How effective are covert vs. overt strategies and the inclusion of a face in Instagram posts and does their effectiveness depend on the post being shared by an established or an
unestablished brand?
Master’s Thesis – Graduate School of Communication Persuasive Communication
Jeanne Bommeljé 11401818
Supervisor: Anne Kranzbuhler Date of completion: 26th of June 2018
Abstract
This research delves into the effects of covert vs. overt advertising and the effects of the inclusion of a face in Instagram posts, and studies whether the strategies may be more or less effective for established or unestablished brands. This was done in an attempt to uncover how specific content elements might affect engagement for particular types of brands (i.e.
established vs. unestablished brands), thereby contributing to an Instagram formula that brand managers will be able to use to increase engagement. This is a crucial matter since consumers are no longer passive recipients of advertisements and their engagement is nowadays highly valued by brands. A content analysis was done in order to carry out a close examination of the elements of two hundred Instagram posts, including both images and captions. The results revealed that brands do not tend to use overt images, while overt captions did not prove to generate any less engagement than covert captions. The research also did it confirm any differences between the effectiveness of overt vs. covert strategies when employed by an established or an unestablished brand. The research also did not confirm any differences between the engagement on images that did or did not include a face, nor did it confirm any significant differences between the generated engagement when an established or an
unestablished brand shares images that include a face. The data did prove that there is a significant difference between engagement on a post when it is shared by an established or an unestablished brand. Having significantly less followers may therefore make unestablished Instagram accounts seem more personal, thereby stimulating followers to leave a comment and/or like the posts. Engagement with Instagram is likely to spill over into the way in which consumers engage with advertisements within the platform, which in turn affects
advertisement evaluations. Brand managers should therefore take into account that
engagement is context specific and consumers engage with Instagram differently compared to other types of media based on its unique characteristics.
Introduction
Instagram is a very popular social network site, particularly among younger consumers (Bakhshi et al., 2014; Latiff & Safiee, 2015). The application allows its users to share images and engage with one another by liking images or by leaving comments. Since its launch it now has over 500 million daily users and over 800 million monthly users (Instagram, 2018), which shows that there is a growing appeal to this type of visual social media.
This platform is becoming increasingly important to brands wishing to gain more popularity. This is because of the high number of users of the platform, as well as the ability to post images that tell something about the brand, allowing brands to reveal a personal touch. This personal touch may help create senses of intimacy with consumers and helps build relationships with those who follow the brand (Hellberg, 2015), thereby potentially increasing the popularity of the brand as well as their engagement.
Consumers are no longer passive recipients of advertisements and their engagement is nowadays highly valued by brands (Hellberg, 2015). Engagement in turn captures
consumers’ activation (Vinerean & Opreana, 2015) and it is reflected by the number of comments and likes on an Instagram post, as well as the number of followers.
The medium is primarily used as way to be entertained, meaning that advertisers ideally need to take this into account before confronting consumers with advertisements on a platform that they use to pass time (Voorveld et al., 2018). Especially in a time where
consumers are becoming more critical and irritated by advertisements (Göbel et al., 2017), brands may want to downplay their blatant marketing messages and try to appeal to the values of the Instagram community. To do this, they need to disguise commercial aims and create posts that fit in with native looking content (Beverland, 2005). This is the practice of covert advertising, whereby the aim to persuade is not as clear. Covert advertising is therefore a sneaky way of advertising as it occurs when consumers are exposed to a subtle
advertisement (Göbel et al., 2017). An example would be when a brand just shows a picture of model wearing their brand, without texts that directly tell the consumers to go to the store.
The opposite of this strategy is overt advertising, whereby the aim to persuade is clearly communicated (Schrindler et al., 2017). It occurs in the form of a typical blatant advertisement whereby there is no doubt that the message communicates commercial aims. An example would be a post that says “Go to the website to shop the whole collection” or “Buy now at www. …”.
Both strategies have their benefits and downsides but it remains unclear which of these strategies (covert or overt) generate a higher engagement on Instagram depending on the brands that employ them. It could be that there is a difference between new brands and established brands because of the ways in which they are perceived by consumers, making it more or less effective for the different types of brands to employ overt or covert advertising on Instagram. This is an important matter because Instagram is an excellent marketplace for both newcomers as well as for established brands (Latiff & Safiee, 2015), while the
difference between the effectiveness of their advertising strategies has not yet been investigated.
Although established brands are more likely to be recognized and tend to be more popular (Muthukrishan, Wathieu & Jing Xu, 2009), this does not mean that unestablished brands necessarily have less engagement since being an established brand is not related to the number of likes on an Instagram account (Voorveld et al., 2018).
The use of human faces in Instagram posts is also said to increase the generated engagement (Bakhshi, Gilbert & Shamma, 2014) as a face is said to increase identification and draw consumers to a post (Walter, 2011). Brands however also tend to share pictures without a face, such as a picture of a quote, a scenery, an animal, food etc. Such images may benefit from sharing an image of something relatable that does not necessarily have anything
to do with the brand (du Plessis, 2017). A question yet to be answered however concerns whether the use of a face is more or less effective for established or unestablished brands. This is worth investigating since research can provide insight on the effectiveness of particular types of content (a face vs. no face) for particular types of brands (established vs. unestablished brands).
With a focus on fashion, it is interesting to see which strategies are often opted by the two types of brands because Instagram is a platform that is most spammed by fashion
(Eriksson & Hansson, 2016). This means that there is an extremely high competition in this genre on Instagram. If we therefore know what the different types of fashion brands do to be successful on Instagram, then we may be able to build upon a successful Instagram marketing formula.
The main question to be answered therefore concerns whether there is a difference in the effectiveness of the types of strategies (overt vs. covert and face vs. no face) depending on the brand (established vs. unestablished) that shares such posts on Instagram. If it could be confirmed that certain types of strategies work better for particular types of brands, then brand managers can use such insight when aiming for a higher engagement. Understanding how specific content might affect engagement can impact design, thereby influencing the way in which brands appear on the platform of Instagram. Engagement is an essential element to photo sharing platforms such as Instagram, which is why it is critical to understand what forms of content drives engagement for what types of brands.
Covert vs. Overt advertising
Covert advertising ensures that the brand is promoted, while the persuasive attempt is not as clear (Muthukumar, 2013). Such posts do not look like commercial content and may
Evans, 2015; Moyer-Gusé, 2008). They usually integrate the advertisement into
entertainment, making the advertisement seem less like an advertisement and more like the surrounding media (Wojdynski & Evans, 2015). An example would be an image of a model wearing a particular brand, with a quote as a caption instead of a commercial message.
