Simplicity is not key: Understanding firm-generated social media images and consumer liking

The many faces of social media in business and economics research: Taking stock of the literature and looking into the future

“…The extant results across many different empirical settings have generally established that visual content (e.g., photos; Li & Xie, 2020), ease of reading (Pancer et al., 2019), emotionality (Tellis et al., 2019), humor (Lee et al., 2018), fit between message and users (Zhang et al., 2017), messages from individuals that share followers and followees (Peng et al., 2018), social influence through contagion (Susarla et al., 2012), and higher levels of firm‐induced social media activity (e.g., more posts; Dhaoui & Webster, 2021) give rise to more engagement in the form of likes and shares. Interestingly, recent research shows that too rich visual design elements (e.g., combining animations and pictographs) may actually harm engagement (Bashirzadeh et al., 2022) and that there is a u‐shaped relationship between visual complexity and consumer liking (Overgoor et al., 2022). There also seems to exist temporal variation in engagement in response to different contents, such that, for example, the engagement with virtue content (i.e., offering long‐term knowledge benefits) is stronger in the morning, whereas the engagement with vice content (i.e., offering immediate gratification) is stronger in the evening (Zor et al., 2022).…”

Differential effects of visual complexity in firm-generated content on consumer engagements: a deep learning approach

“…For example, Shin et al (2020) have found that while pixel complexity increases consumer engagement in terms of the number of likes and shares, object complexity decreases both forms of engagement. Overgoor et al (2022) reported that pixel complexity has an inverted U-shaped relationship with the number of likes. Thus, building on previous research, we investigated the non-linear effects of these two types of visual complexity on likes and shares.…”

“…Object-level complexity represents the complexity of semantic information in scenes and quantifies the diversity of object-related information in an image (Sample et al, 2020). An image with low object-level complexity has fewer central cues than an image that has high object-level complexity, looks clean and easy to process, and is therefore likeable (Overgoor et al, 2022). However, as the level of complexity increases, the image encapsulates a greater quantity of objects and is primarily composed of a set of object categories instead of just a few.…”

“…When object-level complexity is high, it slows down content readers and improves their attitude toward the image, through its collative properties (Pieters et al, 2010). These design intricacies make the image more appealing to readers and generally more likeable (Overgoor et al, 2022). Thus, we propose the following hypotheses:…”

“…Furthermore, we controlled for an image-level variable, namely color saturation, as judicious color saturation can offer advantageous peripheral cues, thereby capturing users' MIP 42,4 attention (Overgoor et al, 2022). After accessing an image's resolution, we extracted the RGB value of a pixel, specifically (200,100,50), and then normalized it to (0.78, 0.39, 0.2).…”