Approximately 65 million social posts include branded content. Every. Single. Day. Jochen Hartmann, Mark Heitmann, Christina Schamp, and Oded Netzer set out to academically measure the power of brand selfies. In their research, the team studied a quarter-million images from nearly 200 brands across Instagram and Twitter using sophisticated automated image analysis and text mining. We all know about selfies as a cultural phenomenon, but do these images affect how we interact with brands?
The team evaluated three types of brand-specific images: brand selfies (photos showcasing a brand with an invisible consumer), consumer selfies (photos showcasing a brand that also include the consumer’s face), and packshots (standalone images of branded products). Ultimately, the type of image does matter when attracting consumers and increasing social media engagement and purchase intentions. More specifically, they found that while consumer selfies generate the most engagement via likes, faces are more powerful when it comes to purchase intentions. In fact, brand selfies received 78% more expressed purchase intentions on Twitter and 70% more expressed purchase intentions on Instagram.
Brand selfies received 78% more expressed purchase intentions on Twitter and 70% more expressed purchase intentions on Instagram.
Given the importance of the topic and its applicability to both academic scholars and practitioners, we were excited to learn more from the authors about their study.
Q: What is the most interesting trend that you’re following regarding social media right now?
A: E-commerce, standing for “entertainment commerce.” In the United States and Europe, e-commerce is still mainly treated as a utilitarian shopping experience. In some cases, in Asia people already use e-commerce as an entertainment experience, with influencers providing content with embedded shopping experiences in it. We are likely to see this trend of better understanding and use of influencers and micro-influencers to generate shopping opportunities. Understanding the role of selfies will play an important role in this new world. Moreover, NFTs could become a game changer for social media by allowing social media content producers to better capture the value of their content. Social media producers might be able to attain similar content rights as achieved by media companies or more expensive distribution channels, leaving more revenue to those who actually create.
Q: How can brand managers create selfie opportunities for consumers? What part of marketing do you anticipate having the highest ROI for “brand selfies?” For example, as a brand manager would you focus on selfies inducing packaging or experiential events or does it depend on the situation?
A: Brand managers can utilize our findings in three ways: (1) Tracking social media brand imagery to obtain early signals on how brand engagement is likely to evolve in social media. Our findings suggest brand selfies (ego perspective) result in more favorable observer responses than consumer selfies (portrait perspective) or simple packshots. (2) Brand managers can stimulate social media users to produce brand selfies, e.g., using dedicated packaging (e.g., the Santero “Hand” bottle) or contests (e.g., Corona beer’s “Find Your Beach” campaign or Starbucks’ Red Cup contest). Note that producing and posting brand selfies likely impacts brand engagement of senders as well. This needs to be taken into account by studying which type of social media brand image (brand selfies, consumer selfies, or packshots) results in the highest level of engagement of senders. (3) Our work calls for brands to move beyond simple measurement of social media engagement by likes and counting of comments and toward looking at the type of engagement, and particularly one that is targeted toward the brand rather than the social media content creator.
Q: Given your expertise in this area, are there any similarities or differences you would expect comparing photos to videos? Coming out of a rise of video in TikTok, how do you feel this could affect the landscape?
A: We have observed that consumer selfies result in higher levels of interest in the sender, while brand selfies drive brand interest. Videos might be able to capture the best of both worlds by allowing dynamic changes in perspectives. However, this needs to be explored by further research. Questions include the relative balance of different perspectives during the video, amount/frequency of changes, length of the video, and potential order effects. Also, it is conceivable that audio voiceovers can reinforce self-reference processing that we identify as a driver of brand engagement (e.g., “imagine enjoying this product,” “this could be yours”).
Q: How can brands utilize these learnings with influencers? Do you see similarities and/or differences with user-generated content from consumers?
A: Influencers are often well-trained and capable of choosing content to maximize reach and engagement with their profiles. Since we know that consumer selfies are more conducive to sender engagement than brand selfies are, we expect influencers to arrive at similar conclusions. This poses a challenge for marketers as high-reach users such as influencers are less likely to post the more desirable brand selfies than lower-reach average users with lower levels of experience. The optimal trade-off between reach and engagement is an empirical question marketers can address via experimentation.
Q: What was the most challenging part of coding all these images collected from social media? Are there any metrics you wish were widely available and utilized that currently aren’t being measured?
A: Only about 15–20% of brand images on social media feature a brand hashtag. This makes it hard to obtain a comprehensive view of visual brand presence. It would be desirable if reliable open-access solutions for brand logo detection would exist that could be applied to a representative sample of social media images. Approaches of computer vision exist that can characterize these images in various ways to better understand which types of images are most desirable. However, a necessary condition for all this research is a representative sample of social media brand images, ideally compared to relevant competition.
Read the full article:
The team would like to thank Jochen Hartmann, Mark Heitmann, Christina Schamp, and Oded Netzer for contributing their time and effort to the Journal of Marketing Research Scholarly Insights by AMA DocSIG. This article was anchored in their expertise and insights.