Overt marketing strategies in turn clearly communicate a persuasive attempt at
stimulating consumers to take action with regards to a brand (Schrindler et al., 2017). Such a strategy commonly aims to directly convince consumers to buy a brands’ product(s). To be able to do this, the brand gives out (new) product information and applies intrusive strategies to increase product awareness (van den Putte, 2002). Such a strategy therefore has a strong product focus and/or uses phrases such as “buy now”, “visit the website”, “now available at…” etc.
Typically, consumers try to guard themselves against advertisements because they do not like to be persuaded (Brown & Krishna, 2004). Consumers have a need for freedom to choose their own attitudes and behaviors, which is why people become reactant when
noticing that someone tries to persuade them (Moyer-Gusé, 2008). The persuasion knowledge model also holds that consumers try to resist being influenced by advertisers, making them skeptical about advertisements and distrusting of the media. Hence, when consumers sense that there is an attempt to persuade, they tend to react defensively (Mohr et al., 2001).
Research also showed that persuasiveness decreased when the motive was perceived as purely aiming at increasing sales. When the motive was considered as caring about
consumers’ interest however, persuasion effectiveness increased (DeCarlo, 2005; Main et al., 2007).
Covert tactics in turn try to circumvent such consumer skepticism by being disguised as a neutral message (Göbel et al., 2017). Covert advertising is therefore the sneaky version of advertising and occurs when consumers are exposed to a subtle advertisement, thereby
creating advertisements that have the appearance of something other than an advertisement (Wojdynski & Evans, 2015). Such sneaky advertisements then diminish viewers’ perception that it is intended to persuade, making consumers more accepting of the message (Moyer-Gusé, 2008). When consumers are more receptive they are also more likely to evaluate the messages more favorably (Muthukumar, 2013), thereby potentially being more likely to press the ‘like’ button on a covert Instagram post. Although this type of advertising is subtle in nature, it has a positive impact on consumers because they seem very natural and non-commercial compared to other forms of promotion (Muthukumar, 2013).
Native looking content, as reflected in covert advertising, also decreases the chance that users will avoid or ignore the posts because consumers do not immediately recognize the persuasive attempt (Wojdynski & Evans, 2015; Moyer-Gusé, 2008). Research showed that such covert advertisements are visually more attractive and gain more attention than commercial messages by offering appealing visual presentations (Tomažic et al., 2014).
This strategy is the opposite of overt advertising which alerts consumers and makes it clear that marketing tactics are employed, thereby activating consumers’ persuasion
knowledge. This may in turn negatively affect the way consumers respond to the brand (Göbel et al., 2017), whilst still ensuring that consumers are not deceived by sneaky tactics. Overt strategies, although recognized as clear persuasion, could still benefit from consumers appreciating the honesty. Consumers dislike being deceived by clever advertisements (Göbel et al., 2017), meaning that they may appreciate the straightforward approach.
Instagram is however intended to be a medium that allows its users to seek entertainment, socially interact and spend spare time (Voorveld et al., 2018). The given research will therefore examine whether the contentions above also apply to this specific platform. No similar line of work has yet provided insight on how specific overt or covert Instagram content can affect aspects of online behavior, such as engagement. When
confronted with commercial content on an entertainment platform, it could however be expected that consumers tend to skip through it as soon as possible and reject the content, thereby evaluating commercial posts as less likable than personal-looking posts. Overt advertisements may even be seen as a disruption of the content flow, making consumers scroll past it and ignore it (Voorveld et al., 2018). This research can test these gaps in knowledge with the use of the following assertion:
H1: Posts on Instagram generate higher engagement when employing covert advertising than when employing overt advertising.
Human face
If a feeling of similarity with the post is high, then this may result in identification with the post and the brand (Moyer-Gusé, 2008). When identification is high, then consumers are more likely to be appealed by the post and may engage with it by liking the picture or by leaving a comment. Identification may in turn happen when the post includes a face because a human face attracts a sense of familiarity among consumers (Walter, 2011). This may be a face of someone famous or the face of a user of the brand1.
People constantly explore their surroundings by looking for something familiar. A face provides this sense of familiarity and therefore draws attention, which is why a face may be a powerful aspect of an Instagram post (Walter, 2011). The catching of attention is crucial on Instagram since consumers are often overloaded by the amount of posts that are added to the platform every minute of the day (Instagram, 2018; Steven, 2016). Familiarity also provides a feeling of comfort while a face adds personality to the application, potentially
1 Effects of influencers vs. non-influencers are not studied in this research. For this part, a focus will only be placed on human faces.
making users feel like being in a social setting instead of on the internet (Walter, 2011). Faces therefore also have the potential to make users feel engaged.
People are automatically triggered to feel something when they see a face since faces provide a human touch that triggers emotions, making it a powerful design element. By using emotional expressions, the feelings attributed to the use of a brand can be visualized. An image of someone expressing their feelings is much clearer and more easily noticed than a piece of text. Faces then also create a particular feel, allowing the transfer of emotions from the picture to the viewer, while faces also have the potential to make a reflective imprint on peoples’ long-term memories (Walter, 2011).
Not all Instagram posts as shared by brands use faces. Brands nowadays also share pictures of a quote, a landscape, a puppy, food etc. This allows brands to share their interest and try to engage with consumers who may be appealed by the same things. An image of for instance a relatable quote may make the brand seem reachable and personal, thereby drawing upon the positive aspects of authenticity, while the senses of reachability and personality stimulate engagement (Vinerean & Opreana, 2015). This may incite consumers to like or comment on a post.
Another way in which brands purposely do not use a face is when they share close-ups of their products, without a model involved. Such posts are purely product focused and emphasize the appearance and the quality of the brands’ items.
A previous study found that people are more drawn to an Instagram post that includes a picture of a human face. Out of the 1.1 million pictures that were analyzed in this study, pictures with a face were 38% more likely to receive likes and 32% more likely to receive comments, compared to pictures without a face (Bakhshi, Gilbert & Shamma, 2014). These findings will be reexamined by testing the following contention:
H2: Posts on Instagram generate higher engagement when sharing images that include a face compared to images that do not include a face.
Established vs. unestablished brands
Established brands are brands that have existed for a longer time. In this research, a “longer time” will be interpreted as being twenty years old or older2. Since they have existed for so long, consumers have established associations within their minds when thinking about established brands (Sjodin & Törn, 2006). Established brands are also generally perceived as having a higher quality, even if some of their attributes may be inferior to that of competing, unestablished brands (Muthukrishan, Wathieu & Jing Xu, 2009). Their quality beliefs are held with such confidence because they are familiar to consumers. Established brands are therefore also considered to be market leaders within the fashion industry (Jun et al., 2015).
