Instagram:
The Influence of Background Features
and Tone of Voice
on Purchase Intention
Assignment: Track: Name: Department: Student ID: Supervisor: Programme: Date of Completion: Master’s Thesis Persuasive CommunicationNina van der Sluis Graduate School of Communication
12041718 Dr. Ivana Bušljeta Banks
MS Communication Science 28th June 2019
Abstract
In this thesis an online experiment was conducted to measure the effectiveness of background type on purchase intention, on Instagram. Furthermore, the tone of voice used in the description of the Instagram post was manipulated in order to assess its moderating effect on the relationship between background type and purchase intention. This subject is relevant for many marketing professionals and brands active on Instagram. Instagram has become one of the most important social media of the moment, and brands and marketing professionals are in need of guidance when it comes to choosing which strategy to adopt in order to attract followers and potential customers. Contrary to the expectations, it was found that portraying a product on a white background leads to higher purchase intention of that product compared to portraying a product in a real-life setting. Next to that, the use of tone of voice in the
description of an Instagram post was found to influence consequent purchase intention of the shown product, depending on the background type. It was concluded that potential customers definitely prefer a white background accompanied with a neutral tone of voice in the
Table of contents
Introduction ... 5
Theoretical framework ... 7
Instagram ... 7
Background features ... 8
Product promotion on Instagram ... 9
The influence of background features on purchase intention ... 9
The moderating role of tone of voice ... 11
Methods ... 12
Design and sample ... 12
Stimuli ... 13 Procedure ... 15 Results ... 16 Randomization ... 16 Measures ... 17 Manipulation ... 18 Main analysis ... 18
Discussion and conclusion ... 20
Limitations and directions for future research ... 22
References ... 25
Appendix A ... 29
Appendix B ... 31
Appendix C ... 32
Introduction
With the rise of social media, an additional touch point for brands has been created (Roncha & Radclyffe-Thomas, 2016). Social media have become a significant part of
people’s daily life. Consumers’ buying behavior has drastically changed due to its advent, as
people have greater purchasing options to choose from (Ramanathan, Subramanian & Parrott, 2017). Furthermore, social media platforms facilitate the interaction between brands and consumers and a two-way flow of information exchange (Nambisan & Baron, 2007).
From the brand’s point of view social media are a benefit, because they simplify consumer reach and engagement with consumers. This engagement can easily trigger
consumer’s interest towards the brand and facilitate the creation of a dynamic relationship and
full immersion between brand and consumer, through brand stories (Roncha & Radclyffe-Thomas, 2016). The creation of a good and loyal relationship, consequently, enhances the likeability of consumers to buy a product promoted by the brand as well as the likeability to recommend that product to other potential consumers (Gee, Coates & Nicholson, 2008). Social media, furthermore, allow brands to reach a vast number of potential consumers in a blink of an eye. This gives brands the opportunity to build solid and long-lasting relationships with consumers, increasing brand loyalty and awareness (Roncha & Radclyffe-Thomas, 2016).
Among diverse social media platforms, Instagram has shown to be especially effective in connecting brands and consumers (Weise, 2015). The platform was created in 2010 for sharing videos and photos on mobile devices with other users, called ‘followers’, with the possibility to ‘like’ each other’s photos and comment on them (Roncha & Radclyffe-Thomas,
2016). Moreover, the platform allows users to discover interesting topics and diverse brands associated with these topics, by the use of hashtags and the “explore” button (Instagram, 2019). This facilitates their search for products and brands they are interested in. The platform
gives brands the opportunity to get a deeper understanding of their consumers, in terms of drivers and preferences. This way, brands can adapt their marketing strategy to the needs of their consumers (Roncha & Radclyffe-Thomas, 2016).
The use of images plays an important role in a marketing strategy of a brand, as it leads to high levels of engagement (Roncha & Radclyffe-Thomas, 2016). Most brands use the emotion strategy on Instagram, in order to influence affective beliefs. This strategy consists in the creation of brand associations by the use of transmitting emotions experienced when using the product (Goor, 2012). These emotions are transmitted by the use of appealing
backgrounds in the images. In fact, according to Harris and Goode (2010), aesthetic appeal is said to attract consumers.
Several researchers have studied the effect of background features on consequent purchase intention, as well in the physical ambience as online (Cheng, Wu and Yen, 2009; Harris & Goode, 2010; Lingling Gao, Xuesong Bai, 2014; Vilnai-Yavetz & Rafaeli, 2006), but not many have taken into consideration the social medium platform Instagram and its potential for brands in influencing consumers by the use of attractive backgrounds in the images promoting a product. Moreover, little research has been done on this platform in how well brands succeed in positioning their products: does it actually leads to more sales?
Based on this, the current study will elaborate on previous research in order to shed more light on the influence of background on purchase intention on Instagram. The aim is to respond to two important questions. First, does the background type (real-life setting vs. white background) of an Instagram post positively influence subsequent purchase intention?
Second, does the tone of voice adopted by the brand in the description of the Instagram post moderate the influence of background type on purchase intention?
The current study will first present a review of the role of background features in influencing consumer’s purchase intention and the role of tone of voice in moderating the
main effect. Subsequently, the methodology of this research will be presented along with the study findings. Finally, the results will be discussed together with some future implications and lastly a conclusion will be provided.
