting their colour palette (Shi, 2013; Aghidae & Honari, 2014).
Based on presented academic research, warm colours in advertising were identified as more likely to stimulate physiological excitement and increase attention (Aghidae & Honari, 2014), effectiveness in processing and remembering information (Kroeber-Riel, 1979) and likelihood of impulsive reactions (Crowley, 1993; Bagchi & Cheema, 2013). Even though most of the aforementioned research focused on offline environments, study by Sokolik, Magee and Ivory (2014) provides evidence that these findings are equally relevant for digital marketers. The authors compared the click-through-rate (CTR) of more than 1.5 million banner ad impressions and concluded that red colour schemes are substantially more likely to generate clicks compared to ads with blue colour schemes.
On the contrary, cool colours were proven to produce greater feelings of relaxation and pleasure than warm colours (Jacobs & Suess, 1975; Gorn et al, 2004). In brick-and-mortar stores, cool colours lead to more favourable product evaluations (Middlestadt, 1990), stronger inclination to browse and increased purchase frequency (Bellizzi & Hite, 1992). Due to these qualities, blue is also favoured in e-commerce context. Blue backgrounds in online stores evoke higher pleasure (Coursaris, Swierenga & Watrall, 2008), greater sense of trustworthiness (Fogg, Lee & Marshall, 2002; Kim & Moon, 1998) and make consumers more likely to consider high price as an indicator of high quality and value, in contrast to red backgrounds, which make
people more high-price averse (Hsieh et al., 2018; Mehta & Zhu, 2009; Babin, Hardesty & Suter, 2003).
When it comes to email design, there is only one study addressing the effect of different colours on email’s effectiveness in generating clicks. Zviran et al. (2006) manipulated the background colour of an email and concluded that the highest CTR was generated for yellow, green and blue colour, respectively. The pink background, which was also tested by the authors, had the worst results. However, this experiment is not fully aligned with nowadays’ commercial email marketing practices. Companies prefer to use white background due to simplicity (Karvonen, 2000), better readability (Hall & Hanna, 2004) and consequent faster information processing (Ling & van Schaik, 2002). Therefore, it is important to test colour effectiveness in an experiment design adjusted to this marketplace reality. This study aims to fill the gap in current literature by analysing the effect of heading and call-to-action colour in an email with a white background.
2.3 Email effectiveness
Looking at email marketing from the consumer perspective, Wilson et al. (2015) have recently provided an empirically tested comprehensive framework of four main determinants of email’s effectiveness. The authors identified four cognitive factors responsible for 70% of person’s intention to comply with an appeal made in an email, which in commercial emails often refers to clicking on a call-to-action button. These determinants are message involvement (relevance), trusting beliefs (in the sender), perception of benefits offered, and perception of effort needed to comply with the appeal. This research uses Wilson et al. (2015) framework as the foundation of its hypotheses, testing the framework’s application on not only the intention but rather the actual compliance with an appeal in an email. Specifically, this research examines to what extent does the level of trusting beliefs affect email recipients’ propensity to click on a link in a marketing email.
2.3.1 Relevance of email marketing
Wilson’s study was performed on traditional direct email marketing, where the relevance of a message is out of marketer’s control. This study will analyse whether Wilson’s findings apply to permission-based (opt-in) email marketing, where the starting point of communication is inevitably different.
The idea of permission-based marketing, when marketers seek consumer’s consent with receiving promotional messages in advance, was first introduced by Godin’s (1999). Following the author’s arguments, in return for one’s consent companies are able to deliver promotional messages that are “anticipated, personalized and relevant”. By agreeing to receive firm’s promotional messages, a person signifies that he finds its content personally relevant and allows the firm to process his personal details and track his past email interactions. Therefore, message relevance in the context of permission-based emails, hence also this research, is a constant variable.
2.3.2 Trusting beliefs in email marketing
Consumer trust in online environment is a multifaceted concept affected by factors beyond the content of the marketing message itself (Wilson et al., 2015; Gill et al., 2005; Gefen et al., 2003). Trusting beliefs essentially represent consumer’s beliefs in honesty of the other party, such as the brand sending the message (Wilson et al., 2015). Previously presented research supports the idea that brand trust is facilitated by accumulating relevant brand experience (Ha & Perks, 2005; Gefen, Karahanna & Straub, 2003; Gefen, 2000). Ha & Perks (2005) investigated the role of brand experience in building brand satisfaction and brand trust, concluding a strong direct effect on both. Based on the aforementioned studies the following sequence of effects is proposed; the higher the receiver’s level of experience with the sender, the higher are his trusting beliefs (Ha & Perks, 2005; Gefen, Karahanna & Straub, 2003) and the higher is his intention to comply with the appeal in the email (Wilson et al., 2015). Given
that the appeal in the email employed in this experiment is clicking on a call-to-action button, we develop the following hypothesis:
H1: There is a positive relationship between one’s level of brand experience and his propensity to click on a link in an email newsletter.
