How descriptive models work
Descriptive models are used to quantify data relations and create segments. From this segments, publicity and personalised offers can be created.
For example, for Trump’s presidential campaign, data analysts determined millions of individual profile types. Here’s where the controversial case of Facebook and Cambridge Analytica comes up. With these profiles, the communications team created specific campaigns for each cluster.
According to the preferences of a citizen that is in favor of the right to bear arms, he was shown images of men protecting their home or hunters. The same would happen if he had a more adventurous or stay-at-home profile.
The objective –
Establish communication lines that would make potential voters loyal
The gathered data –
Profile information from Facebook users, shared content and likes
The analysis has uncovered a lot of supposedly illegal practices and even some that could be considered immoral, unethical or dangerous. Of course, it also forced a lot of people who work or are related to digital marketing to ask themselves if their activities could also be considered immoral, unethical or dangerous. Even though we often are relatively sure that campaign tactics are on the right side of the legality/illegality division, ethical exercises live in a great spectrum and are open to subjectivity and shading: How can we be sure that we are playing by the rules and also abiding by the law? Does it matter to us?
What do you think? Did you know about this case?
You may also be interested in:
Cómo funcionan los modelos descriptivos