Theoretical framework Instagram
From the point of view of the consumers, Instagram gives them the possibility to follow brands they are interested in and to search information based on different hashtags (Roncha & Radclyffe-Thomas, 2016). For brands on Instagram, hashtags have become very important as well, as they give them the opportunity to get easily known to other users and to get consumers to follow them (Wallsbeck & Johansson, 2014). This illustrates how this platform enables brands to easily engage and reach consumers (Ginsberg, 2015). Further, the platform gives companies and brands the opportunity to promote themselves in a low-cost, high quality, competitive way (Wallsbeck & Johansson, 2014). For this reason, recent years showed an increase in brand registrations on Instagram (Phua, Jin & Kim, 2016). Instagram also gives brands the opportunity to get a deeper understanding of their consumers, in terms of drivers and preferences. This way, brands can adapt their marketing strategy to the needs of their consumers (Roncha & Radclyffe-Thomas, 2016). In fact, according to the Uses and Gratifications Theory people go on social media to fulfill needs which will lead to
gratifications (Phua, Jin & Kim, 2016). According to Phua, Jin & Kim (2016), users have various motivations of being on social media, such as Instagram: mainly information-seeking, leisure and entertainment.
In sum, Instagram seems to be an attractive platform for brands to engage with consumers and promote themselves: it gives brands the opportunity to use this platform to create content and engage these possible consumers with various content (Carah & Shaul,
2016). Further, the platform is a suitable tool to increase brand exposure and brand awareness in a short time span (Phua, Jin & Kim, 2016).
Background features
Kotler (1973) was one of the firsts to study the relevance of background features as a marketing tool. He called these features ‘atmospherics’ and defined them as “the space designed to create a certain effect in buyers” (Kotler, 1973). Kotler (1973) also demonstrated that background features can have an impact on behaviors and attitudes of consumers,
because they influence their emotions and satisfaction. More specifically, background features create a setting which transform background features into implicit cues, influencing perceived quality of the brand (Fowler & Bridges, 2012).
According to Jaakonmaki, Müller and vom Brocke (2017), content features are mostly a combination of text, visual, and audio content. Online measures of these features are said to be lacking (Harris & Goode, 2010). These features are mostly referred to as ‘servicescapes’. Bitner (1992) coined the word “servicescapes” referring to the physical surroundings which
are created by marketeers to facilitate the provision of product offerings to customers, by impacting consumers’ subconsciousness. These servicescapes are made up by both tangible
and intangible features that create the product experience (Ariffin, Bibon and Abdullah; 2012).
A study from Ariffin, Bibon and Abdullah (2012), which took place in a restaurant, has shown that pleasant scent, pleasant music, comfortable temperature, low levels of noise, and adequate lightning, all in harmony with other elements of the restaurant, result in
customers having favorable perceptions and positive experiences (Ariffin, Bibon & Abdullah, 2012). It could be caused by the different cues that people perceive through their visual senses. In fact, the atmosphere of a particular setting is delivered through the sensory
channels, and consequently encoded accorded to the motivation of information processing the cues people perceive elicit in the consumer (Cheng, Wu & Yen, 2009).
Product promotion on Instagram
With the rise of social media many companies and brands are now also present on social media to promote their products (Hanna, Rohm & Crittenden, 2011). An important reason for brands to engage in social media is the fact that it is much cheaper and more effective than most other marketing strategies, besides the larger consumer reach and the fact that there are no geographical or time constraints. In fact, through social media one can easily and instantaneously access information about a product or service. This increases the
relevance for marketing professionals to promote brands in a way that is attractive to consumers. Their aim, in fact, is to increase the percentage of consumers who think of their products (Martinus & Anggraini, 2018).
A good strategy to increase the percentage of consumers who think of a specific product is having innovative and appealing promotions through social media. This is why marketing professionals started to use “visual storytelling”: the promotion of real experiences through photo and video content (Ginsberg, 2015). By means of this strategy consumers can perceive the quality and experiences they will encounter when purchasing a particular product. This way, brands can add value to both their brand as well as to the daily lives of their consumers (Ginsberg, 2015).
The influence of background features on purchase intention
An appropriate use of backgrounds is considered to be important for online shoppers (Harris & Goode, 2010). Research about online aesthetic appeal shows that the interpretation of the servicescape by consumers can be attractive or alluring (Harris & Goode, 2010). Servicescapes are, in fact, seen as critical when consumers engage in sales, because they are the artefact representing the brand (Vilnai-Yavetz & Rafaeli, 2006). Particular background
features allow consumers to form an interpretation of the product’s use in real-life. Therefore,
the visual element is really important, as it is the one which is immediately accessible (Breazeale & Ponder, 2011). Cheng, Wu, and Yen (2009) demonstrated that a certain setting is likely to influence people’s emotions and thus affect purchase behavior.
