Annual Conference 2019

Facebook and Instagram became subject of debate as it was revealed that the Internet Research Agency (IRA) advertised on the platforms extensively before, during, and after the election. Many have shown that Russian-backed ads had broader aims than electing a political candidate. Still, little work has explicated whether content was successful in engaging audiences. Benchmark data gathered shows that Russian advertisements were engaged with dramatically more than typical. Here we point to the microtargeting efforts of the IRA. Messages were tailored and targeted to specific audiences, which appears to have boosted engagement. As more targeting parameters were used, clicks increased. Moreover, IRA ads were not subject to industry or regulatory oversight. Ads were unprecedented in the use of insults, hatred, and threats. A computational analysis of negativity through Google’s Perspective API shows ads indeed demonstrated all three characteristics, but specifically insults boosted engagement. Untraditional, pejorative descriptors for minorities and groups may have stimulated audiences to engage. In all, this study provides a proactive method in which advertising platforms can screen for new types of negativity and limit the exposure these ads have.