In the course of recent years, the use of personal data by political parties and campaigns in the context of national elections and referendums has risen vastly and proven to be a useful tool for successful political online communication. And yet, these practices raise serious ethical questions which have been widely discussed in international media, in particular in light of the far reaching Cambridge Analytica scandal and the 2016 US election. When thinking outside of the box, however, personal data use as a tool of political marketing has not only taken place in North America and Europe – with Brazil’s recent elections being a prime example.
Bots sending out invites to a WhatsApp group of Jair Bolsonaro (Brazil’s current President) supporters and Instagram accounts flooded by election advertisements have been common practice in Brazil’s 2018 election campaign, where personal data usage has played a key role. According to the Global Digital Report (2018), Brazil is the third country with the largest number of active Facebook users (130 million), and the second one in terms of WhatsApp (120 million) and Instagram (50 million) users. As a consequence, social media platforms inevitably played a major role in the elections. After an Electoral Reform in 2017 had led to new rules for political online advertising, sponsored electoral content produced by candidates and their parties was, for the very first time in 2018, allowed to be boosted on social networks.
Collecting, Clustering, Communicating
What followed came as no surprise: So-called data-brokers, companies specialised in analysing consumer data, and advertising agencies began to use individual data such as WhatsApp contact phone numbers and socio-economic profiles “scraped off” of social media to gather information about (potential) voters for the sole purpose of tailoring ads specifically to the target audience. Afterwards, they applied five different data mining techniques to prepare the data:
- mapping audiences
- clustering by niches and patterns
- identifying consumption patterns
- defining impact drivers amongst users
- gaining insights and making recommendations
Data brokers, Numbr Group for instance, used bots to scrape data off of people’s social media profiles, such as geolocation and likes. Then, the data firm identified groups of people who were, as an example, either in favour or against military intervention in Rio de Janeiro. With that initial information clusters were formed and data was mined. Political strategists were then delivered a handbook on for what these people advocate, the subjects that they are talking about, whether videos, pictures or texts are more appealing and received a media plan on how to segment communication in order to reach this group of people exclusively.
How to (Ab)Use Data
We might ask ourselves, what’s the entire fuss about? Big Data has been used for the purpose of political campaigning before. Yes – but to such an invasive extent? Not only do ad targeting practices and personal data use to improve online campaigning expose voters to much more micro-targeting, resulting in potential public fragmentation and less access to balanced information; they also grant data firms access to personal phone numbers and e-mail, leading to a massive invasion of people’s private sphere. The more private the environment which a conversation is held in is, the more susceptible people are to the message itself – which makes it even more challenging to resist its content, even if the initial invite was sent by bots.
Yet, it’s rather a question of how this data was deliberately (ab)used, given that every piece of personal information was and still is available for political parties, campaign strategists or anyone who can actually buy private data. So, where does that leave us? Where can we draw the line between what’s personal and what’s public data? The answer, at least for Brazilian data firms, is clear: There is no difference.