Unestablished brands are those yet to establish themselves within consumers’ minds (Lei Shi, 2014). They are therefore new and can be recognized as start-ups. In this research, “new” will be characterized as being three years old or younger3. There is a general
consensus which holds that brands that are less privileged in the market place, which usually tend to be the unestablished brands, are associated with feelings of friendliness and those with more power (i.e. established brands) are associated with being dominant (Prell, 2011; Jun et al., 2015). This means that such unestablished brands may benefit from advantages of
2 No research has yet set boundaries as to what it means to be an “older” brand. Twenty years however seems like long enough for the brand to develop itself and become recognized by consumers.
3 No research has yet set boundaries as to what it means to be a “new” brand. Three years however seems like young enough for consumers to not typically be familiar with the brand.
being perceived as an ‘under-dog’ brand, that struggles against ‘top-dog’ brands despite their competitive disadvantage (Jun et al., 2015). Top-dog and under-dog positioning therefore comes back in established brands being the market leaders and unestablished brands being new and yet to be developed.
It could be that there are certain matches between types of brands (established vs. unestablished) with types of advertisements (covert vs. overt). This fit between the type of brand and type of advertisement may in turn influence the generated engagement on a post.
Attitudes towards new and unestablished brands are linked to feelings of warmth and empathy (Jun et al., 2015). A feeling of warmth and empathy may in turn be better reflected by an unobtrusive advertisement which stimulates entertainment and aims to engage with consumers (Wojdynski & Evans, 2015), compared to an advertisement that emphasizes commercial aims such as “go to the site” or “buy now”. Unobtrusive, subtle advertisements in turn do not treat consumers as potential buyers, but instead, treat them like followers that are part of an engaging network (Göbel et al., 2017). This type of subtle, covert
advertisement may be a better match with unestablished brands because it suits their under-dog, friendly profile.
Prior research also proposes that perceptions of authenticity increase brand preference and influence positive responses to an advertisement (Chalmers & Price 2009). Authenticity may in turn be more easily provoked by unestablished brands because they are more easily associated with genuineness, while established brands are more likely to be associated with mass marketing (Jun et al., 2015). Because of favorable perceptions of authenticity that are linked with unestablished brands, it could be that such brands are more successful when sharing native, authentic, genuine covert advertisements, compared to their commercial established counterpart. The personal nature of unestablished brands may therefore match
with covert advertisements and make covert advertisements more successful when shared by a brand that seems authentic.
Research also showed that advertisements as spread by unestablished brands are more likely to directly create certain associations because unfamiliar brands do not yet have set links within the minds of consumers (Miller, 2015). This means that it may be even more important for unestablished brands to not only be associated with sales attempts but also be primarily linked with other types of content, as consumers are getting more critical and irritated by clear advertisements (Göbel et al., 2017). When an unfamiliar brand therefore successfully shares entertaining, non-sales related posts that show interest in consumers, consumers may decide to like and/or start following the brand based on their positive first impressions.
Researchers provided alternative justifications for advertising effectiveness of established brands. Effectiveness could therefore go two ways when, for instance, an established brand tries to come across as personal and sincere, thereby not directly stimulating sales but instead, aiming to entertain consumers with the use of a covert
Instagram post. Some claim that brand attitudes for such established brands can be changed by manipulating brand beliefs (Miller & Allen 2012). Hence, it could be that subtle,
entertaining, covert advertisements amplify an emotional response, thereby stimulating positive attitudes (Sjodin & Törn, 2006). Others hold that the perceptions of established brands are unlikely to be influenced by particular covert advertisements because consumers already have stable associations in mind with regards to the familiar brand (Machleit et al, 1993). Such stable associations are in turn difficult to change (Sjodin & Törn, 2006).
Additionally, when a brand that is considered to be a mass advertiser tries to come across in a personal manner, consumers may be distrusting of the advertisement and consider the
Consumers trust in established brands may also be lower due to associations with their commercial interests (Jun et al., 2015), making it harder for such brands to come across in a personal manner. As previously discussed, it is important for brands to come across as sincere and genuine in order to increase the brands’ persuasiveness. The persuasiveness of a covert post is however expected to be less successful when a brands’ motive is considered to be commercial (DeCarlo, 2005; Main et al., 2007). This makes it more challenging for established brands to be successful in sharing covert posts on Instagram. Associations with mass marketing and with being a leading brand (Jun et al., 2015) do however mean that established brands may be a better match with overt advertisements that clearly stimulate consumers to buy their products.
To test the contentions that there may be a better fit between certain types of brands (i.e. established vs. unestablished brands) with certain types of advertisements (i.e. covert vs. overt advertising), the following hypotheses are proposed:
H3a: Covert advertising generates higher engagement for unestablished brands than for established brands.
H3b: Overt advertising generates higher engagement for established brands than for unestablished brands.
With regards to the use of faces in Instagram posts, it would be interesting to know whether the use of a human face is more effective for established or unestablished brands. It could very well be that the use of a face is more effective for an unestablished brand as the established brand already has a familiar name and/or logo. The unestablished brand therefore struggles more to make their posts noticeable, and a face, as argued before, has this potential to draw attention (Walter, 2011). The established brand may generate higher engagement when posting a picture that does not include a face since their name alone has a greater ability to catch attention compared to an unfamiliar name.
Furthermore, consumers may engage with pictures of a face when shared by an unestablished brand because they are connecting with the face as a brand they feel closer to. Unestablished brands usually have a lower number of followers (Muthukrishan, Wathieu & Jing Xu, 2009), meaning that their posts are shared with a smaller crowd. When an
unestablished brand therefore shares an image that includes a face, consumers may be triggered to feel something because of the more intimate and personal feel to an
unestablished brand. When shared by an established brand with an immense network of followers, a face may appear more as a generic object (Jun et al., 2015), instead of as a face that they feel closer to. Consumers may then feel more inclined to engage with a post that they feel closer to compared to a post that feels more distant. To examine these contentions, the following assertions are tested:
H4a: Instagram posts that include a face generate higher engagement for unestablished brands than for established brands.
H4b: Instagram posts that do not include a face generate higher engagement for established brands than for unestablished brands.
Methodology
This thesis delves into the characteristics of a shared photo. It does not look at the kind of comments, the timing of the posts, carrousels, videos or Instagram stories. Instead, it is based on pictures alone. This way, the types of picture-posts shared by the types of brands are subject of close examination, delving into textual and photographic representations and how these express covert and/or overtness. A content analysis is therefore suitable for the given research since it makes use of a close study of the details and content (Krippendorff, 2004). Both pictures and captions are analyzed separately since it could for instance be that an image is not necessarily overt while the text is.
