Targeted Advertising and Election Result Implications
The 2016 election was controversial due to unconventional campaigning tactics and rumors of Russian interference on social media. Although fake news is mentioned tirelessly when reminiscing about the election, the simple use of targeted advertising on social media, specifically Facebook, is rarely discussed. Just as retail companies use “cookies” to advertise their products in a targeted way on social media, Donald Trump and Hillary Clinton used targeted advertising based on available personal Facebook information, such as gender, political affiliation, and location. By looking at the micro-targeted advertising tactics used as a whole, regardless of their content, Trump’s targeted advertisements were extremely effective in persuading undecided voters and registered Republicans to go out and vote for him. There was no statistically significant information about Clinton’s targeted advertisements, but these ads used a Facebook toolkit that harnessed predictive analytics to specify which users would be good targets for which voting advertisements. These algorithms take into account your search history, previous engagements with advertisements and news articles, and personal information from Facebook to determine whether you are a good target for various advertisements.
Targeted advertising works by harnessing the ideas of matching markets and auction pricing as discussed in class to assign a value to a Facebook user. That value is then used to determine what targeted advertisement, if any, is shown to what users and what the advertisement should cost. Targeted advertising, especially in the case of election cycles, is a smart way to utilize the widespread influence of the World-Wide Web by spreading an idea to specific people in Facebook’s network. Once specifically targeted people in the network have received the information from the advertisement, they are more likely to spread that ideology to other people that would not have been as easily persuaded by a simple targeted advertisement. This models the concept in the last question of problem set 8 where, depending on who is friends with who, an idea can either spread to an entire group or stay contained in one space. By using targeted advertising during the 2016 election, Trump was able to take advantage of Facebook’s large network of users and targeted those who were most likely to be swayed as well as those who already were strong supporters, in the hopes that he would cause an information cascade where these users would convince their friends and so on to also vote for Trump. Studies showed that this tactic increased the probability that an undecided voter would vote for Trump by 5% which was an influential amount in such a close election thanks to targeted advertising and Facebook’s social network capabilities.
https://phys.org/news/2018-11-impact-facebook-advertising-elections.html