The effect of graphic design on brand trust was not extensively investigated, however, there is empirical evidence suggesting its ability to positively influence one’s trusting beliefs (Wang & Emurian, 2005; Crowley, 1993). Findings from e-commerce website design literature identify cool hues to increase the perceived trustworthiness of a seller (Fogg, Lee & Marshall, 2002; Kim & Moon, 1998). Just like websites, email differs from other ad formats in its "pull" nature of access, meaning that the access is controlled by consumer, while banner ads are "pushed" onto consumer unwantedly and without his consent (Thompson & Wassmuth, 1999). Moreover, the content of a website and email requires much more complex information processing than the few sentences found in a banner or pop-up ad (Rodgers & Thorson, 2000). Based on presented similarities in content complexity and access nature, this study presumes that findings from web-design context are, to certain extent, relevant for email marketing. Applying the findings from web-design context; cool hues induce trust in the brand (Fogg, Lee & Marshall, 2002; Kim & Moon, 1998) and higher trusting beliefs increase the magnitude of consumer’s intention to comply with an appeal (Wilson et al., 2015), namely to click on the call-to-action button.
H2a: Cool hues used in email’s heading and call-to-action button lead to higher click-to-open rate for email newsletters.
H2b: Warm hues used in email’s heading and call-to-action button lead to lower click-to-open rate for email newsletters.
2.4 Conceptual Model
Figure 2: Conceptual model
3. Methodology
In this section, the empirical part of the study will be provided. First, the collected sample will be described. Second, the research design and variable measures and measurements scales will be explained.
3.1 Sample and data collection
The population of interest of this study were Western European email users subscribed to, and receiving, any type of marketing emails or commercial newsletters. A sample representative of the underlying population was derived from the contact database of an international consultancy service provider company. The contacts were filtered based on location, and this experiment only considered contacts located in one of the following countries; Luxembourg, Denmark, Norway, The Netherlands, Sweden, Finland, Germany, United Kingdom and Belgium. These countries were chosen as the empirical setting for two main reasons. First, they have the highest share of individuals who send and receive emails (Statista, 2018b). Second, the marketing spending in these countries generates more than 60% of total
Email colour (blue, green, orange, red)
Brand experience (1-low, 5-high) Email marketing effectiveness (Click) H2a+b H1
marketing spending across Europe (Knapp & Chen, 2017). These statistics suggest that the email marketing awareness in these countries is advanced, and their audience has broad experience with email marketing practices. As a result, 19 642 contacts were involved in the online field experiment.
The research was administered using a contact database of an international B2B consultancy service provider, which was appropriate for its permission-based nature, large amount and suitable nationality mix of the contacts. Although the data necessary to this study are not typically available to the public or to academic researchers, regional marketing managers of this company agreed to make the click data available for the purposes of this research. After filtering the email recipients by country, the resulting sample size was 19 642 recipients. The blue version of the email was delivered to 3972 people, green to 3983 people, orange to 3907 people and the red email was received by 3847 people. The control group that has received the email version with black heading and white call-to-action button consisted of 3933 people.
3.2 Procedure
To answer the research question and to test the developed conceptual model, a two-way factorial treatment experiment was conducted. The experiment was administered online via an email marketing automation software used by the involved company. The experiment tested four treatment groups (blue, green, orange and red email colour) and one control group (white) on five groups of contacts, assigned randomly using the email marketing automation software’s random sampling algorithm. With colour as the independent variable, the effect on recipients’ engagement with the email (measured in clicks) was evaluated.
The emails were sent to the sample group on Wednesday, 18th of April 2018 at 9:00 GMT, and the clicks were collected over a 7-day time period. The final click-through data were exported from the email automation software and for each email delivered, the colour scheme,
recipient nationality and success in generating a click were recorded. Additionally, the number of emails opened in the past and purchase history were collected and recorded for each email recipient individually. The facets that were kept the same for all emails are timing, content and structural design. All emails were sent on the same date and hour of the day, to avoid timing bias (Drèze & Bonfrer, 2008). Similarly, all emails had the same subject line, sender name, content, structure, background colour, body text font style, colour and size to avoid the effect of content and design customization, which have the potential to increase CTR by over 60% (Ansari & Mela, 2003).