If we dive deeper into background features on the internet, a connection can be created with the last statement made by Cheng, Wu and Yen (2009). In fact, the study of Lingling Gao and Xuesong Bai (2014) showed that website features are created to evoke positive emotional and cognitive state in online consumers, in order to enhance willingness to spend more time and money for products. People are willing to spend more when their expectations are met regarding the aesthetics: when a customer feels satisfied, he will have a more positive attitude towards the brand and consequently will be more likely to engage in purchasing a promoted product (Vilnai-Yavetz & Rafaeli, 2006). Purchase intention has been defined as the probability that someone purchases a product, based on the intention and percentage that people will actually buy the product (Gageler & Van Der Schee, 2016). Purchase intentions of consumers are one of the most important inputs needed for marketing managers to determine eventual future sales and determine how a certain product promotion will impact consumers’
buying behavior (Morwitz, 2012). Product promotion has been shown to influence purchase intention (Vilnai-Yavetz & Rafaeli, 2006).
A shift has taken place in the online buying behavior: consumers are no longer only driven by utilitarian goals, but also by hedonic goals (Bridges & Florsheim, 2008). More precisely, hedonic aspects of the online service have an important effect on purchase intention, such that positive feelings towards the online service and satisfaction about the service lead to more purchase intention (Babin & Attaway, 2000). Satisfaction is,
Kim and Kandampully (2009) further confirm that aesthetics are indispensable for the attraction and retention of consumers.
In sum, background features have been shown to influence consumers’ purchase
intention (Cheng, Wu, and Yen, 2009; Lingling Gao & Xuesong Bai ,2014; Vilnai-Yavetz & Rafaeli, 2006). No research has been conducted yet about background features on Instagram. Therefore, this study will investigate the impact of background features of an Instagram post, promoting a product on subsequent purchase intention. Basically, previous research regarding servicescapes will be extended to the context of Instagram. Therefore, the first hypothesis states:
H1. An Instagram post containing the promotion of a product in a real-life setting
leads to a higher purchase intention than an Instagram post promoting a product on a plain, white background.
The moderating role of tone of voice
Tone of voice used by brands has been demonstrated to be decisive in forming
consumer attitude, determining the beginning of the relationship between consumer and brand (Keeling, McGoldrick & Beatty, 2010). A brand can adopt various types of tone of voice when addressing the consumer. Previous research has mainly focused on a conversational human voice or a more neutral tone of voice, which can also be referred to as a professional tone of voice (Kelleher, 2009).
A conversational human voice can be defined as a tone of voice which makes a brand feel like a real human: it is considered to be important for consumer’s experience, because it
addresses emotions and it contains attributes such as closeness and humanness (Barcelos, Dantas & Sénécal, 2018). A formal tone of voice is perceived to be more persuasive and profit-driven (Van Noort & Willemsen, 2011). The formal tone of voice can be referred to a factual tone of voice, based on attribute-value information. The arguments in such
information are mostly specific, clear and rational (Grabner-Kräuter & Waiguny, 2015). A formal tone of voice has been demonstrated to match consumer’s utilitarian buying
motivations (Keyzer, Dens and de Pelsmacker, 2017).
Research about the tone of voice adopted by brands in the description of an Instagram post is lacking. For this reason, it would be interesting to further investigate this. More precisely, current study will investigate whether tone of voice influences the likelihood of people to engage in consequent purchase intention. It is assumed that tone of voice moderates the effect of background type on purchase intention, as it is a feature which a brand can decide to either include to the Intsagram post or not.
In sum, current study will investigate the moderating effects of a description with a more human tone of voice, a neutral description with a more formal tone of voice and no description, on subsequent purchase intention:
H2. The effect of background type on consequent purchase intention is moderated by
the tone of voice used in the description of the post, such that (a) the influence is stronger when a brand uses a conversational human voice in the description, compared to a formal tone of voice or no description at all; (b) the influence is stronger when the brand uses a formal tone of voice in the description, compared to no description at all.
Methods Design and sample
The study will be a 2 (real-life setting vs. blank background) x 3 (presence of a description with a conversational human voice vs. presence of a description with a formal tone of voice vs. absence of description) between subjects, full-factorial, online experiment.
The study sample consists of people who have an Instagram account and who enjoy drinking some wine, as the product shown in the stimulus material is a bottle of white wine from the brand Wine Wizz. Wine has been chosen as product, because wine is a product
consumed for hedonic purposes and, as mentioned before, according to Bridges and Florsheim (2008) consumers’ buying behavior is now driven mostly by hedonic goals. In
relation to this, participants had to be older than 18, the legal age for drinking alcohol. In total, 287 participants were recruited for this study with the use of a snowball effect, through email and Whatsapp text messages. Of these 287 participants, 32 had to be excluded from the sample because they quit the survey before it had reached the end. Another eight participants were excluded due to missing values on several items. At the end, one more participant was excluded due to that person’s poor level of English proficiency. The reason for this last exclusion is related to the fact that participants with a poor level of English proficiency are assumed to be unable to understand the questions correctly.
After this, a final sample of 246 participants was left for the analysis. In order to control whether the age of participants was equally divided over the conditions, the average age was calculated. The average age of the participants was (M = 31.82; SD = 13.63). In order to control whether the gender was equally divided over the conditions, the percentage of males, females and other was calculated. The percentage of males (42.3%, N=246) was slightly lower than that of the females (56.9%, N= 246). Participants were also able to check the answer “other”. Results showed that 0.8% of the participants gave that as an answer.
