Data analysis and mining have been growing considerably in recent years. Almost every industry has found different purposes for big data and the tools used to analyze big data. These efforts have brought in massive revenues for data mining companies. However, there are some challenges created by the recent controversy over Cambridge Analytica. One firm has the potential to considerably shift the activities of data mining companies that work with advertisers and social networks. Data mining companies have to adapt if they hope to survive this issue.
What is Cambridge Analytica?
Cambridge Analytica was a large, politically-motivated data mining operation that was in existence for several years up to their recent folding in early 2018. They were a firm that specialized in voter targeting. The firm was bankrolled by conservative donors Robert and Rebekah Mercer and guided by conservatives like Steve Bannon. This company became deeply entwined with the 2016 presidential election and the Donald Trump campaign.
Cambridge Analytica allowed the Trump campaign to micro-target voters and use a data system that was on par with the more well-funded Hillary Clinton campaign. Analytica’s connections to Trump multiplied immensely when Steve Bannon became the Trump campaign CEO in August of 2016. The formula for Analytica’s success did not become clear until media reports surfaced in early 2018. Those reports argued that Analytica actually gained their sophisticated voter profiles through manipulation of Facebook’s data.
The Analytical Approach
A British professor created an app on Facebook that gave individuals cheap personality tests for free. Individuals who signed up for the app had all of their personal data made available to Analytica along with the personal data of all of their friends. Then, Facebook changed their rules and restricted the ability of companies to collect such a large amount of data. They moved away from Analytica’s approach and requested that Analytica stop the practice and no longer use their data. However, Facebook provided no enforcement mechanism to ensure that Analytica would delete the data they had already illicitly collected. Analytica’s usage of this data eventually helped lead to Trump’s electoral victory.
This lack of oversight and broad policy towards the privacy of their consumers led to a considerable scandal. Analytica was so tarred by the experience that they were forced to close. Mark Zuckerberg, the founder of Facebook, was hauled before Congress and testified for several hours. Politicians who had otherwise not cared about Facebook were now talking about potentially imposing regulations about how the social network collects data from its users. The effort led to a considerable sea change in the way individuals viewed social networks and data mining in general.
What has Cambridge Analytica Done?
Analytica has curbed the ability for data mining companies to monetize from social media. Previously, this approach had been particularly profitable. Companies could offer tools that mined large quantities of data to micro-target a massive number of individuals through advertisements. There was almost no upper limit to this data collection and little understanding of rules and controls.
Companies could compete with other companies on simply how much data they could collect in general. There was an understanding that data analysis companies could collect more and more because an individual was using a free service. While data breaches and poor data security measures could make the news, simple established data analysis business practices rarely made the news.
Now, individuals and regulators are more careful about data overuses and breaches. Business practices that would have earlier been ignored are now causing consumer outrage in the media. Every company that offers a free service based off of data mining has, therefore, become suspect. Companies are cutting into their bottom lines to restrict their business services. They are controlling this effort, not because of business reasons but to pacify lawmakers and prevent them from regulating the industry. Such an effort takes a considerable amount of time and money. It also distracts the company from being able to focus entirely on advancing their technology.
What can Data Mining Companies do to Respond?
Companies that provide data mining software have to respond accordingly. First of all, they must create suites that require explicit buy-in from the individuals involved. There must be safety procedures that allow for individuals who do not want their data to be shared to opt out of any data sharing system. This approach will only reduce the information by a small amount.
Many individuals are fine with sharing their data for what is otherwise a free product. But the days of data being shared without their consent are over. Also, data mining companies have to restrict their collections to individuals who have signed up for a particular platform. Collecting data on individuals that have no connection to Facebook or some other company will cause a massive amount of problems for a data mining company.
Data Mining Algorithms
Data mining algorithms must be fine-tuned to provide quality rather than quantity. The workings of the algorithms were one aspect of Facebook’s data mining procedure that got them in trouble. Since the algorithm worked better with more data, the company offered a massive number of ways in which a client could micro-target individuals.
Inevitably, some of those potential categories were deeply offensive. Data mining algorithms need to respond by offering more dictated solutions to customers. They can compete to analyze the same amount of data with their own proprietary formula. Data mining companies could go to advertisers and argue that their particular parameters and categories are the most effective. Then, if the advertiser is not happy with the company’s selection, he or she could simply move on to another company.
Finally, big data firms need to find more ways to monetize. They should prioritize detailed reports and logistical analyses that bring in revenue along with advertising. In addition, companies such as Facebook are moving towards fee-based models where they do not have to constantly update all of their advertising approaches. Data mining companies could help social networks target different services for different prices. It could help produce and analyze surveys suggesting how much people would pay and if they would spend extra money on different services.
What Have We Learned from Cambridge Analytica?
Data analysis companies have adapted to considerable technological changes before. These companies have moved from labs to limited companies and then to the entirety of the global economy. They can most certainly adapt again to governmental and regulatory changes. Data analysis companies will have to shift their business practices and their revenue streams in order to adapt to this new situation. The process will take a considerable amount of time and may require a number of hiccups. It is understandable that data mining companies are somewhat alarmed. But in a cutting-edge industry like this one, they are certainly up for the challenge.