3.3 Variables and Measures 3.2.1 Dependent Variable
Click: The click behaviour of email recipients was measured as a two-item categorical variable. The email performance report exported from the email marketing automation software assigned email recipients with a status identifying where their interaction with the email ended. Three statuses observed were Sent for people who received an email but did not open it, Opened for people who opened the email and Clicked Link for people who opened the email and clicked on a link within an email. Whether a person decides to open an email is an accumulation of various of different factors such as the subject line, timing or to which inbox was the email assigned by their email service provider. To eliminate the bias and influence of these factors and amplify the focus on the effectiveness of the emails’ content (O’Connell, 2008), only recipients marked as Opened and Clicked Link were included in the analysis. Therefore, the performance measure of interested to this study is the click-to-open rate (CTOR) calculated as the number of unique clicks as a percentage of unique email opens (Lingley & Campos, 2009). In the database used in the statistical analysis, people who solely opened the email were assigned the Click variable value no, and people who clicked on a link were assigned a Click variable value yes.
3.2.2 Independent Variables
Colour hue: Colour hue corresponds to the wavelength of light, representing colour’s value of red, yellow, blue and green pigment (Valdez & Mehrabian, 1994; Zviran et al., 2006). Based on hue, colours may be divided into a warm, cool and neutral (Crowley, 1993). Long wavelength colours such as red, orange and yellow, represent the warm group. Cool colours including green, blue and violet, comprise of short wavelength. Throughout this experiment two cold hues, namely green and blue, and two warm hues, namely red and orange, were randomly assigned to four treatment groups. Email colour was manipulated by changing the colour of the heading and the call-to-action button included in the body of an email. The fifth, control group, has received an email with a black heading and white call to action button. A categorical variable was used to represent different email colours in the dataset analysed (0 – white, 1 – blue, 2 – green, 3 – orange, 4 – red). These four colours were chosen because of the limitations imposed by the company’s corporate guidelines, which did not allow for the use of yellow and violet in corporate communications. Nevertheless, the colours tested succeed to proportionally represent both warm and cool colour groups. In addition, they provide for good visibility on screens regardless of their brightness level, which could have been a problem if using a less dominant colour such as yellow.
Brand experience: Brand experience is defined as a construct formed by accumulation of relevant brand information in form of ad exposure, information search or product usage (Ha & Perks, 2005; Baker et al., 1986). While the extent of search performed by an individual email recipient could not be recorded and controlled for, it was possible to collect precise information about the other two sources of brand experience; ad exposure and product usage. The company’s email automation software allowed the number of exposures to brand’s promotional emails to be tracked, by measuring the number promotional emails opened by each recipient. In addition, the company’s customer relationship management software allowed for the
purchase history of each recipient to be recorded, namely whether the person is or is not a previous customer. This information was then used to measure and evaluate the level of brand experience on a point scale (1 – low brand experience, 5 – high brand experience). A five-point scale was adopted from Ha & Perks (2005) and Labrecque & Milne (2012). However, it is important to note that while all past studies working with the construct of brand experience allowed for participants’ self-evaluation of their level of brand experience, this experimental design did not allow to do so, thus the evaluation was performed by the researcher based on available online behavioural data. Using the data about recipients’ purchase history and ad exposure, this research accounts for one of four factors of brand experience identified by Brakus et al. (2009), behavioural experience, by answering to the authors’ question whether individuals engaged in physical actions and behaviours when using the brand.
First, consumers who have made at least one purchase of a product or service from the brand were assigned a high level of brand experience. Consumers who have not purchased a product or service before, were classified based on the number of the brand’s promotional emails they have opened. Table 1 provides a detailed explanation of the brand experience level categorization conditions.
Table 1: Brand experience categorization conditions
Brand experience Purchase history (primary condition)
Ad exposure - number of promotional emails opened (secondary condition)
5 - high Customer 20+
4 – medium high Not a customer 15-19
3 – medium Not a customer 10-14
2 – medium low Not a customer 5-9
4. Results
In this section, the results of performed data analysis will be presented. First, a summary of the recipients included in the analysis will be provided. Second, the data preparation will be described, followed by the actual statistical model and hypotheses testing.
4.1 Recipients
Out of the 19 642 relevant recipients, 5 899 people opened the email. Therefore, 5 899 people were taken into account in the analysis. The nationality mix of the recipients was the following; 54,6% were from the United Kingdom, 16,9% from The Netherlands, 11,4% from Belgium, 9,1% from Germany, 3,9% from Finland, 1,5% from Norway, 1,2% from Sweden, 0,9% from Luxembourg and 0,5% from Denmark. 53,98% of these individuals hold a managerial position and the rest consists of business professionals and specialists.