Furthermore, participants had to indicate their level of English proficiency, which has been divided into native, advanced, moderate and low. Participants with a low level of English proficiency were considered to be inadequate to answer participate in the experiment. The level of people with a low level of English proficiency was 0%, after one person was taken out of the data set, the percentage of people with an intermediate level of English proficiency was 12.6%, the percentage of people with an advanced level of English proficiency was 61.5% and the percentage of Native English people was 25.9%.
Six groups were created for this experiment. For what concerns the background type, three groups were confronted with a real-life setting and other three groups were confronted with a white background. More precisely, the real-life setting figured the green colors of the grass and the trees, the blue sky with some clouds and the shadow of the tree due to the sun. The other one showed the same attributes related to the wine bottle, in the same perspective, but on a white background. Furthermore, the tone of voice used in the Instagram post was adapted to each of the groups, such that two groups had a more conversation human voice, two groups had a more formal tone of voice and two groups had no description at all.
The groups consisted of the following stimuli: the first group was shown a wine bottle within a real-life setting: more precisely outside in a garden, on a wooden table, with the sun shining. The description which accompanied the post of the first group had a conversational human voice (see Figure 1, Appendix A). The second group was shown a wine bottle within the same setting, with a more neutral tone of voice accompanying used for the description accompanying the post (see Figure 2, Appendix A). The third group was shown a wine bottle within the real-life setting, with no description of the post (see Figure 3, Appendix A). The fourth group was shown a bottle of wine with a white background, accompanied with a description having a conversational human voice (see Figure 4, Appendix A). The fifth group was shown a bottle of wine with a white background, accompanied with a more neutral tone of voice accompanying used for the description accompanying the post (see Figure 5, Appendix A). The last group was shown a bottle of wine with a white background, with no description of the post (see Figure 6, Appendix A).
The glass full of wine and the cork accompanying the bottle of wine are considered to be part of the product. Next to that, all the elements are seen to contribute to a greater picture of the experience of the product. Moreover, the choice of the real-life setting consisted in a
‘regular’ garden of someone’s home, so that participants could easily identify themselves and
picture themselves in that same setting.
Procedure
By means of a link, participants could access the online survey between 8 and 25 May, 2019. Participants were provided with the estimated time of duration, which was
approximately 5 minutes. All participants voluntarily chose to participate. Once people had clicked on the link, they were directed to the introduction page of the online survey. Here, participants were told that the study was about wine brands on Instagram. In the following page, participants were asked to provide informed consent, before they could head on to the beginning of the survey. Participants were informed about the privacy statements and they were informed that they could withdraw from the experiment at any time. Participants were only able to participate if they had reached the age of 18. Without their agreement,
participants weren’t able to start the survey.
First of all, participants were asked a couple of general questions about their Instagram use, their habits of drinking wine and their interest in wine brands on Instagram. These questions were meant as filler questions. After that, they were presented with an
Instagram post of the brand Wine Wizz, which portrayed a bottle of wine within two different backgrounds and with three different descriptions. Each participant was randomly assigned to one of the six experimental groups (Group 1,2,3,4,5,6).
After participants had been randomly exposed to one of the stimuli, they were asked questions related to what they had previously seen. The questions were the same for all groups, accept for one word which was manipulated according to which group they were in: real-life setting or white background. Items, such as “the real-life setting makes the product interesting”, measured on a 7-point scale (1. Strongly agree – 7. Strongly disagree) and “the white background matched the product”, measured on a 7-point scale (1. Strongly agree – 7.
Strongly disagree), were asked (see Table 1, Appendix B). These items were based upon the original scale by Harris & Goode (2010).
At this point, only the four groups which saw an Instagram post with a description answered the next set of questions, which aimed to measure the tone of voice. Items such as “the description of the post… makes me feel close to the brand” or “creates a sense of
distance from the brand”, and “the description of the post… is very personal” or “is very impersonal”, were on a bipolar matrix (see Table 1, Appendix B). These items are adapted from the scale of Barcelos, Dantas & Sénécal (2017).
Subsequently, every participant had to answer questions about their buying intention of the product seen in the Instagram post. Items such as “I can imagine myself enjoying the product”, measured on a 7-point scale (1. Strongly agree – 7. Strongly disagree), and “I would consider buying the shown product”, measured on a 7-point scale (1. Strongly agree – 7.
Strongly disagree), were asked (see Table 1, Appendix B). These items are adapted from the scale of Gageler & van der Schee (2016).
Finally, the questionnaire concluded with some demographic questions about the participant’s age, gender, nationality and their estimated level of English proficiency.
Before the end of the survey, participants were asked to answer the manipulation check, which was a question about whether they had seen either an Instagram post portraying a product in a real-life setting or a white background.
Results Randomization
In order to check if participant’s gender was comparable across the conditions with
different domains, a Chi-square test was conducted. The Chi-square test had gender as dependent variable and the condition as independent variable. The results showed that participants’ gender was not significantly different across the conditions, 2
p= .901. This means that randomization of participants across conditions was successful in
terms of participants’ gender. In order to check if participants’ age was comparable over the
different conditions with different domains, a one-way ANOVA with age as a factor was conducted. Results of the one-way ANOVA showed that participant’s mean age was not significantly different across the conditions, ƒ (1,44) = 1.576, p = .019, meaning that randomization of participants across conditions according to age was successful.