For this study, twenty established and twenty unestablished brands are randomly selected. Established brands are randomly selected by browsing through retailer websites such as “de Bijenkorf” and “Zelando” and verifying whether the brands are 20 years old or older. Unestablished brands are randomly selected by searching for headings such as “the newest brands of 2016/2017/2018” and/or “upcoming brands of 2017/2018”. By then looking at which brands these brands follow, more unestablished brands are gathered. It is then also verified whether these brands are 3 years old or younger. All brands operate in the fashion industry.
Out of the twenty gathered brands, ten brands of each category have been randomly selected by putting the names in a random selection generator. This lead to the following brands to be analysed for this research4:
Table 1.
The analysed brands of this research
Unestablished brands Established brands Ovelia Transtoto (2016) – UK Minimum (1997) DE
Kitri (2017) UK H&M (1927) SW
Les favorites (2017) – NL ZARA (1974) SP
Bellerose (1989) BE Mango (1984) SP
Noumenon (2016) – NL Forever21 (1984) US
Premme (2017) – US Calvin Klein (1942) US
Elliette (2016) US C&A (1841) NL
Hola Couture (2017) BE Maison Scotch (1885) NL
Elvine (2016) SW G-Star RAW (1986) NL
Fabienne Chapot (2016) – NL Tommy Hilfiger (1951) – US
Ten random posts of each page are analysed. To ensure that the investigated posts are randomly chosen, a guideline is set by picking every 8th post that does not contain a video or a carousel. In case none of the analysed photos of a particular brand contain ‘a face’ or ‘no face’, while it can be seen that the brand does use photos with/without a face, three random photos with/without a face are added to the analyses of the brand in question by selecting the first three images that do/do not include a face, as shared by the brand. This is done to ensure that relevant data is collected in an unbiased manner. The pages can be accessed through Instagram.
With regards to the coding, covertness can be recognized by not offering any information on sales or prices. It may occur in the form of content marketing whereby
content is shared that does not explicitly promote a brand, but instead, is intended to stimulate interest (du Plessis, 2017). Such content aims to enhance consumers’ daily life and aims to be engaging, and can pertain to fashion as well as to any other topics, whilst disguising or not having commercial persuasive aims. Covert images can therefore include:
• An image of a quote, a landscape, an animal, a person, food, clothes etc. • Exclusive content
o A behind-the scenes photo o A day-in-the-life-of photo
An image is not necessarily considered as covert when a picture of a person appears to be spontaneous. Both covert and overt images may make use of spontaneous-looking images as well as staged photos. Staged photos may however also take on the looks of a spontaneous photo in order to make the content appear as generated by users. Staged and/or non-staged photos are therefore not included in the analysis of the images.
Covert captions do not tell the consumers to go buy their products, but instead, serve as entertainment and/or brand interactivity. Covert captions can include:
• Questions to the readers, such as “what do you think?”, “how do you like this?” or personal questions such as “how’s your day?” etc.
• Quotes • Neutral texts
o Texts that do not mention sales, promotions, price or premium quality. Neutral texts could include short narratives, phrases, emoticons etc.
An image or caption is not considered as covert when it mentions or shows any information about product quality, pricings, discounts, advantages of the brand etc.
Overtness can be detected by looking at the use of obvious sales strategies. Such images include:
• An image on prices or sales.
An image is not necessarily overt when it shows a close up of a product, without any texts on advantages of the brand or products. A sole image of a product is not necessarily telling consumers to go buy their products, instead, it merely showcases the brands’ clothing.
Overt captions could include: • (Temporary) advantages
o Discounts o Promotions o Free products
• Captions that include short phrases such as “buy now” or “get yours” • Captions that direct consumers to go to the website and/or store
o e.g. Link in bio, www.(nameofstore).com, come visit our store etc. • Captions that highlight to be quick before it is sold out
o e.g. Soon in store/ limited editions
• Captions that offer (new) product information, promoting advantages of their clothes o e.g. Premium quality, origin of the clothes etc.
A post is always considered as overt when it includes one or more of the items above. When the caption includes both overt and covert elements it is also considered as overt since there is an element that clearly communicates persuasive aims.
The presence of faces is also noted when there is at least one face in the image. A face is always noted as present when it is clearly visible, whether it is a close-up or a full-body picture. It is not noted as present when someone’s back is photographed or in any other pictures that do not show a face5.
Table 2 provides insight on the brands that scored highest with regards to their brand engagement index on Instagram in the Netherlands in the fashion industry, and it makes clear that both established and unestablished brands can be found in this top 10.
Table 2.
Ten fashion brands that scored the highest engagement rates in the Netherlands in 2017 6
Brand Ranking Followers
G-Star RAW 10 440.134
Loavies* 15 336.092
C&A 17 307.289
Scotch & Soda 20 287.456
Daily Paper* 51 109.000 CoolCat 113 49.153 Hutspot* 118 46.420 Je M’appelle* 203 21.064 WE 230 16.433 MSTHV* 515 1.218
Note: The brands with a star (*) are unestablished fashion brands. The ranking shows the
place on the list of the most engaged brands on Instagram in the Netherlands.
Established brands may therefore usually have a larger audience, which results in significantly more likes and comments. Engagement is however calculated by dividing the total number of likes and comments by the number of followers, multiplied by a hundred. The number of likes is a strong social signal which quantifies the users who liked the post. The number of comments reflects the extent to which users took explicit action to leave remarks on a post. The number of followers indicates the size of the profiles’ audience (Bakhshi, Gilbert & Shamma, 2014). The engagement formula leads to a certain percentage per post, whereby a higher percentage indicates a higher engagement rate. Besides the computation of the engagement ratio of the posts, each profile is also noted on the number of followers alone.
6 Source: "Brand Engagement Index: Dé ranglijst van merken in Nederland op Instagram", 2018
Results
The data showed that out of the two-hundred posts that were analyzed, only two used an overt image. This means that the differences between overt images and covert images cannot be tested since there is not enough data to compare the two groups7. When investigating H1, H3a and H3b8, only covert captions vs. overt captions are therefore tested instead of also checking for the effects of overt vs. covert images. Furthermore, the data proved to be reliable as checked through the inter-coder reliability test using Krippendorff’s Alpha (see table 5).
An independent sample t-test was conducted to compare the levels of engagement9 on Instagram in covert and overt captions. Covert captions scored an average of 3.32%
engagement (SD = 5.21%) while overt captions core an average of 2.17% engagement (SD = 4.93%). This difference is however not significant, t (198) = 1.56, p=.119, CI = [-0.30, 2.60],
7 An attempt was made to randomly select additional overt images in the same way in which additional images of faces would have been collected in case there was not enough data (see methodology). It however proved to be impossible to analyze an extra three overt images per profile since such images were rarely used by the given brands.
