How Joe Biden’s team leveraged data analytics in the US elections — CrunchMetrics
With the unprecedented challenges brought on by the coronavirus pandemic and record voter turn around, the 2020 US presidential race between Joe Biden and Donald Trump was one of the most-watched elections of all time. A lot has gone into the political campaign that helped in Biden’s landslide victory with nearly five million vote lead. Data also played a critical role in his election campaign. Biden’s team worked with Civis Analytics — an analytics platform backed by former Google chairman Eric Schmidt to optimize their online presence and rally their cause.
The role of artificial intelligence and machine learning in elections
The application of data science in predicting electoral behavior is not a new concept. Political campaigns of today are strategized and executed based on advanced AI-enabled forecasting models and systems. Political candidates have used data science and predictive systems to forecast electoral polls and drive winning campaigns. In addition to analyzing voting habits, advanced analytics helps in targeting voters with the right messages on the right channels.
Statistical models and advanced analytics were used in the 2016 political advertising campaigns by both Hillary and Trump. Even Barack Obama had a strong data analytics team to support his campaigns back in 2008. The 2020 US elections were no different. Advanced analytics has helped political parties to go beyond just basic polling research. Here are some ways in which Biden’s team leveraged analytics for his campaigns:
1. Analysis of complex electoral data to gauge voter sentiment
Voting habits of citizens are influenced by multiple variables. To predict voter behavior, parties must analyze voter data from multiple sources such as Census, social media channels, and third-party platforms. By leveraging artificial intelligence, political parties can gather and analyze large volumes of behavioral data to gauge the popular sentiment and probable response of voters to important issues such as immigration, healthcare, economic crisis, pandemic, and racism. This information helps in crafting campaign messages most suited to a specific type of demographic. For instance, in the recent US elections, Joe Biden’s team launched Ad campaigns focused on economic growth to win the confidence of working-class voters in Ohio.
2. Microtargeting to attract floating voters
Elections are won by majority votes. In addition to retaining their supporters, political parties have to woo floating voters or votes of people who have not yet decided their votes. Machine learning algorithms help in microtargeting of Ad campaigns to attract these voters. Political candidates can also get the early visibility of the states that need more campaigning effort. This is especially useful in winning the votes of undecided voters in states that have no clear majority. Voter results in swing states often decide the fate of the entire election. In the 2020 presidential elections, Biden’s team focused aggressively on battleground states such as Arizona, Michigan, Pennsylvania, and Wisconsin with targeted video ads like “Unprepared”.
3. Social media sentiment analysis
People today are more likely to express their political opinions freely on social media. Political parties can leverage advanced analytics to monitor rapidly changing engagement metrics such as popular hashtags, relevant discussions from several social media websites on the internet to strategically respond to the information. In real-time, candidates can monitor the changes in sentiment to know how voters might respond to a specific campaign message. To gain the support of young voters, Biden’s team had extensively worked on improving his online presence including recruiting influencers in the months prior to his victory in the US elections.
4. Data analytics for digital Ad strategies
More political parties today are moving from traditional advertising channels to digital ones that offer the option of personalization. In fact, digital ads form a major part of the campaign budget. Just the Facebook Ad spends by Trump and Biden ($ 107 Million and $94 Million) individually is more than the combined Ad spends by the candidates in 2016 election ($81 Million). AI-enabled analytics enable political parties to autonomously ingest, monitor, and analyze streaming ad data from popular social media channels such as Facebook and Instagram. By tracking key metrics in real-time such as impressions and click-through rates for ads, parties can optimize their advertising spend and reach more supporters online. Campaigners can go beyond the basic demographic metrics such as location, age, etc to target highly personalized social media ads. AI enables parties to run different variations of the digital ad based on the viewer. Kamala Harris’s team launched a series of ads to reach different voter demographics including women and Black votes.
Data science helps political parties in gauging their popularity, the attractiveness of their campaign message, and in choosing the right channels for targeting voters. Algorithms can help political candidates in understanding voter engagement metrics and popular sentiment. It essentially helps in understanding the voter sentiment across states. In the end, it is the voter’s intend that decides the majority in a democracy — the united voice of the people for one great, strong nation.
“Democracy is the government of the people, by the people, for the people.” — Abraham Lincoln
Summon the power of Augmented Analytics to identify risks and business incidents in real-time with CrunchMetrics.
Originally published at https://www.crunchmetrics.ai on November 13, 2020.