4.2 Data preparation
First of all, a check of frequencies was performed to examine if there were any errors in the data collected. As a result, no errors and missing values were found. Afterwards, skewness, kurtosis and normality tests were performed. The tests revealed that neither of the variables is normally distributed. The 0.015 skewness value of Email colour represents a symmetrical distribution, yet the -1.305 kurtosis value suggests that the distribution is platykurtic; flatter as compared with normal distribution. Brand experience has 0.504 skewness and -1.127 kurtosis values, suggesting that the distribution is left-skewed and without a dominant peak (Figure 3). The dichotomous dependent variable Click had a value of 1.949 for skewness and 1.800 for kurtosis, meaning that the data were heavy-tailed and strongly skewed to the left. This is because of all the responses monitored, only 15,2% of the recipients clicked on the link in email. Yet, according to Tabachnick & Fidell (2001), reasonably large samples reduce the negative impacts of extreme skewness and kurtosis on the analysis. Since this study
was performed on a sample size of 5899 individuals, these values should not substantially harm the quality of the analysis.
In order to avoid bias caused by unequal distribution of brand levels across colour hues, a crosstabulation of their distribution was created (Table 2), revealing a relatively proportionate distribution. The differences between individual groups lie within the 5% range, which makes the data suitable for further analysis.
Table 2: Distribution of different levels of brand experience across colour hues
A correlation analysis was performed to examine the relationship between all variables; Email colour, Brand experience and Click. From the correlation matrix provided by SPSS, Brand experience is the strongest predictor of Click with a Pearson correlation coefficient of 0.131 and the significance value less than 0.01 (Table 3). Email colour follows a Pearson correlation coefficient of -.012 and the significance value 0.343, suggesting no significant correlation with the click rate of an email.
Table 3: Means, Standard deviation and Correlations
4.3 Hypothesis testing
A binary logistic regression analysis was conducted to examine the relationship between
Brand experience Basic (white)
blue green orange red
Low 19.5% 19.7% 19.3% 20,2% 21.3% Medium Low 18.5% 20.7% 20.8% 20.2% 19.8% Medium 20.2% 21.0% 19.4% 18.6% 20.8% Medium High 22.0% 20.0% 19.8% 19.4% 18.7% High 22.5% 20.4% 19.2% 17.9% 20.0% Variables Means SD 1 2 Click 0.15 0.359 - Email colour 2.00 1.415 -.012 - Brand experience 2.87 1.495 -.131** -.031
Brand experience). A logistic regression was performed to research the ability of Email colour and Brand experience to predict Click on link in an email, while keeping the email content, structural elements, subject line and timing unchanged. First, after general overview of the frequencies, a crosstabulation was conducted to get an idea of the distribution of clicks across different colour and brand experience levels.
Table 4: Crosstabs effect of different colour hues on email effectiveness
Table 5: Crosstabs effect of different levels of brand experience on email effectiveness
Brand experience
Clicked (no) Clicked (yes) Pearson Chi-Square df Asymptotic Significance (2-sided) low 71.2% 28.8% medium low 87.3% 12.7% medium 90.5% 9.5% medium high 90.1% 9.9% high 88.5% 11.3% Total 84.8% 15.1% 230.567 a 4 .000
a 0 cells (0.0%) have expected count less than 5. The minimum expected count is 68.58.
Table 4 suggests that the email colour did not make any significant difference for the CTOR of the marketing newsletters tested. There is a slight decrease in CTOR of the red colour email when compared to the control group and the remaining three treatment groups. However, the asymptotic significance of this relationship is at p = .858 level, thus highly insignificant. Table 5 provides insight into the frequencies of clicks for different levels of brand experience. It shows that people with the lowest level of brand experience clicked on the link within email
Email colour
Clicked (no)
Clicked (yes) Pearson Chi-Square df Asymptotic Significance (2-sided) basic 84.6% 15.4% blue 84.4% 15.6% green 84.7% 15.3% orange 84.9% 15.1% red 85.9% 14.1% Total 84.9% 15.1% 1.319 a 4 .858
newsletter more than twice as much as the remaining the groups. This effect is evaluated as significant with p = 0.000.
In order to give meaning to the crosstabulations, a binary logistic regression was performed (Table 6). Logistic regression is a multiple regression with a categorical outcome variable and either continuous or categorical predictor variables (Field, 2012). Since the outcome variable Click is of binary nature (click/no click), a binary logistic regression was employed to develop the statistical model. The logistic regression model was statistically significant (2 = 108.901, p = 0.000). However, only one predictor variable was rated significant, and the model explained only 3,2% (Nagelkerke R2) of the variance in Click. Therefore, it can be concluded that most variance in the outcome variable Click is explained by other factors that are not included in the model.