8 H1: Posts on Instagram generate higher engagement when employing covert advertising than when employing overt advertising.
H3a: Covert advertising generates higher engagement for unestablished brands than for established brands.
H3b: Overt advertising generates higher engagement for established brands than for unestablished brands.
9 Engagement is computed by dividing the total number of likes and comments by the number of followers, multiplied by a hundred.
thereby rejecting H1 (see table 4 of the appendix). These results suggest that the level of engagement on a post does not differ depending on the post employing an overt or a covert caption.
An independent sample t-test was also conducted to compare the levels of
engagement on Instagram posts that did and did not include a face. Images with a face scored an average of 3.17% engagement (SD = 5.76%) while images without a face scored an average of 2.27% engagement (SD = 3.61%). This difference is however not significant, t(193.27) = 1.36, p = .175, CI = [ -0.41 , 2.22], thereby rejecting H2 (see table 6 of the appendix). These results suggest that the level of engagement on a post does not differ depending on the presence or absence of a human face in the image.
A two-way ANOVA was also conducted that examined the effect of a face vs. no face and established vs. unestablished brands on the level of engagement of the Instagram posts. The test revealed that there was no significant interaction between the presence or absence of a face and established vs. unestablished brands on the level of engagement, F = (1, 196) = 2.45, p = .119. This means that there is no difference between established and unestablished brands’ generated engagement on Instagram when using a face in their posts, thereby
rejecting H2a and H2b (See table 7 of the appendix).
A two-way ANOVA was conducted that examined the effect of covert vs. overt captions and established vs. unestablished brands on the level of engagement of the
Instagram posts. The test revealed that there was no significant interaction between covert vs. overt captions in advertising and established vs. unestablished brands on the level of
engagement, F= (1, 196) = .38, p = .536. As previously demonstrated, the main effect of covert vs. overt captions proved to be insignificant, F = (1, 196) = .40, p = .527. The effect of established vs. unestablished brands however did prove to be significant, F (1, 196) = 48.24,
brands’ generated engagement on Instagram when using covert or overt advertisements, thereby rejecting H3a and H3b (See table 8 of the appendix), while the data did prove that there is a difference in the engagement rate when a post is shared by an established or and unestablished brand.
A one-way analysis of variance (ANOVA) was therefore conducted in order to compare established vs. unestablished brands with regards to engagement, followers, likes and comments. There proved to be a significant difference between the effect of established vs. unestablished brands on engagement, F = (1, 198) = 51.97, p < 0.001. Unestablished brands were found to have a significantly higher engagement (M = 5.18%, SD = 6.45%) than established brands (M = 0.53%, SD = 0.30%) (see table 9 of the appendix). Established brands were in turn found to have significantly more followers (M = 9,474,920, SD = 9,505,538) (See table 10 of the appendix), likes (M = 36,993, SD = 45,244) (see table 11 of the appendix) and comments (M = 134.43, SD = 205.05) (see table 12 of the appendix), than unestablished brands (M = 22,725, SD = 28,185), (M = 380, SD = 627), (M = 5.95, SD = 9.06).
Discussion
This research did not confirm any differences between the effectiveness of overt or covert captions, nor did it confirm any differences between the effectiveness of overt vs. covert caption strategies when employed by an established or an unestablished brand. The research also did not confirm any effects of the presence of a face in Instagram posts, nor did it confirm any significant differences in engagement when an established or an unestablished brand shares images that include a human face.
The data did prove that there is a significant difference between engagement on a post when it is shared by an established or an unestablished brand. In fact, unestablished brands generated a significantly higher engagement on their posts (5.18%) compared to established
brands (0.53%), meaning that a higher proportion of the followers of an unestablished brand interact with the brands posts, compared to established brands.
Unestablished brands in this research tended to have significantly less followers than established brands, while having significantly less followers may in turn make the
unestablished Instagram accounts seem more personal, thereby stimulating followers to leave a comment or like the posts.
This is also in line with the theory on top-dogs vs. under-dogs, whereby the established brands represent the top-dogs and the unestablished brands represent the under-dogs. The less privileged, under-dog type of brand is associated with feelings of friendliness and the more powerful, top-dog type of brand is associated with being dominant (Prell, 2011; Jun et al., 2015). It may in turn be less appealing to interact with the dominant, leading brand that has millions of followers, and more appealing to interact with the smaller, ‘friendly’ brand that has significantly less followers.
It should also be noted that ‘effectiveness’ in this research was calculated by the
engagement rate, while it could be that effectiveness, in the world of advertisers, is calculated by looking at the number of likes and comments alone. In the latter case, established brands would be seen as being more effective due to their higher reach.
The use of a human face in an Instagram posts was expected to increase the generated engagement (Bakhshi, Gilbert & Shamma, 2014) as a face is said to increase identification and draw consumers to a post (Walter, 2011). This research however did not confirm such relationships and/or any differences between an established or an unestablished brand sharing images that included a face. It could be that some of the analyzed posts included an image of an influencer. Influencers were however not taken into account in this research, while it could be that such influential individuals had an effect on the results.
Except for two of the analysed posts, all images were covert. This means that most brands tend to share images of a quote, a landscape, an animal, a product, a behind-the-scenes or a person, instead of sharing an image with texts that inform consumers about prices and/or sales. The two posts that did employ overt images highlighted sales, while overt images that display (low) prices were never used. It therefore seems like brands, in this research,
acknowledge that users of Instagram may be more appealed by content marketing than by overt images on sales or prices. Additionally, brand managers on Instagram may have taken into account that consumers do not like to be confronted with clear advertisements on a platform that they use to pass time (Voorveld et al., 2018).
The most frequently used tactic of overt captions included texts that directed consumers to go to the website and/or store (e.g. ‘find your own polka-dot perfection online and in stores’). This happened significantly more often (66 times) than any other types of overt captions. Since there proved to be no effect of established or unestablished brands adopting such overt captions on engagement rate, it could be that consumers do not mind being directed to a website or store. Perhaps this type of advertising does not bother consumers as much as texts that tell consumers to ‘buy now’. Future research should therefore gather more data that explores the individual differences of overt tactics (e.g. Tactic 1 – low price
advantages, tactic 2 – ‘buy now’ tactic 3 ‘go to the site’) in order to be able to test their differences.
The results do not align with the theory on the persuasion knowledge model, which holds that consumers try to resist being influenced by advertisers, making them skeptical about advertisements (Mohr et al., 2001). In this research, overt posts did not suffer from getting less engagement, even when a post uses overt captions that tell users to visit the website and buy their clothes. Consumers therefore did not react defensively when confronted with clear
attempts to persuade (Mohr et al., 2001), thereby rejecting the theory of the persuasion knowledge model on Instagram captions.