Measures
Two new variables have been computed before the analysis:
Purchase intention. Purchase intention concerning the bottle of wine portrayed in the Instagram post was assessed using three items adapted from the scale of Gageler & van der Schee (2016), answered on a 7-point Likert-scale ranging from ‘strongly agree’ (1) to
‘strongly disagree’ (7). Item examples are: “I can imagine myself enjoying the product” and
“I would consider buying the shown product”. A factor analysis with oblimin rotation
indicated that the scale was unidimensional. Initial Eigen Values indicated that the first factor had an Eigen Value of 2.40, explaining 79.98% of the variance. The second and third factor had an Eigen Value of .32 and .26, explaining respectively 11.51% and 8.50% of the variance. Results of the factor analysis showed a significant KMO of .74, meaning that there are
enough participants to run the analysis on. Next to that, the analysis showed a significant Bartlett’s Test, which means that the items are related. After the factor analysis, a reliability
test has been done, in order to make sure that the items are internally consistent. Results showed a significant Chronbach’s Alpha of .88, which proved that the scale was reliable. Therefore, a new variable has been computed, including all 3 items of the scale. The new variable was named “PurchaseIntentionMean” (MIN=1.00; MAX=7.00; M=3.44; SD= 1.36).
Tone of voice. The moderating factor concerning the tone of voice used in the description of the Instagram post was assessed using six items adapted from the scale of
Barcelos, Dantas & Sénécal (2017), answered on a 7-point Bipolar-scale. Item examples are: The description of the post… “makes me feel close to the brand” (1) “creates a sense of distance to the brand” (7) and, the description of the post… “is very personal” (1) “is very impersonal” (7). A factor analysis with oblimin rotation indicated that the scale was
unidimensional. Initial Eigen Values indicated that the first factor had an Eigen Value of 3.38, explaining 56.39% of the variance. The other five items had Eigen Values of .97, .51, .47, .37 and .30, explaining respectively 16.13%, 8.54%, 7.76%, 6.18% and 5.01% of the variance. Results showed a significant KMO of .85, telling us that there are enough participants to run the analysis on. The Bartlett’s Test was also significant, meaning that the items are related.
After that, a reliability test was done to measure internal consistency. Results showed a significant Chronbach’s Alpha of .84, which would have increased slightly to .864 if one item would have been deleted. As the increase wasn’t substantial, the item was kept. Finally, a new
variable was computed, including all 4 items of the scale. The new variable was named “ToneOfVoiceMean” (MIN=1.00; MAX=6.25; M=3.12; SD=1.27).
Manipulation
To check if participants perceived the condition they were in as intended, a Chi-square test was conducted with condition (real-life setting vs. white background) as independent variables and perceived background type as dependent variable. A significant effect of
background type was found on perceived background type, 2 = .69, 2 (1, N=246) = .57, p = .000.
95.9% of the participants who actually were in the real-life setting condition perceived to be in that condition, and 70.6% of the participants who were in the white background condition perceived to be in that condition. This indicates that the manipulation was successful.
In order to test Hypothesis 1, which proposed that an Instagram post portraying a product in a real-life setting vs. a white background would lead to a higher purchase intention of that product, an independent samples t-test was conducted with background type (real-life setting vs. white background) as the grouping variable and purchase intention as a test variable. The results showed a significant effect of background type on purchase intention, t (244) = -3.64, p= .000. The direction of the effect was opposite to the prediction: participants had higher intentions to purchase the product when they were presented with the white background (M= 3.74, SD= 1.29) than when they were presented with a real-life setting (M= 3.12, SD= 1.37). This means that Hypothesis 1 can be rejected: An Instagram post portraying a product in a real-life setting doesn’t lead to more purchase intention of that product
compared to an Instagram post portraying a product on a plain, white background.
Hypothesis 2 was tested with a two-way analysis of variance (ANOVA), with type of background (real-life setting vs. white background) as independent variable, purchase
intention as dependent variable and tone of voice used in the description of the Instagram post (conversational human voice vs. neutral tone of voice vs. no tone of voice) as fixed factor. The hypothesis proposed that an Instagram post promoting a product in a real-life setting vs. a white background would lead to more purchase intention of that product, (a) with the effect being stronger for people who were presented with a post containing a description with a conversational human voice compared to those who were presented with a post containing a neutral tone of voice; (b) with the effects being stronger for people who were presented with a post containing a description with a neutral tone of voice compared to those who were
presented with a post containing no description at all. Results showed a significant interaction effect between background type and tone of voice used in the description of the post on
purchase intention, ƒ (2, 239) = 5.69, p= .004, 2 = .05 (see Table 1, Appendix C) In the real-life setting condition, participants who were presented with the Instagram post containing a
conversational human voice had a higher purchase intention (M= 3.37, SD= 1.64) compared to those who were presented with the Instagram post containing a neutral tone of voice (M= 2.73, SD= 1.13). On the other hand, participants who were presented with the Instagram post containing no description at all showed a higher purchase intention (M= 3.27, SD= 1.24) compared to those who were presented with the Instagram post containing a neutral tone of voice (M= 2.73, SD= 1.13).