Table 6: Binary logistic regression – colour hues and brand experience * clicked email
B S.E. Wald df Sig. Exp(B) 95% C.I. EXP(B) Lower Upper Brand exper -.282 .029 98.083 1 .000 .754 .713 .797 White 1.827 4 .767 Blue -.004 .114 .001 1 .973 .996 .796 1.246 Green -.032 .115 .076 1 .783 .969 .773 1.214 Orange -.064 .116 .299 1 .585 .938 .747 1.179 Red -.136 .117 1.358 1 .244 .873 .694 1.097 Constant -.965 .107 81.698 1 .000 .381 Hypothesis 1
Testing the first hypothesis about the effect of recipient’s brand experience on his propensity to click on a link within an email, we identify the level of brand experience as a continuous predictor (1-low brand experience, 5-high brand experience) and the response of either clicking or not clicking on the email is set as a dichotomous outcome. The regression pointed out that increased brand experience significantly decreases the likelihood of clicking on a link ( = -.282; p = 0.000). The odds ratio demonstrates that for every unit increase in
Brand Experience, the propensity to Click decreases by 20.46%. These results oppose the suggestions presented in H1, hence it is rejected.
Hypothesis 2
Testing the hypotheses H2a and H2b about the effect of email colour on recipient’s propensity to click on a link within an email, we identify the colour as a categorical predictor and the response of either clicking or not clicking on the email is set as a dichotomous outcome. The control group where the colour of an email is not manipulated (remains black and white), was used to examine whether making a call-to-action button colourful has any effect on recipients’ propensity to click on attached links. The crosstab analysis showed that email with red colour led to the lowest CTOR of 14.1%, which was 1.3% lower than the CTOR of the control group (Table 3). For the remaining colour treatments, the difference was less than 0.5%. The regression analysis revealed that the < 0 and p > 0.05 for each of the colour treatments. Therefore, while neither of the email colour treatments led to an increase in the odds of clicking on a link within an email, this effect was not significant. To conclude, there is no significant effect of email colour on the propensity to click on a link within a marketing newsletter, thus both H2a and H2b lack empirical support.
5. Discussion
5.1 General discussion
Email as a marketing channel is popular for its easy customization, measurability, integration with other customer relationship management systems and superior ROI (Aufreiter et al., 2014; Jenkins, 2009; Sharma & Seth, 2004). Yet, as the increasing number of online brands leads to increasing information overload (Eppler & Mengis, 2004) and the number of promotional emails generated (Hartemo, 2016), there is an everlasting need to enhance the understanding of email marketing effectiveness. Since the effectiveness of marketing emails is
2008), this research aimed to investigate how a specific email and consumer characteristics affect email click-to-open ratio (CTOR).
Specifically, this study enquired into the effect of email colour and recipient’s brand experience on the propensity to click on a link within a promotional email. Past research proved that cool colours (Fogg, Lee & Marshall, 2002; Kim & Moon, 1998; Baker et al., 1994) and high levels of brand experience (Ha & Perks, 2005; Gefen, Karahanna & Straub, 2003) have direct positive effect on brand trust, and brand trust has a positive effect of intention to comply with an appeal made in an email (Wilson et al., 2015). Thus, it was assumed that brand trust mediates the positive effect of cool colours and high levels of brand experience on recipients’ intention to behave in a way desirable for the email sender; click. To examine the different relationships, an online factorial treatment experiment was performed on an international consultancy company’s contact database. In an online field experiment, 4 email colours were tested on groups more than 1000 recipients, whose levels of behavioural brand experience (Brakus et al., 2009) were subsequently evaluated. The clicks were collected over a 7-day time period and SPSS program was used to perform statistical analysis on the resulting data.
Both hypotheses were tested using the binomial logistic regression analysis. As a result, neither of the hypotheses was supported. The effect of higher levels of brand experience on one’s propensity to click on a link was found to be significantly negative, which is why Hypothesis 1 was fully rejected. No relationship was found between colour of an email and its CTOR, hence Hypothesis 2a and 2b lack empirical support. Consequently, it could be stated that email marketing effectiveness is not affected by the colour of email headings and call-to-action buttons, and that lower levels of brand experience lead to higher propensity to click on a link within an email.
Hypothesis 1
The relationship found between the brand experience level and propensity to click was against expectations posed in the reviewed literature. The direction of the relationship between brand experience level and number of clicks on a link was negative, thus opposite the assumption presented in Hypothesis 1. With an odds ratio of 0.754, one unit increase in brand led to 24.6% decrease in the likelihood of clicking on a link within an email.