An explanation for the unexpected insignificant effect of covert vs. overt captions on engagement could relate to the aspects of low-involvement processing. As previously mentioned, Instagram allows users to scroll through feeds which are filled with images that are constantly renewed. The constant renewal of images may lead to a content overload as each minute over 3000 posts get added to the platform (Instagram, 2018), leading to a
decreased attention span (Steven, 2016). When engaged in low involvement processing, little working memory is used (Petty & Cacioppo, 1986). This means that consumers are not engaged in critical thinking and therefore not as likely to activate their persuasion knowledge. When engaged in low involvement processing, consumers are also more drawn by contextual factors than by the actual message content and base their judgements on simple cues (Yanh & Lin, 2014), such as nice looking images.
Previous research also showed that most brand communication is processed with little attention using low involvement processing. This low involvement processing is in turn poor at drawing conclusions based on advertisements. It does however store visual impressions as simple associations with the brand. This way, the more often a brand shares nice looking, covert images, the stronger the mind starts to link such images with the brand. These repeated associations therefore define brands within consumers’ minds (Heath, 2001). Hence, brands may be purposely and repeatedly sharing covert images on Instagram in an attempt to create favourable associations within the minds of their followers. Such associations can have a powerful influence on intuitive decision making (Heath, 2001). Consequently, consumers often find it hard to explain their reasons for choosing for particular brands and tend to deny that advertising had a significant impact on their behaviour (Heath, 2001). Such associations and their impact on behaviour would be more obvious if consumers were repeatedly
confronted with overt looking images, which in turn make it clear that there is an attempt to persuade. Covert images may however make consumers feel like they are in charge of choosing the brands they like, which may be the reason why brands do not often use overt images on Instagram.
Additionally, it is possible that the features of Instagram such as engagement, relaxation or inspiration (Voorveld et al., 2017) guide consumers’ interpretation of the commercial posts. The positive experience of Instagram can become associated with the advertising. This associative transfer may supersede the increased skepticism and negative attitudes towards advertisements. Accordingly, even if one recognizes a clear advertisement within the Instagram environment, the superiority of the positive experience over the activation of advertising persuasion knowledge marginalizes the assumed occurrence of skepticism and negative attitudes. It could therefore be that the platform of Instagram is not affected by negative attitudes associated with advertisements, thereby making Instagram an exception to consumers’ skepticism with regards to advertisements.
For managers of a brands’ Instagram page, the findings shed light on how to balance overt or covert captions with covert images. Overt captions did not prove to trigger less engagement than covert captions, while covert images tend to be favourable to both brand managers and consumers. Supposedly this is because covert images are visually more attractive and gain more attention than commercial messages by offering appealing visual presentations (Tomažic et al., 2014). Future research should try to gather more data of overt Instagram posts to verify this contention. Consumers may also be more appreciative of content marketing than clear advertisements since this is a better match with the platform of Instagram (Voorveld et al., 2018). Brand managers may in turn benefit from appreciative consumers.
A major theoretical implication draws on the finding that the persuasion knowledge model does not work for covert captions on Instagram, meaning that consumers are not put off by texts that tell them to go to the site and/or buy their clothes. Hence, the characteristics of the platform of Instagram may lead to a different experience. Engagement with Instagram is likely to spill over into the way in which consumers engage with advertisements within the platform, which in turn affects advertisement evaluations (Voorveld et al., 2018). Brand managers should therefore take into account the way in which Instagram is used and treat it differently to other types of social media platforms. Engagement is context specific and consumers engage with Instagram differently based on its unique characteristics.
References
Bakhshi, S., Gilbert, E. & Shamma, D. (2014). “Face It: Instagram Pictures With Faces are More Popular”. Georgia Institute of technology. Retrieved April 16th 2018, from http://www.news.gatech.edu/2014/03/20/face-it-instagram-pictures-faces-are-more-popular
Beverland, M. (2005) "Brand management and the challenge of authenticity". Journal of
Product & Brand Management, Volume, 14 Issue: 7, p. 460-461, DOI:
10.1108/10610420510633413
Brand Engagement Index: Dé ranglijst van merken in Nederland op Instagram. (2018). Retrieved March 18th from https://brandengagementindex.nl/merk
Brown, C. & Krishna, A. (2004), “The Skeptical Shopper: a Metacognitive Account for the Effects of Consumer Default Options on Choice”. Journal of Consumer Research, Volume 31, Issue: 3, p. 529-539, DOI: 10.1086/425087
Chalmers, T. & Price, L. (2009). “Perceptions of Authenticity in Advertisements:
Negotiating the Inauthentic.” Advances in Consumer Research, Volume 36, p. 72– 75. Retrieved from http://acrwebsite.org/volumes/14456/volumes/v36/NA-36 DeCarlo, T. (2005). Effects of Sales Message and Suspicion of Ulterior Motives on
Salesperson Evaluation. Journal of Consumer Psychology, Volume 15, p. 238–249, DOI: 10.1207/s15327663jcp1503_9
Du Plessis, C. (2017). The Role of Content Marketing in Social Media Content Communities
South African Journal of Information Management, Volume 19, Issue: 1, DOI:
10.4102/sajim.v19i1.866