However, in the white background condition this tendency was different. More
specifically, purchase intention was higher when description of the Instagram post contained a neutral tone of voice (M= 4.19, SD= 1.42) compared to when the description contained either a conversational human voice (M= 3.52, SD= 1.17) or no description at all (M= 3.61, SD= 1.23) (see Figure 7, appendix D). Overall, this means that both hypothesis 2a and hypothesis 2b can be rejected: the description of an Instagram post containing a conversational human voice vs. a neutral tone of voice leads to a higher purchase intention when a product is placed in a real-life setting, but not compared to when a product is placed in a white background. No description of the Instagram post leads to higher purchase intention of the product placed in a real-life setting compared to a neutral tone of voice used in the description.
Discussion and conclusion
The purpose of this research was to examine whether a particular background
displayed with a product would influence subsequent purchase intention of that product, and whether the tone of voice used in the description of the Instagram post would moderate this effect.
Surprisingly, results revealed the opposite to what was the expected outcome: a white background demonstrated to have a higher effect on subsequent purchase intention of the displayed product compared to a real-life setting. Next to that, in the real-life setting group, a conversational human voice (vs. neutral tone of voice) adopted in the description of the post
did indeed lead to higher consequent purchase intention of the displayed product, but a neutral tone of voice adopted in the description of the post didn’t lead to higher consequent purchase intention compared to no description at all. In fact, no description at all showed to be more effective than a neutral tone of voice adopted in the description of the post. On the other hand, in the white background group, a neutral tone of voice used in the description lead to more purchase intention compared to both a conversational human voice as no description at all.
Besides not confirming the hypothesis, this research shows an unexpected turning point: results suggest that the assumed moderator, tone of voice used in the description of the Instagram post, might be the variable influencing directly consequent purchase intention. This way, the moderator becomes the independent variable and the assumed independent variable (background type) becomes the moderator. In fact, the tone of voice adopted had a different effect on purchase intention depending on the type of background it is accompanied with.
The assumption that the background type would influence consequent purchase
intention was based upon previous research. For instance, the study of Harris & Goode (2010) demonstrated how people’s purchase intention was positively influenced by attractive
background features and the study of Breazeale & Ponder (2011) showed that particular background features enhance people to get a good idea of how a certain product should properly be used in the environment. In relation to these results, current hypotheses were formed, but unfortunately weren’t confirmed. This means that results of this study aren’t
comparable with previous studies.
Current results can be explained by various reasons. First of all, no previous research had investigated the combination of background type, chosen to display a certain product in, with tone of voice, used in the description of the Instagram post. In fact, this research had the intent to take a closer look at one of the most popular social media of the moment, Instagram, and investigate the influence of brands on this social medium. More specifically, how brands
can trigger purchase intention of its followers by means of a certain background type
accompanied with a certain description of the product portrayed in their Instagram posts. The fact that current research investigated a topic which wasn’t fully investigated previously could explain the reason why these results were unpredictable.
Second of all, participants could have preferred the white background due to the fact that it wasn’t disturbing or distracting them from the product displayed in the Instagram post.
Hence, the wine bottle placed in the real-life setting could have directed participant’s attention mainly to the background elements and not to the product, leading to a lower memory of the product and thus to less purchase intention. Besides that, heuristic use of the background by participants could also explain the fact that participants seeing the Instagram post with a product placed in a real-life setting, accompanied by a description containing a conversational human voice showing lots of emoticons in the text, could have been influenced positively by the visual of the emoticons, having a positive effect on consequent purchase intention. This could also be the reason why participants in the white background group had more intention to purchase the shown product when the description contained a neutral tone of voice: as they were fully paying attention to the product and what was written in the description of the post.
Third, it could be that the description of the Instagram post is more important to people, in the sense that they can get a better idea of what kind of product is shown in the Instagram post and how to experience it. On the other hand, it gives brands the opportunity to be creative in writing an appropriate text which enlightens the costumer of the value of the product and how the product should be experienced. This is more important for a customer.
Limitations and directions for future research
This study had several limitations. First of all, as it was an online experiment, it was impossible to monitor the participants in order to see if they were following the instructions of the experiment. As mentioned before, participants were directed to the experiment by clicking
on a link. At this point they were left alone in conducting the entire experiment. This leads to the fact that the environment in which they conducted the experiment might not have been the right one, as it could have contained a lot of distracting factors. This could definitely explain the elevate number of respondents who abandoned or didn’t finish the experiment before it
had reached its end. Next to that, participants could have had the wrong attitude while answering the questions. Overall, it expresses a weak external validity. A solution for this could be the replication of this study in a controlled setting, where researchers can follow the entire experiment.
Second, the term “real-life setting” is a vast understanding: it could be everything. The real-life setting chosen to include in this study, as a background type, was considered to be as much related to the product shown in the Instagram post as possible. Since the product was a bottle of white wine, it made sense to portray it in the backyard of someone’s house, showing the green of the grass, the wooden table at which people often sit outside to have a drink and the clear blue sky which gives a warm summery faeling. The limitation of this setting is that not everybody feels attracted to this kind of real-life setting and not everybody considers this setting to be nice. More precisely, people’s tastes are objective, also when it comes to
preference for a certain background. Future research should therefore replicate this research and see if different real-life settings have different effects on people.