However, there is a logical explanation for the negative relationship between these variables. Brand experience is correlated with brand expertise and brand knowledge (Alba & Hutchinson, 1987). As shown by Moorthy et al. (1997), brand experience initially increases the amount of brand information search, but the knowledge effect takes over beyond a certain level of experience and leads to reduction in need for further information and decline in the amount of information search. This would explain why consumers with lowest brand experience level generated more than double the open rate of consumers with second lowest level of brand experience. Consumers who have only recently joined the mailing list of the brand were not familiar with its offers and expertise, which is why they were more prone to click links in its marketing emails in search of more information. After opening several emails and gaining sufficient knowledge about the brand, recipients are able to better evaluate the relevance of individual emails and click only if they find the offer personally pertinent. As a result, the CTOR for more experienced recipients decreases. Consequently, it can be stated that low levels of brand experience increase recipients’ propensity to click on email advertisement due to lower levels of brand knowledge leading to higher levels of information seeking.
Hypothesis 2
Despite the fact that the red email received the lowest number of clicks, the results of logistic regression did not show any significant relationship between the colour of an email and its CTOR. Furthermore, while green and blue led to higher CTOR than red, the CTOR of orange
exceeded the CTOR of both cool hues, and no colour exceeded the CTOR of the control (black and white) email. Therefore, the assumption that cool hues in an email lead to a higher number of clicks on link in an email was not supported.
There are two main reasons providing an explanation for the failure to support these hypotheses. First, the hypotheses were predominantly based on findings from web-site design context, which favours cool hues in web-store background as leading to more positive consumer attitudes towards the e-retailer (Fogg, Lee & Marshall, 2002; Kim & Moon, 1998). However, while both email and website environment can be characterized as “pull” advertising platforms (Thompson & Wassmuth, 1999), they pose several crucial differences. Websites are more complex, interactive and the amount of time that people spend on engaging with them is significantly higher (Nielsen, 2011; White, 2017). Accordingly, information and visual cues processing on a website is ultimately different from an email. Hence, findings related to website’s design effect on consumer attitudes and responses should not be automatically extended to email design. All web-design findings ought to be tested prior to application to other online advertising platforms, which indeed was the one of the purposes of this research. Secondly, as people spend little time on processing promotional emails (Jensen & Jepsen, 2007; White, 2017), colour needs to be dominant to make an impact. This explains why Zviran et al. (2006), who manipulated the full background colour of an email, discovered significant effect of colour on email’s click-through-rate. This research tested colour effects in a present-day email marketing design, which favours white backgrounds for the sake of better readability (Hall & Hanna, 2004) and faster information processing (Ling & van Schaik, 2002). The results suggest that under these circumstances, colour of email heading and CTA button do not significantly affect consumer behavioural responses.
5.2 Theoretical implications
This study contributes to academic literature by shedding light on the impact of recipients’ experience with the sender brand on his responsiveness to its email marketing communications. While past literature predominantly favours brand experience as desirable, and benefiting consumer attitudes such as brand trust, familiarity and satisfaction, this research points out that it does not necessarily benefit consumer behaviour and his responsiveness to brand’s email advertising. In fact, higher levels of brand experience lead to lower levels of clicks on links within tested promotional emails, owing to the fact that brand experience reduces one’s inquisitiveness and need to learn more about the brand. As a result, this study emphasizes the ambiguous effect of brand experience on one’s interactions with a brand.
Furthermore, this study reveals that the colour of a call-to-action button and heading is not one of the determinants of nowadays’ email marketing effectiveness. Colour plays an important role if used in the background of an email (Zviran et al., 2006), however, as long as the email background is white, the structure is clear and text readable, the colour of small structural elements does not significantly help nor harm the effectiveness in generating clicks. Lastly, this study supports the belief that discoveries related to e-commerce and web design effectiveness should not be automatically assumed to apply to email marketing. Despite the similar nature of websites and emails, email as a platform is less complex and receives less processing time than website. Hence, each finding from the web-design and e-commerce context should be first tested on email marketing before widely applied.
5.3 Managerial implications
This study found that higher brand experience levels lead to lower levels of behavioural engagement with email advertising. This points out to the fact that email marketing in form of newsletters and promotional offers is an effective tool for inciting interest and awareness for new contacts. On the other hand, email marketing effectiveness is lower for existing and
experienced consumers, as their propensity to click on links in these emails is lower. Managers shall keep these tendencies in mind when designing and evaluating their marketing strategies. For instance, their marketing strategy could be tailored differently for people with different levels of brand experience. In fact, brand experience level could become another consumer segmentation variable. Given that more experienced consumers are less prone to click on links, they might require higher levels of personalization, while new contacts could receive a greater amount of more general marketing emails.
With regard to email colour, this study found that there is no significant difference between CTOR of emails with blue, green, orange, red or white call-to-action buttons. This means, that managers can willingly choose their colour palette as long as the email background is white, the structure is clear, and the text is readable. Accordingly, this study accentuates that marketing managers should not rely on colour to compensate for the lack of other email qualities. Rather than email’s colour palette, email marketing effectiveness is defined by its timing (Drèze & Bonfrer, 2008; Merisavo & Raulas, 2004), level of personalization and resulting personal relevance (Hemersma, 2013; Edwards, 2016) and benefits offered in return for engaging with the message (Wilson et al., 2015), and these are the attributes which managers should focus on and treat with vigilance.
