Comparative analysis of advertising attention to Facebook social network: Evidence from eye-movement data

Highlights

• The main is to navigate online advertising during social networking sites.

• A behavioral eye tracking exploratory study is conducted on Facebook advertising.

• Eye tracking metrics are correlated to Facebook advertising.

Abstract

With the rapid development of social networking sites, Internet advertising has quickly become an important marketing tool. Previous studies either utilized the concept of click-through-rate (CTR) to estimate the number of people who responded to the presence of the advertisements, or employed a questionnaire to measure the advertisements' effectiveness. However, these results couldn't accurately reveal potential customers due to the lack of a common understanding on Internet advertising. This study used eye tracking technology to record the participants' eye gaze behaviors and analyze the differences in the distribution of visual attentions while browsing the researcher's Facebook page. The participants consisted of 65 students whose eye movements were tracked and recorded by the eye tracking system. These participants came from a northern university of science and technology in Taiwan, of which 35 were assigned to Facebook friend group and the other 30 were assigned to non-Facebook friend group according to their interpersonal relationships with the researcher. Eye tracking measurements, including total fixation duration (TFD) and number of fixations (NOF) on the defined regions of interest (ROIs) of the Facebook page were compared to indicate their visual attentions. The experimental results showed that 1) there was a significant difference on the right-hand side (RHS) ads between the two groups based on TFD and NOF; 2) there was a significant difference on the desktop news feed (DNF) ads between the two groups based on TFD and NOF; 3) there was the same viewing sequences between the two groups based on stimulus with RHS ads; and 4) there was a difference of viewing sequences between the two groups based on stimulus with DNF ads.

Introduction

Social networks has enjoyed a surge in popularity worldwide in recent years. Among the social networking platforms, Facebook has risen at the fastest pace. According to data compiled by Astralweb in 2015, 56.5% of American citizens use Facebook and an average of 450 million Asians use Facebook each month. In addition, statistics from the market survey firm comScore showed that the time Americans spent on Facebook exceeded Google for the first time (Anderson, 2010). From this data, it is evident that most Internet users today spend their time on Facebook, sharing recent developments, uploading photos or playing games. Clearly, Facebook has become a core part of people's daily routine even replacing information search as the most common purpose of Internet use. Thus, lots of renowned companies quickly realized the value of brand promotion by using Facebook, such as Coca-Cola, Samsung Mobile, Starbucks, Nike Football, Pringles, and so on. From marketing campaigns by leveraging Facebook's features, it creates a platform which satisfies marketers' needs to build interaction with customers and opens a transparent circumstance where information can flow freely between consumers and marketers. (Dunay & Krueger, 2010). Besides that, Vahl (2011) highlighted the businesses magnetism towards Facebook and mentioned that Facebook revenue from businesses advertisements reached approximately US$2.5 billion in 2011.

This flourishing Internet-based commercial activity is different from previous traditional consumer markets in that it does not possess human and interactive mechanisms. This has made Internet media an extremely effective marketing and advertising channel (Wang, Wang, & Farn, 2009). However, there is still only a limited and fragmentary understanding on how to utilize Internet advertising effectively. Zhang and Kim (2008) believed that there are four main factors behind Internet advertising usage: viewers, Internet advertising, advertisers and the advertised products or services.

Owing to desktop computer features, Facebook advertising placement options include right-hand side (RHS) ads and desktop news feed (DNF) ads. RHS ads appear on the bar located on the right side of Facebook, sandwiched between news feed and ticker. Benefits of placing ads on the RHS are because RHS ads are smaller and appear off to the side rather than front and center on the screen, RHS ads are less invasive. If advertisers choose to place their ads on Facebook's desktop news feed, ads will appear on news feed, the section of Facebook where posts by friends and family appear. It's important to remember, however, that placing ads on DNF means the ads will only appear on news feeds accessed through desktop devices or laptops. Benefits of placing ads on DNF are that consumers favor DNF ads because their resemblance to regular posts makes them more appealing to Facebook users. According to consumer surveys, people hate intrusive ads, such as pop-up ads and video ads that play automatically with sound on. On the contrary, DNF ads blend with its environment and do not interrupt the user experience. Although Facebook offers the two ads placement on desktop computers, it is necessary to have a more realistic understanding of how, how often, and what consumers have read on Facebook ads while browsing Facebook page. Therefore, how to effectively attract viewers to the Internet advertising placed on social networking sites is one of the issues that advertisers are concerned about.