Third, in the groups where participants were presented with a bottle of white wine on a white background, a tiny part of the wooden table was visible. The bottle of wine was
standing on that table. It could be questionable whether this table is considered to be part of the background, influencing people’s perception of the white background in which the
product is placed. The reason for keeping the wooden table for the Instagram post with the white background, was to keep both conditions as similar to each other as people, except for the background. Furthermore, this choice made the white background in which the product
was placed, less fake. A solution to this limitation could be to place the bottle of wine on a white table instead of the wooden table, that way the perception of the white background could be as a whole and it is possible to rule out any other influencing factors of the likeability of the background.
Lastly, it would be interesting for future research to find out the reasons why results turned out the way they did. For instance, it could be possible to run a qualitative research (e.g. interviews) in order to get a closer look at the reasons behind the preference of a certain background in relation to a tone of voice used in the description of the Instagram post
portraying a certain product. finally, it would be interesting to see if there are differences in results in relation to different products.
References
Ariffin, H., Bibon, M., Abdullah, R. (2012). Restaurants Atmospheric Elements: What the Customer Wants. Procedia – Social and Behavioral Sciences 38, 380-387.
Babin, B.J., Attaway, J.S. (2000). Atmospherics affect as a tool for creating value and gaining share of customer. Journal of Business Research, Vol.49(2), 91-99.
Barcelos, R. H., Dantas, D. C., & Sénécal, S. (2018). Watch your tone: How a brand's tone of voice on social media influences consumer responses. Journal of Interactive
Marketing, Vol.41, 60-80.
Breazeale, M., Ponder, N. (2011). Get the picture? Visual servicescape and self-image congruity. Journal of Business Research, Vol.66, 839-846.
Bridges, E., Florsheim, R. (2008). Hedonic and utilitarian shopping goals: The online experience. Journal of Business Research, Vol.6, 309-314.
Carah, N., Shaul, M. (2016). Brands and Instagram: Point, tap, swipe, glance. Mobile Media
& Communication, Vol. 4(1), 69-84.
Cheng, F., Wu, C., Yen, D. (2009). The effect of Online Store Atmosphere on Consumer’s Emotional Responses – An Experimental Study of Music Behavior. Behavior &
Information Technology Vol. 28, No. 4, 323-334.
De Keyzer, F., Dens, N., De Pelsmacker, P. (2017). Don’t be so emotional! How tone of voice and service type affect the relationship between message valence and consumer responses to WOM in social media. Online Information Review, Vol.41, 905-920. Fowler, K., Bridges, E. (2012). Service environment, provider mood, and provider-customer
interaction. Managing Service Quality, vol.22(2), 165-183.
Gageler, L., van der Schee, J. (2016). Product placement on social media: A study on how generation Y’s brand perception and purchase intention are influenced. Jönköping
Gee, R., Coates, G., Nicholson, M. (2008). Understanding and profitably managing customer loyalty. Marketing Intelligence and Planning, Vol.26(4), 359-374.
Ginsberg, K. (2015). Instabranding: Shaping the personalities of the top food brands on Instagram. The Elon Journal of Undergraduate Research in Communications, Vol.6 (1), 78-91.
Goor, M. (2012). Instagram: a content analysis into marketing on Instagram. Master thesis,
University of Amsterdam, Department of Communications, Amsterdam.
Grabner-Kräuter, S., Waiguny, M.K. (2015). “ ‘I believe more in factual reviews’ – but not so much when the reviewer is similar to the reader and the product is hedonic”, in
Bušljeta Banks, I., De Pelsmacker, P., Okazaki, S. (Eds), Advances in Advertising
Research, Springer Fachmedien, Wiesbadem, 89-102.
Hanna, R., Rohm, A., & Crittenden, V. L. (2011). We’re all connected: The power of the social media ecosystem. Business horizons, 54(3), 265-273.
Harris, L.C., Goode, M.H. (2010). Online servicescapes, trust, and purchase intentions.
Journal of Services Marketing, vol. 24(3), 230-243.
Instagram (2019). Making Explore an even better home for your interests. Retrieved from: https://instagram-press.com/blog/2019/05/16/making-explore-an-even-better-home-for-your-interests/
Jaakonmäki, R., Müller, O., vom Brocke, J. (2017). The impact of content, context, and creator on user engagement in social media marketing.
Keeling, K., McGoldrick, P., Beatty, S. (2010). Avatars as salespeople: Communication style, trust, and intentions. Journal of Business Research, Vol.63, 793-800.
Kelleher, T. (2009). Conversational voice, communicated commitment, and public relations outcomes in interactive online communication. Journal of Communication, Vol.59(1), 172-188.
Kim, J., Kim, M., Kandampully, J. (2009). Buying environment characteristics in the context of e-service. European Journal of Marketing, Vol.43(9-10), 1188-1204.
Kotler, P., Zaltman, G. (1971). Social marketing: An approach to planned social change.
Journal of Marketing, Vol.35(3), 3-12.
Kotler, P. (1973). Atmospherics as a marketing tool. Journal of Retailing, Vol.49(4), 48-64. Lingling, G., Xuesong, B. (2014). Online Consumer Behavior and its relationship to website
atmospheric induced flow: insight into online travel agencies in China. Journal of
Retailing and Consumer Services, Volume 21, issue 4, 653-665.