5.4 Limitations and future research
This study was dedicated to the examination of whether colour and higher brand experience levels have the potential to benefit recipients’ propensity to click on a link, which is the ultimately desirable action for the email sender. The proposed hypotheses were founded on the assumptions of an empirically tested framework presented by Wilson et al. (2015), identifying trusting beliefs in the sender as one of the of four main determinants of email’s effectiveness. Neither of the hypotheses presented at the beginning of this research was supported. Although they were founded on previous research (e.g. Wilson et al., 2015; Ha &
Perks, 2005; Zviran et al., 2006; Fogg, Lee & Marshall, 2002), they were not proven to be correct in this experiment. The lack of support for these hypotheses may be due to methodological limitations.
First, the experimental design did not allow for controlling of several potentially influential variables. Although the software used to perform the analysis allowed for access to a relatively large sample, it also imposed several restrictions on the nature and amount of data that could be collected. The email marketing software used does not record recipients’ gender, which could not be included as a control variable in the analysis. Similarly, subjective attitudes to the brand and interest in its offerings could not be recorded. As a consequence, it can be concluded that the model developed did not control for all potential determinants of recipients’ propensity to click on a link in an email newsletter. Second, the assessment of consumer brand experience only accounted for one’s behavioural brand experience and did not account for its remaining three dimensions; sensory, affective and intellectual brand experience, presented by Brakus et al. (2009). This could lead to a certain level of bias, as some people who have opened several emails could still consider themselves to have low level of brand experience, while people who have received only one email might consider themselves as experienced and familiar with the brand. Nevertheless, these limitations can also serve as new opportunities for future research, in which the effect of brand experience on consumer behavioural outcomes can be investigated in a more controlled setting of a lab experiment. Such design would allow for measurement and manipulation of all four factors of brand experience, gender and even subjective constructs such as individual perception of benefits offers by the email.
Another notable limitation is related to the spectrum of colours tested throughout the experiment. Due to restrictions imposed by the corporate guidelines, the experiment could only operate with colours identified as corporate colours of the sender company. As a consequence, yellow and violet colour had to be excluded from the experiment. Furthermore, these corporate
guidelines did not allow for manipulation of colour brightness and saturation, which might also have an impact on the way consumers process marketing cues and respond to email advertising. Future research should investigate this area and study the effect of colour brightness and saturation on consumer responsiveness to online and email marketing. In addition, this experiment did not examine specific combinations of multiple colours together. Therefore, the results of this study cannot be generalized to emails with other than white background, or emails utilizing more than one colour in its structural elements, as it does not address the joint effects of different colours, and whether warm and cool colour hues combined could attenuate the effect of each other.
Lastly, the online factorial treatment experiment was performed on marketing emails sent by a company operating in a business-to-business sector, thus the results have application limited to B2B sector only. Future research could investigate the effect of cool hues and different levels of brand experience on responsiveness to marketing emails in a business-to-customer (B2C) sector.
6. References
Aghidae, S., & Honari, R. (2014). Investigating the Psychological Impact of Colors on Process of Consumer Shopping Behavior. Internatioinal Review of Management and Business Research, 3 (2), 1244.
Alba, J. W., & Hutchinson, J. W. (1987). Dimensions of consumer expertise. Journal of consumer research, 13(4), 411-454.
Ansari, A., & Mela, C. (2003). E-Customization. Journal of Marketing Research, 40 (2), 131-145.
Aufreiter, N., Boudet, J., & Weng, V. (2014). Why marketers should keep sending you e-mails.
McKinsey&Company. Retrieved from:
http://www.mckinsey.com/insights/marketing_sales/why_marketers_should_keep_sending_y ou_emails
Babin, B. J., Hardesty D. M., & Suter, T.A. (2003). Color And Shopping Intentions: The Intervening Effect Of Price Fairness And Perceived Affect. Journal of Business Research, 56, 541– 551.
Bagchi, R., & Cheema, A. (2013). The Effect of Red Background Color on Willingness-to-Pay: The Moderating Role of Selling Mechanism. Journal of Consumer Research, 39, 947–60.
Baker, J., Grewal, D., & Parasuraman, A. (1994). The influence of store environment on quality inferences and store image. Journal of the academy of marketing science, 22(4), 328-339.