In the past, measurement of Internet advertising was based on Cost Per Impression (CPM) (Novak & Hoffman, 2000) but relatively quickly click-through rate (CTR) became the standard measure. The purpose of click-through rates was to measure the ratio of clicks to impressions of an Internet advertising (Dréze & Hussherr, 2003). Unfortunately, CTRs were plummeting in the real world because of four critical questions as follows: (1) why do banner ads seem to be ineffective; (2) what can advertisers do to improve their effectiveness; (3) does an immediate measure such as click-through rate undervalue online advertising; and (4) are memory-based measures such as recall or awareness more appropriate? (Briggs & Hollis, 1997). Therefore, no consensus can be reached about effective advertising transmission from this information. In other words, by considering the basic question of whether viewers are attracted to advertising on social network platforms, the measurement standard should be how much attention an advertisement receives. Although attention effectiveness has caught enormous concern in marketing (Hernández-Méndez & Muñoz-Leiva, 2015), it is difficult to understand an impact on consumers' decisions and emotion using CTR measurement. In recent years, researchers have been using eye tracking technology to examine the cognitive processing in reading, visual perception, and information processing (Rayner, 1998) and paid attention to advertising to make up for deficiencies in past research (Chung et al., 2013, Radach et al., 2003), even predicted the traditional advertising effectiveness for video advertising (Zhang & Yuan, 2018). In addition, Wedel and Pieters (2008) regarded the use of eye tracking as to obtain the viewers' visual attention which information enters to reach higher-order cognitive processes of increased interest. In order to address the above-mentioned questions about CRTs, eye tracking technology can be used to understand visual attention distribution by analyzing the viewers’ eye movement data and how attention changes from the eye movement paths. Therefore, to meticulously probe the level of attention given to Internet advertising makes it an extremely persuasive form of measurement.

Recent studies have used eye trackers to examine visual behavior of consumers, such as visual behavior on advertisements (Ju and Johnson, 2010, Pieters et al., 1999, Wedel and Pieters, 2000, Wedel and Pieters, 2008). By using eye trackers, digital advancement has improved the precision and usability. Throughout the history of eye tracking technology, researchers often define regions of interest (ROIs) to reveal how attention is allocated according to research questions and main areas of concern before the experiment to observe the relationship between eye movement variables and the main areas of interest of the experiment (Ho, 2014). Depending on the defined ROIs, several key variables have emerged as significant indicators of visual behaviors, including fixations, saccades, pupil dilation, and scan path (Rayner, 1998). Based on these key variables, this study adopted three kinds of dependent variables for processing eye movement data used in visual behavior observations, which are the total fixation durations (TFD), the number of fixations (NOF), and the latency of first fixation (LFF) on each ROI used to assess participants’ attention distribution. These data were related to endogenous and exogenous attention controls. The longer the TFD, the more attractive the ROI is to the participant; the larger the NOF, the more attractive the ROI is to the participant; the shorter the LFF paid on an ROI, the more attractive it is to the participant (Calvo and Lang, 2004, Hewig et al., 2008, Ho, 2014).

The main purpose of this study was to determine whether participants with different interpersonal relationships within Facebook look at the RHS and DNF ads displayed; thus, ascertaining the differences in visual attentions of Facebook ads were generated as follows.

• To investigate how RHS and DNF Facebook ads were gazed by participants with different interpersonal relationships.

• To explore the differences in visual attentions of the ROIs between participants with different interpersonal relationships while browsing RHS and DNF Facebook ads, respectively.

Based on the previous literature review, this study intended to examine the differences in eye measurements of TFD and NOF of Facebook ads among participants with difference interpersonal relationships. Thus, the experimental analysis was conducted to observe the participants' visual patterns via eye tracking technology. To thoroughly examine how participants gazed at ROIs of RHS and DNF Facebook ads, we conducted this study to examine participants’ visual attention in terms of their eye-movement patterns as they were given a Facebook page. Therefore, this study proposed four research questions and hypotheses:

• Is there a significant difference in TFD of RHS ads among participants with different interpersonal relationships?

  • There is a significant difference in TFD of RHS ads among participants with different interpersonal relationships.

• Is there a significant difference in TFD of DNF ads among participants with different interpersonal relationships?

  • There is a significant difference in TFD of DNF ads among participants with different interpersonal relationships.

• Is there a significant difference in NOF of RHS ads among participants with different interpersonal relationships?

  • There is a significant difference in NOF of RHS ads among participants with different interpersonal relationships.

• Is there a significant difference in NOF of DNF ads among participants with different interpersonal relationships?

  • There is a significant difference in NOF of DNF ads among participants with different interpersonal relationships.

• Is there a significant difference of viewing sequences of Facebook ads among participants with different interpersonal relationships?

  • There is a significant difference viewing sequences of Facebook ads among participants with different interpersonal relationships.