Eriksson, E. & Hansson, C. (2016). Fashion Brands and Engagement on Instagram: How can Fashion Brands Apply Content Marketing on Instagram to engage Millennials and Increase Brand Equity? Uppsala University Library, Halmstad University
Göbel, F., Meyer, A., Ramaseshan, B. & Bartsch, S. (2017). Consumer Responses to Covert Advertising in Social Media. Marketing Intelligence & Planning, Volume 35, Issue: 5, p. 578-593, DOI: 10.1108/MIP-11-2016-0212
Heath, R. (2001) Low Involvement Processing - a New Model of Brand
Communication. Journal of Marketing Communications, Volume 7, Issue: 1, p. 27-33, DOI: 10.1080/13527260123019
Hellberg, M. (2015). Visual Brand Communication on Instagram: a Study on Consumer Engagement. Department of Marketing. Hanken School of Economics
Holt, D. (2002). “Why Do Brands Cause Trouble? A Dialectical Theory of Consumer Culture and Branding.” Journal of Consumer Research, Volume 29, Issue: 1, p. 70– 90
Instagram (2015). Instagram statistics. Retrieved April 16th 2018 from https://instagram-press.com/our-story/
Jun, S., Sung, J., Gentry J. & McGinnis L. (2015) Effects of Under-dog vs. Top-dog
Positioning Advertising. International Journal of Advertising, Volume 34, Issue: 3, p. 495-514, DOI: 10.1080/02650487.2014.996199
Krippendorff, K. (2004). Content Analysis: An Introduction to Its Methodology. Second Edition. Thousand Oaks, CA: Sage. DOI: 10.1177/1094428108324513
Lange, F. & Dahlèn, M. (2003). Let’s be Strange: Brand Familiarity and ad-brand
Incongruency. Journal of Product and Brand Management, Volume 12, Issue: 7, p. 449– 461, DOI: 10.1108/10610420310506010
Latiff, Z., & Safiee, N. (2015). New Business Set Up for Branding Strategies on Social media – Instagram. Procedia Computer Science. Volume 72, p. 13–23
Lei Shi (2014). On Branding in the Western World with Figurative Content Factors: A Case of Established Versus Non-Established Brands. Journal of Promotion Management, Volume 20, Issue: 3, p. 358-371, DOI: 10.1080/10496491.2014.908804
Machleit, K., Allen, C. & Madden, J. (1993). The Mature Brand and Brand Interest: An Alternative Consequence of Ad-Evoked Affect. Journal of Marketing, Volume 57, Issue: 4, p. 72– 82, DOI: 10.2307/1252220
Main, K., Dahl, D, & Darke, P. (2007). Deliberative and Automatic Bases of Suspicion: Empirical Evidence of the Sinister Attribution error. Journal of Consumer
Psychology, Volume 17, p. 59–69, DOI: 10.1207/s15327663jcp1701_9
Miller, F. (2015) Ad Authenticity: An Alternative Explanation of Advertising's Effect on Established Brand Attitudes. Journal of Current Issues & Research in Advertising, Volume 36, Issue: 2, p. 177-194, DOI: 10.1080/10641734.2015.1023871
Miller, F. & Allen, C. (2012) How Does Celebrity Meaning Transfer? Investigating the Process of Meaning Transfer with Celebrity Affiliates and Mature Brands. Journal of
Consumer Psychology, Volume 22, p. 443, 452, DOI: 10.1016/j.jcps.2011.11.001
Mohr, L., Webb, D. & Harris, K. (2001) “Do Consumers Expect Companies to be Socially Responsible? The impact of Corporate Social Responsibility on Buying Behavior”.
Journal of Consumer Affairs, Volume 35, Issue: 1, p. 45-73, Retrieved from
https://doi-org.proxy.uba.uva.nl:2443/10.1111/j.1745-6606.2001.tb00102.x Moyer-Gusé, E. (2008). Toward a Theory of Entertainment Persuasion: Explaining the
Persuasive Effects of Entertainment-Education Messages. Communication Theory, Volume 18, Issue 3, p. 407-425. DOI: 10.1111/j.1468-2885.2008.00328
Muthukumar, M. (2013). Covert Advertising–What does Covert Advertising do? Advances
https://search.proquest.com/openview/09e79550732d2eed42f371b7fdfb535b/1?pq-origsite=gscholar&cbl=2030322
Petty, R. & Cacioppo, J. (1986), “The Elaboration Likelihood Model of Persuasion”,
Advances in experimental social psychology, Volume 19, p. 123-205, DOI:
10.1016/S0065-2601(08)60214-2
Sjodin, H. & Törn, F. (2006). When Communication Challenges Brand Associations: A Framework for Understanding Consumer Responses to Brand Image Incongruity.
Journal of Consumer Behavior, Volume 5, Issue: 1, p. 32-42, DOI 10.1002/cb.44
Steven, J. (2016). Decreasing Attention Spans and Your Website Social Media Strategy. Retrieved May 31st from https://www.adweek.com/digital/john-stevens-guest-post-decreasing-attention-spans/
Tomažic, T., Boras, D., Jurišic, J. & Lesjak, D. (2014) "Covert Advertising as Inadmissible Presentation of Information". Industrial Management & Data Systems, Volume 114, Issue: 1, p.107-122, DOI: 10.1108/IMDS-04-2013-0204
Van den Putte, B. (2002). An integrative framework for effective communication. In W. Nelissen (Ed.), Marketing for sustainability: towards transactional policy-making p. 83-95. Amsterdam: IOS Press, retrieved from http://hdl.handle.net/11245/1.210241 Vinerean, S. & Opreana, A. (2015) Consumer Engagement in Online Settings:
Conceptualization and Validation of Measurement Scales. Expert Journal of
Marketing. Volume 3, Issue: 2, p. 35-50, retrieved from
http://marketing.expertjournals.com/ark:/16759/EJM_307vinerean35-50.pdf Voorveld, H., van Noort, G., Muntinga, D. & Bronner, F. (2018): Engagement with Social
Media and Social Media Advertising: The Differentiating Role of Platform Type.
Journal of Advertising, DOI: 10.1080/00913367.2017.1405754
http://ebook.eqbal.ac.ir/Web-Design/Designing%20for%20Emotion.pdf Wojdynski, B. & Evans, N. (2015): Going Native: Effects of Disclosure Position and
Language on the Recognition and Evaluation of Online Native Advertising. Journal
Appendix
Figure 1. Theoretical model and hypotheses tested
Table 1.
The analysed brands of this research
Unestablished brands Established brands Ovelia Transtoto (2016) – UK Minimum (1997) DE
Kitri (2017) UK H&M (1927) SW
Les favorites (2017) – NL ZARA (1974) SP
Bellerose (1989) BE Mango (1984) SP
Noumenon (2016) – NL Forever21 (1984) US
Premme (2017) – US Calvin Klein (1942) US
Elliette (2016) US C&A (1841) NL
Hola Couture (2017) BE Maison Scotch (1885) NL
Elvine (2016) SW G-Star RAW (1986) NL
Fabienne Chapot (2016) – NL Tommy Hilfiger (1951) – US
Table 2.
Ten fashion brands that scored the highest engagement rate in the Netherlands
Brand Ranking Followers
G-Star RAW 10 440.134
Loavies* 15 336.092
C&A 17 307.289
Scotch & Soda 20 287.456
Daily Paper* 51 109.000 CoolCat 113 49.153 Hutspot* 118 46.420 Je M’appelle* 203 21.064 WE 230 16.433 MSTHV* 515 1.218 Table 3.