Baker, W., Hutchinson, J., Moore, D., & Nedungadi, P. (1986). Brand familiarity and advertising: effects on the evoked set and brand preference. ACR North American Advances. Retrieved from http://www.acrwebsite.org/search/view-conference-proceedings.aspx?Id=6570
Bellizzi, J. A., & Hite, R. E. (1992). Environmental color, consumer feelings, and purchase likelihood. Psychology & marketing, 9(5), 347-363.
Brakus, J. J., Schmitt, B. H., & Zarantonello, L. (2009). Brand experience: what is it? How is it measured? Does it affect loyalty? Journal of marketing, 73(3), 52-68.
Braunsberger, K., & Munch, J. M. (1998). Source expertise versus experience effects in hospital advertising. Journal of Services Marketing, 12(1), 23-38.
Campbell, M. C., & Keller, K. L. (2003). Brand familiarity and advertising repetition effects. Journal of consumer research, 30(2), 292-304.
Chaudhuri, A., & Holbrook, M. B. (2001). The chain of effects from brand trust and brand affect to brand performance: the role of brand loyalty. Journal of marketing, 65(2), 81-93.
Chittenden, L., & Rettie, R. (2003). An evaluation of e-mail marketing and factors affecting response. Journal of Targeting, Measurement and Analysis for Marketing, 11(3), 203-217.
Constine, J. (2016). Facebook encourages text statuses with new colored backgrounds. TechCrunch.com. Retrieved from: https://techcrunch.com/2016/12/19/facebook-status-background-color/
Coursaris, C. K., Swierenga, S. J., & Watrall, E. (2008). An empirical investigation of color temperature and gender effects on web aesthetics. Journal of Usability Studies, 3(3), 103-117.
Crowley, A. E. (1993). The two-dimensional impact of color on shopping. Marketing letters, 4(1), 59-69.
Demangeot, C., & Broderick, A. (2010). The Marketers’ Prismatic Palette: A Review of Color Research and Future Directions. Psychology & Marketing, 30 (6), 461-469.
Drèze, X., & Bonfrer, A. (2008). An empirical investigation of the impact of communication timing on customer equity. Journal of Interactive Marketing, 22(1), 36-50.
Edwards, S. (2016). Why business personalization is the key to long-term success. Inc.com. Retrieved from: https://www.inc.com/samuel-edwards/why-business-personalization-is-the-key-to-long-term-success.html
Eisinga, M. (2017). The effect of different levels of personalization on ad effectiveness (Unpublished master’s thesis). University of Amsterdam, Amsterdam, The Netherlands.
Elliot, A. J., Maier, M.A., Moller, A.C., Friedman, R., & Meinhardt, J. (2007). Color and Psychological Functioning: The Effect of Red on Performance Attainment. Journal of Experimental Psychology, 136(1), 154–168.
Ellis-Chadwick, F., & Doherty, N. (2012). Web advertising: The role of e-mail marketing. Journal of Business Research, 65(6), 843-848.
Eppler, M.J., & Mengis, J. (2004). The concept of information overload: A review of literature from organization science, accounting, marketing, MIS and related disciplines. The Information Society, 20 (5), 325 – 344.
Fogg, B.J., Lee, E. and Marshall, J. (2002). Interactive Technology and Persuasion. The Persuasion Handbook: Developments in Theory and Practice. ThousandOaks. CA: Sage, 765– 88.
Fogg, B.J., Soohoo, C., Danielson, D.R., Marable, L., Stanford, J., & Tauber, E.R. (2003). How do users evaluate the credibility of Web sites? In Proceedings of the 2003 conference on Designing for user experiences. ACM, San Francisco, California, 1-15.
Gefen, D. (2000). E-commerce: the role of familiarity and trust. Omega, 28(6), 725-737.
Gefen, D., Karahanna, E., & Straub, D. W. (2003). Trust and TAM in online shopping: An integrated model. MIS quarterly, 27(1), 51-90.
Gill, H., Boies, K., Finegan, J. E., & McNally, J. (2005). Antecedents of trust: Establishing a boundary condition for the relation between propensity to trust and intention to trust. Journal of business and psychology, 19(3), 287-302.
Godin, S. (1999). Permission Marketing: Turning Strangers into Friends and Friends into Customers. New York: Simon & Schuster.
Gorn, G. J., Chattopadhyay, A., Sengupta, J., & Tripathi, S. (2004). Waiting for the web: How screen color affects time perception. Journal of Marketing Research, 41, 215–225.
Gorn, G., Chattopadhyay, A., Yi, T., & Dahl, D. (1997). Effects of Color As an Executional Cue in Advertising: They're in the Shade. Management Science, 43(10), 1387-1400.
Ha, H. Y., & Perks, H. (2005). Effects of consumer perceptions of brand experience on the web: Brand familiarity, satisfaction and brand trust. Journal of Consumer Behaviour, 4(6), 438-452.
