Effects of Information Provision in Digital Markets

Bairathi, M (2023) Effects of Information Provision in Digital Markets. Doctoral thesis, University of London: London Business School.

Abstract

Digitization of economic activities has transformed the way businesses operate. Continuous changes in digital markets, coupled with the explosion of available data, represent an exciting opportunity to develop a better understanding of the roles of different marketing tools in digital markets and its implications for consumer behavior, firm profits, and policy. Consequently, this dissertation is organized around two main themes: (1) platform economy and (2) online advertising.

The first chapter examines platform endorsement in the context of one of the world’s leading freelance platforms. Many digital platforms with large product assortments endorse a selected group of items to facilitate user choice. While it is intuitive that endorsed items may enjoy considerable benefits from increased sales, little is known about the effect of such platform endorsement on unendorsed items and on the platform. Using data from a field experiment conducted on an online freelance platform, I examine the effect of platform endorsement on user search and purchase behavior. I find that platform endorsement leads to an increase in search and purchases not only for endorsed services but also for unendorsed services. I find that this increase in search and purchases is mainly driven by an increase in overall quality perception of the services offered on the platform. I further explore heterogeneity in the effect of platform endorsement and find that the effect of platform endorsement on purchase is more pronounced for users with a higher propensity to purchase. I discuss implications for platforms, merchants, and regulators.

In the second chapter, I focus on a recent innovation in advertising – influencer marketing. The recent growth of the influencer marketing industry means brands are becoming more likely to contract with influencers. However, there is little empirical evidence regarding 1 consumer engagement with sponsored content relative to organic content. In this paper, I examine whether consumers engage less with sponsored content relative to organic content. Moreover, in light of the recent regulations across the globe mandating influencers to disclose advertising in sponsored content, I examine the effect of advertising disclosure in sponsored content on consumer engagement. I collect a dataset of 180,404 posts created by 510 Instagram influencers operating across ten categories. I identify sponsored posts in the dataset using advertising disclosure and supervised learning. Leveraging timing of regulatory actions and industry-level advertising trends as instrumental variables, I causally identify consumer engagement, measured by the number of likes, with sponsored content relative to organic content. I find that consumers engage less with sponsored content relative to organic content. However, the results show disclosure of advertising in sponsored content increases likes for sponsored content. Given the popularity of influencer marketing with brands, I further explore what characterizes successful influencer content. I rely on previous theory in consumer psychology and argue authenticity of content attenuates the negative effect of sponsorship on likes. I measure authenticity as – topic alignment of a post with other content shared by the influencer, influencer’s propensity to share brand related content, and the number of times a brand is mentioned in the post. I find that authenticity of content fully mitigates the negative effect of sponsorship on likes. My findings are relevant for regulators who are concerned about lack of advertising disclosure in influencer marketing and can also inform influencers and advertisers on their content creation strategies.

In the final chapter, I investigate systematic differences in online ratings based on gender in the context of an online labour market. Ratings have a direct effect on sales. I leverage a unique dataset from an online labour market that elicits both private and public ratings from buyers after completion of a job. While public ratings are displayed as star ratings on the website, the platform uses private ratings in internal evaluations. I find that conditional on having the same private rating, female freelancers receive lower public ratings compared to male freelancers. Further, I demonstrate that gender bias in rating is more pronounced 2 when buyers hail from locations with greater gender inequality. This suggests that buyers’ online behavior can reflect of their cultural biases. These results are important for platforms and merchants because systematic differences in consumer evaluation based on gender can lead to undesirable consequences for platform participants.