Abstract

Social media platforms such as Facebook show ads with popularity or general endorsement signals such as “likes”. Additionally, these ads can display social endorsement from friends (i.e., friends’ “likes”). This paper examines the effectiveness of displaying these different signals on social media ads in generating user attention and actual conversions in the form of app installs through field experiments on Facebook. We partnered with a mobile app company and conducted an ad campaign on Facebook to randomly target unique user groups to install the mobile application. We find that the overall “likes” associated with the ad do not help the user’s decision on clicking the ad and the app-installing decision conditional on clicking. Further, ads endorsed by friends have a lower clicking and install performance as compared to the ones without such endorsement. However, the negative effect of “likes” on the app install performance conditional on clicking is attenuated if displayed simultaneously with social endorsement. We confirm our results using an ad campaign for another app. Further analysis, using an Amazon Mechanical Turk experiment and field experiment, suggests that a low similarity between users and their social connections likely result in the negative effect of social endorsement. Our results have implications for the design of ads on social media platforms.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call