How Cambridge Analytica managed to influence American Voters using their Brand Preferences

Cambridge Analytica, "special information-sharing service", partnerships with the US Government, Facebook? User Profiles? 50 million...87 million...120 million...fake news...apologies...more fake news...more apologies...and the headlines went on...

You gotta agree. 2018 was not quite an year for Facebook.

The 2018 Facebook Timeline (No Pun Intended)

In January 2018, when Zuck posted a New Year's Resolution to "fix" Facebook's failings, no one could have possibly imagine what fate had in store for him, or for those 87 million unlucky users, for that matter. The matter started in February 2018, when some Russian organizations were revealed to have interfered US political and electoral processes, including the 2016 presidential elections. The Facebook ads cited during this time, it was later speculated, had been seen by more than 11.4 million Americans. In March, the renowned The New York Times, revealed in a breakthrough article that a British consulting firm, Cambridge Analytica, had harvested 50 million Facebook user profiles for use in political campaigning in the 2016 elections. The number of breached accounts later increased to 87 million. Not only this, the UN, investigating genocide of Rohingya Muslims of Myanmar, also claimed that Facebook played a major role in spreading hate speech against the Muslims.

Amid all these speculations and revelations, Zuckerberg in April, apologized before a joint hearing of the US Senate for Cambridge Analytica failings. He admitted Facebook scans the links and images that people send each other on Facebook Messenger, and reads chats when they're flagged to moderators. And if this wasn't enough, Facebook later admitted that it shared data with four Chinese smartphone makers: Huawei, Lenovo, Oppo and TCL. In June, a bug in FB's system caused it to publicly share the posts of 14 million people who thought they were making private updates, for which Facebook later apologized, again. In July, Facebook was fined €500,000 by Britain's Information Commissioner's Office over the data misuse. Facebook shares suffered the biggest one-day wipeout in US stock market history, dropping $US120 Billion. In September, hackers breached Facebook's database again, taking names and contact details of 29 million users. In December, 223 pages of private communication between the FB officials was published, showing how it shared user data with companies including Netflix and Airbnb, and discussing generating revenue by selling access to user data.

It seems Zuckerberg and Sandberg, the faces of the company, have quite expectedly, lost goodwill, which would take more time than expected to restore. Over the course of the whole year, serious allegations like Data misuse, the proliferation of fake and misleading news that lead to disputes and genocides, have plagued the company. The company also faced the exit of a lot of its prominent officials, including WhatsApp co-founder and CEO Jan Koum, Instagram co-founder and CEO Kevin Systrom, and Oculus co-founder Brenden Iribe. Facebook is gradually running out of customers, it seems, since after years of steady growth, the stocks began to decline this year. Some users are deleting their accounts, others are simply using it less.

Now here comes the interesting part of the article...

The deep thinkers wanna dig deeper. A question that's on everybody's mind:

How It All Works?

My dear reader, have you ever heard about this term called profiling? It's like categorizing information from a huge array of unorganized data, as in criminal profiling: pickpockets, robbers, thieves, murderers, rapists, etc. Now hold your senses dear, because what I'm gonna reveal would be hard to believe for you. Surprisingly, Fashion Profiling played a very important role in 2016 American presidential elections, according to Christopher Wylie, the CA whistleblower. Now what is Fashion Profiling? The practice of targeting individual profiles based on their clothes brand preferences. You see, your choice DOES matter in this world, after all. So you should be happy , in a way that it means a lot to Facebook.

OK, back to Chris now. In November, at a fashion conference held in Britain, Chris revealed that clothing preferences was a key factor in determining the business of Cambridge Analytica: selling voter profiles drawn from Facebook database. He said to the media that, Fashion data from Facebook was used to build AI models, which reportedly helped Steve Bannon identify conservative backers. Preferences in clothes and music are the leading factors in political leaning, he said. The narratives of various American fashion brands, which play on the legend of Western frontiers, are also the narratives of various political parties. Those who choose on the former are susceptible to the latter. He mentioned Wrangler and LL Bean, in particular, that CA aligned with political traits. Thanks to latest Machine Learning & Data Mining techniques, it is possible now to find out how people think and feel, based on algorithms of their preferences of various fashion brands. For instance, Humberto Leon and Carol Lin, the founders of fashion brands Kenzo and Opening Ceremony, appealed in 2016 to their customers to vote for Liberals, automatically turning many liberal supporters' inclination towards the brand. 

It was noted that more organized and aggressive use of fashion data was groomed and mined regularly by the political candidates themselves to steer the political campaigns in their favor. Like if a user encountered an ad about Tees from Hilary Clinton's campaign brand on his Facebook feed, and he clicked on that ad, the data would automatically be stored on CA's servers. Purchases, or even the number of clicks made on each candidate's campaigning brand, could clearly identify potential issues that could galvanize a voter.

"It's all about learning who your supporter base is." said Marshal Cohen, chief industry analyst of the NPD group during the 2016 elections, "Why customers do what they do? How do they live? What are their trigger points? What words resonate with them? It's worth its weight in gold, in the political field, just like in marketing or sales. We call it demographic profiling."

Clothing choices says a lot about an individual's sense of behavior and personal desires. Cambridge Analytica preyed on that human reality via algorithm, using data from Facebook profiles of more than 50 million people without their permission. Fashion Profiling is another facet of this arena, using the concepts of ML & Data Analysis to determine the way brands are perceived among the common people, particularly, with respect to the pop culture. The infamous data breach plunged a tech firm such huge into hot water as it sought to clarify how so much of its users' data could have been used in manipulating a campaign such large.

Assessing values, goals & priorities, based on the clothes people wear, has been a tradition, be it in the professional arena, or political arena, for centuries. What is Fashion Profiling, you ask? The "dress for the job you want", the "dress you're gonna wear in your campus interview" is an expression of fashion profiling. Calling someone a "Gucci Person" (in Pop Culture) is fashion profiling, adopting the Levi's brand instead of Tommy Hilfiger in your lifestyle makes a statement about your associations and personal preferences and opens you up to fashion profiling.

This event was indeed a brilliant example of how personal data, pertaining to one's shopping choices, can be used manipulatively in unanticipated and potentially damaging ways. Wylie said in the conference that Facebook, in a way, was damaging society by separating people based on their cultural preferences while the opposite must have been true. Most users, be it on FB or Instagram, are more concerned with sensitive data theft rather than being victims of targeting political campaigning, based on the trails of cultural crumbs they leave deliberately online to be used against them later on. Nothing can be speculated right now about the rapidity at which Artificial Intelligence and Machine learning are expanding, or how they would turn out for the people in future.

:)