Professors Students Researchers Enthusiasts

DIGITAL ADVERTISING

How would the optimal strategy be affected by economic inequality? We explore this using field experiment data from Facebook in India.

In these field experiments, consumers were shown ads that encouraged them to join an online community that shared reviews of wedding vendors. These ads randomized the extent to which the images of weddings showed global versus local aspirations. In general, the globalized images performed poorly. However, this is moderated by the degree of inequality in the surrounding region. People living in more unequal regions are more likely to embrace globalized images. We present suggestive evidence that this has to do with consumers trying to distance themselves from an environment of poverty, rather than because inequality is correlated with other regional characteristics that make the globalized image appealing.

Service Quality

Digital platforms and firms in the sharing economy routinely invest in service- quality improvements and still grapple with service quality inconsistency.

In this project, We have used unique data from an online medicine delivery platform in India that relies on delivery persons for their service delivery. Exploiting idiosyncratic variation in delivery person allocation, we found that when a customer encounters a well-performing delivery person, she/he will give a poor rating to the delivery person in the next encounter and vice-versa. This cyclical nature can be explained by the customer’s prior expectations about the next delivery person based on her previous encounter with a delivery person. We further show that this effect is moderated by the nature of the ailment and the monetary value of the transaction.

Retargeted Advertising

In collaboration with an online seller of home-improvement products, the authors conduct a large-scale randomized field experiment to study the effects of retargeted advertising, a form of internet advertising in which banner ads are displayed to users after they visit the advertiser’s website.

Results show that switching on experimental retargeting causes 14.6% more users to return to the website within four weeks. The impact of retargeting decreases as the time since the consumer first visited the website increases—indeed, 33% of the effect of the first week’s advertising occurs on the first day. Furthermore, the authors find evidence of the existence of complementarities in advertising over time: the effect of advertising in week 2 of the campaign is higher when the user was assigned to a nonzero level of advertising in week 1. The authors discuss mechanisms that can explain their findings and demonstrate a novel low-cost method that can be applied generally to conduct valid online advertising experiments.

Paid Search

We investigate the causal effect of position in search engine advertising listings on outcomes such as clickthrough rates and sales orders. Because positions are determined through an auction, there are significant selection issues in measuring position effects.

A simple mean comparison of outcomes at two positions is likely to be biased due to these selection issues. Additionally, experimentation is rendered difficult in this situation by competitors’ bidding behavior, which induces selection biases that cannot be eliminated by randomizing the bids for the focal advertiser. Econometric approaches to address the selection are rendered infeasible due to the difficulty of finding suitable instruments in this context.

We show that a regression discontinuity (RD) approach is feasible to measure causal effects in this important context. We apply the approach to a large and unique data set of daily observations containing information on a focal advertiser as well as its major competitors. Our RD estimates demonstrate that there are significant selection biases in the more naive estimates. While mean comparison of outcomes across positions would indicate very large position effects, we find that our RD estimates of these effects are much smaller and exist only in some of the positions. We further investigate the moderators of these effects. Position effects are stronger when the advertiser is smaller, and when the consumer has low prior experience with the keyword for the advertiser. They are weaker when the keyword phrase has a specific brand or product information when the ad copy is more specific as in exact matching options, and on weekends compared to weekdays.