The list of brands before random selection
Unestablished brands Established brands
Atlein (2016) – FR Abercrombie & Fitch (1892) – US Claud agency (2017) – NL America Today (1989) – NL Elliette (2016) – US Bellerose (1989) – BE
Elvine (2016) – SE Calvin Klein (1942) – US
Everybody.World (2016) – UK C&A (1841) – NL Fabienne Chapot (2016) – NL Desigual (1984) – SP Hola Couture (2017) – BE Diesel (1978) – IT
Hornoff (2016) – DE Forever21 (1984) – US
I.AM.GIA (2017) – IT G-Star RAW (1986) – NL
Kitri (2017) – UK H&M (1927) – SW
Ovelia Transtoto (2016) – UK Maison Scotch (1885) – NL
Kitri (2017) UK Mango (1984) – SP
La Doyenne Vintage (2017) – NL Marks & Spencer (1998) – UK Les favorites (2017) – NL Michael Kors (1983) – US
Mafcsm (2017) – UK Minimum (1997) – DE
Neous (2017) – UK Samsoe & Samsoe (1983) – DK Noumenon (2016) – NL Sissy Boy (1982) – NL
Oak + Oar (2016) – CA Tommy Hilfiger (1951) – US Ovelia Transtoto (2016) – UK WE (1962) – NL
Figure 2. Coding rules
0 = Absent/ No 1 = Present/ Yes 1. Covertness
Does the covert image include: • 1.1 An image of a quote • 1.2. An image of a joke • 1.3 An image of a landscape • 1.4 An image of an animal • 1.5 An image of a person • 1.6 An image of food • 1.7 Clothes • 1.8 Other • 2. Exclusive content
o 2.1 A behind-the scenes photo o 2.2 A day-in-the-life-of Does the covert caption include: • 3.1 Questions to the readers
o Such as “what do you think?”, “how do you like this?” or personal questions such as “how’s your day?” etc.
• 3.2 Quotes
• 3.3 Short narratives • 3.4 Neutral text
o Text that does not mention sales, promotions, price or premium quality • 3.5 Other
2. Overtness
Does the overt image include: • 4.1. An image on prices
o Costs
• 4.2 An image on sales o Discounts Does the overt caption include:
• 5.1 Temporary advantages
o e.g. Discounts, promotions, free products • 5.2 Phrases such as “buy now” or “get yours” etc.
• 5.3 Texts that direct consumers to the website and/or store o e.g. Link in bio, come visit our store etc.
• 5.4 Texts that highlight to be quick before it is sold out o e.g. Soon in store or limited editions
• 5.5 Texts that offer (new) product information, promoting advantages of their clothes o e.g. premium quality, origin of the clothes etc.
Table 4. Results of the t-test, testing H1
Group Statistics – Engagement on Posts
N Mean Standard
deviation
Standard Error Mean
Covert caption 119 3.32% 5.21% 0.48%
Overt caption 81 2.17% 4.93% 0.55%
Independent Samples Test – Engagement on Posts Levene’s
test for equality
of variances
t-test for equality of Means 95% confidence interval of the difference F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference Lower Upper Equal variances assumed 1.38 .242 1.56 198 .119 1.15% 0.73% -0.30% 2.60% Equal variances not assumed 1.58 178 .115 1.15% 0.72% -0.28% 2.59%
Table 5. Inter-coder reliability test
Krippendorff’s alpha Percentage agreement
Presence/ absence of a face 1.0000 100%
Presence of overt captions .9588 98.5%
Presence of overt images 1.0000 100%
Table 6. Results of the t-test, testing H2
Group Statistics – Engagement on Posts
N Mean Standard
deviation
Standard Error Mean
Face 130 3.17% 5.76% 0.50%
No Face 70 2.67% 3.62% 0.43%
Independent Samples Test – Engagement on Posts Levene’s
test for equality
of variances
t-test for equality of Means 95% confidence interval of the difference F Sig. t df Sig. (2-tailed) Mean Difference Std. Error Difference Lower Upper Equal variances assumed 6.43 .012 1.19 198 .234 0.91% 0.76% -.60% 2.40% Equal variances not assumed 1.36 193.27 .175 0.91% 0.66% -.41% 2.21%
Table 7. Results of the two-way analysis of variance, testing for H2a and H2b Test of Between-Subjects Effects
df Mean Square F Sig.
Face 1 53.32 2.60 .109
Established * Face 1 50.31 2.45 .119
Error 196 20.526
Table 8. Results of the two-way analysis of variance, testing H3a and H3b Test of Between-Subjects Effects
df Mean Square F Sig.
Covert vs. Overt Captions 1 8.41 .40 .527
Established vs. Unestablished 1 1,012.80 48.24 .000 Covert vs. Overt Captions *
Established vs. Unestablished
1 8.05 .38 .536
Error 196 20.995
Note. Dependent variable: engagement on posts.
Table 9. Results of the one-way analysis of variance (ANOVA) – Engagement ANOVA – Engagement on Posts
df Mean Square F Sig.
Between Groups 1 1,083.96 51.97 .000
Within Groups 198 20.857
Total 199
Descriptives – Engagement on Posts
95% Confidence Interval for Mean N Mean Std. Deviation Lower Bound Upper Bound Unestablished brands 100 5.18% 6.45% 3.90% 6.46% Established brands 100 0.53% 0.30% 0.47% 0.59% Total 200 2.86% 5.11% 2.14% 3.57%
Table 10. Results of the one-way analysis of variance (ANOVA) - Followers ANOVA – Number of Followers
df Mean Square F Sig.
Between Groups 1 4,466,348,670,000,000.00 98.861 .000 Within Groups 198 45,178,032,630,000.00
Total 199
Descriptives – Number of Followers
95% Confidence Interval for Mean
N Mean Std. Deviation Lower Bound
Upper Bound Unestablished brands 100 22,725.20 28,185.48 17,132.59 28,317.81 Established brands 100 9,474,020.00 9,505,538.96 7,587,914.85 11,360,125.15 Total 200 4,748,372.60 8,209,443.983 3,603,660.91 5,893,084.21
Table 11. Results of the one-way analysis of variance (ANOVA) – Likes ANOVA – Number of Likes
df Mean Square F Sig.
Between Groups 1 67,026,137,650.00 65.44 .000
Within Groups 198 1,024,244,482.00
Total 199
Descriptives – Number of Likes
95% Confidence Interval for Mean
N Mean Std. Deviation Lower Bound Upper Bound
Unestablished brands 100 380.21 627.321 255.74 504.68
Established brands 100 36,993.36 45,255.888 28,013.61 45,973.11
Table 12. Results of the one-way analysis of variance (ANOVA) – Comments ANOVA – Number of Comments
df Mean Square F Sig.
Between Groups 1 825,355.52 21.67 .000
Within Groups 198 38,090.61
Total 199
Descriptives – Number of Comments
95% Confidence Interval for Mean
N Mean Std.
Deviation
Lower Bound Upper Bound
Unestablished brands 100 5.95 9.06 4.15 7.75
Established brands 100 134.43 275.86 79.69 189.17