Marinus, H., Anggraini, L. (2018). The effect of sales promotion in social media on the students: purchase intention of face cleaner water product. Humaniora, Vol.9 (1), 15-22.
Michael Hall, C. (2008). Servicescapes, designscapes, branding, and the creation of place identity: South of Litchfield, Christchurch. Journal of Travel & Tourism Marketing,
Vol.25(3-4), 233-250.
Morwitz, V. (2012). Consumers’ purchase intentions and their behavior. Foundations and
Trends in Marketing, Vol. 7(3), 181-230.
Nambisan, S., Baron, R.A. (2007). Interactions in virtual customer environments implications for product support and customer relationship management. Journal of Interactive
Marketing, Vol. 21(2), 42-62.
Phua, J., Jin, V.S., Kim, J.J. (2016). Gratifications of using Facebook, Twitter, Instagram, or Snapchat to follow brands: the moderating effect of social comparison, trust, tie strength, and network homophily on brand identification, brand engagement, brand commitment, and membership intention. Telematics and Informatics 34, 412-424.
Ramanathan, U., Subramanian, N., Parrott, G. (2017). Role of social media in retail network operations and marketing to enhance customer satisfaction. International Journal of
Operations and Production Management, Vol.37(1), 105-123.
Roncha, A., Radclyffe-Thomas, N. (2016). How TOMS’ “one day without shoes” campaign brings stakeholders together and co-creates value for the brand using Instagram as a platform. Journal of Fashion Marketing and Management, Vol. 20(3), 300-321. Schwemmer, C., Ziewiecki, S. (2018). Social media sellout: The increasing role of product
promotion on YouTube. Social Media + Society, 1-20.
Van Noort, G., Willemsen, L.M. (2011). Online damage control: The effects of proactive versus reactive webcare interventions in consumer-generated and brand-generated platforms. Journal of Interactive Marketing, Vol.26, 131-140.
Vilnai-Yavetz, I., Rafaeli, A. (2006). Aesthetics and professionalism of virtual servicescapes.
Journal of Service Research, vol. 8(3), 245-259.
Wallsbeck, F. E., Johansson, U. (2014). Instagram marketing: when brands want to reach generation Y with their communication. Halmstad University, School of Business and
Engineering, 1-91.
Weise, S. (2015). Instagram’s potential for brands: We are social. Retrieved from
Appendices Appendix A
Figure 1. Product displayed in a real-life setting with a conversational human voice in the description
Figure 2. Product displayed in a real-life setting with a formal tone of voice in the description
Figure 4. Product displayed on a white background with a conversational human voice in the description
Figure 5. Product displayed on a white background with a formal tone of voice in the description
Appendix B
Table 1. Dependent variables, independent variables and moderator
Variable Scale Items Chronbach
Alpha Purchase Intention Gageler & van der
Schee (2016)
How much do you agree with the following statements?
…I can imagine myself enjoying the product …I am interested in getting more information about the product
…I would consider buying the shown product (1 to 7: strongly agree – strongly disagree)
.875
Real-life setting Harris & Goode (2010) Do you agree or disagree with the following statements?
…The real-life setting fits the product displayed in the Instagram post
…The real-life setting makes the product interesting …The real-life setting gives me a good idea of how the product should be used
…I like the real-life setting in which the product is placed
…The real-life setting makes me experience the product better
…The real-life setting matches the product (1 to 7: strongly agree – strongly disagree)
.907
White background Harris & Goode (2010) Do you agree or disagree with the following statements?
…The white background fits the product displayed in the Instagram post
…The white background makes the product interesting
…The white background gives me a good idea of how the product should be used
…I like the white background in which the product is placed
…The white background makes me experience the product better
…The white background matches the product (1 to 7: strongly agree – strongly disagree)
.908
Tone of voice Barcelos, Dantas & Sénécal (2017)
The description of the post…
… (1 to 7: makes me feel close to the brand – creates a sense of distance to the brand)
… (1 to 7: helps me experience the product better – doesn’t help me experience the product better) … (1 to 7: fits the display of the product – doesn’t fit the display of the product)
… (1 to 7: is very personal – is very impersonal) … (1 to 7: is very professional – is very unprofessional)
… (1 to 7: makes the brand very attached to its audience – makes the brand very detached from its audience)
Appendix C
Table 2. Relationship between background type and tone of voice on purchase intention Source Type III
sum of squares
df Mean Square F Sig. Partial Eta Squared Corrected Model 44.966a 6 7.494 4.368 .000 .099 Intercept 260.555 1 260.555 151.876 .000 .389 Background Type 25.882 1 25.882 15.087 .000 .059 Tone of Voice Type 2.077 3 .692 .404 .751 .005 Background Type* Tone of Voice Type 19.525 2 9.763 5.691 .004 .045 Error 410.023 239 9.716 Total 3362.111 246 Corrected Total 454.989 245 Appendix D
Figure 7. Moderating effects of tone of voice on the main effect of background type on purchase intention
0 1.000 2.000 3.000 4.000 5.000 conversational human voice
neutral tone of voice no tone of voice
Tone of voice moderating the effect of
background type on purchase intention
real-life setting white background
Pu rch ase in ten tio n me an s
