The Importance of Data Collection for Businesses and Their Customers

  As data driven businesses continue to grow, so does the amount of data collected by these businesses. Companies that thrive on technologies driven by data collection, analyze that data and improve their technology with the help of A/B testing. A/B testing is a revolutionalry experimentation platform that is used by nearly every technology company today. A/B testing helps find the best version of a platform or product for a particular individual. It allows companies to grow their customer base and enables its customers to receive personalized services. This paltform is used by stores, campaigns and even schools. For example, A/B testing is used to make movie recommendations for Netflix, Amazon product recommendations, advertisements and many other personalized results and even visual cues to which users are most responsive.

  A/B testing helps develop algorithms that predict people’s preferences. First, vast amounts of information are collected on a user’s behavior. For example, a person hovers over a button, scrolls the page to see a product, reads a description or lingers on a particular part of the page longer, these actions emit an event that identify the actions and these events are logged. The person does not even need to be logged on to a specific website or account to be identifiable, because all of the logged actions carry a ton of identifiable meta information such as the computer’s IP address, browser version, agent, type of action taken, and geographical location. These data points, by themselves, could be enough to guess a person’s identitity, but more often than not, the data is not isolated. Events get saved and often combined with data available from third party partners. This aggregate of data is then analyzed to determine that person’s behavior. Algorithms statistically evaluate these experiences and compare them to others, analyze resluts and draw conclusions about what version of an advertisement would most affect that person, or what is the best recommendation to that person for a product, or what content will most influence that person. This insight into individuals’ behavior allows companies to make intelligent marketing decisions and improvements to websites, services and mobile applications.

  Once explained to an individual, most people become uncomforatble with this process of data aggregation. After all, even though our society grows increasingly dependant on wireless technology in business and personal affairs, most people still adhere to the notion that if noone is physically present to see their actions then those actions remain relatively private. This uncomfort grows and sometimes even turns to disbelief when a person, with limited knowledge of how their data is used, is confronted with the fact that collected data is not only used for recommendations and advertising but is also sold to companies to help them make decisions about the the scope or cost of a persions’s medical, life or auto insurance coverage, credit extenision or to influence people’s political opinions.

  In a TED talk presented by Zeynep Tufekci at TED Global in September 2017, Ms. Tufekci expertly explained the influence artificial inteligence technologies that use A/B testing models have on our life. Ms. Tufekci spoke about how the aggregation of data about people’s online and offline activities is used to figure out their preferences and how the personalized results provided by these algorithms don’t just show people more of what they like, but have the ability to steer people’s views into a particular direction. Ms. Tufekci talked about how in order to keep us on the site longer, companies like Facebook, Youtube or Google show us the content that we are more likely to keep viewing. In practice, when people are viewing certain content, they are more likely to keep watching the videos or viewing articles that show a more extreme version of the previously viewed content and, in turn, other points of view are then eliminated from what is shown to that individual because the viewer is less likely to click on that content. Therefore, that individual’s beliefs are shaped and reinforced by this one-sided information that over time becomes more and more radical.

  The reprecussions of such algorithmic recommendations are grim. With manipulation of policitical and social information, our political and social views are also manipulated. This technology, however, is not all bad. Rather, in many ways it’s necessary. In today’s data driven society, the monetazation of data allows us to enjoy free valuable products and technologies. The mining and collection of data also lets us find what we need faster and more efficently. Data driven products are developed to make our experiences better, faster, and more conveniet.

  How do we ensure then, that the data we give away so freely is not used against us? It may seem that there’s little we can do when internet is such an essential part of our society that using services provided by companies like Google, Amazon, and in some instances even Facebook, is no longer optional, but necessary. It is also not feasible to make companies use our data only in a way that would be beneficial to us, because delineating that type of use would present a myraid of challenges, competing interests and perpetual debate about the scope of these benefits. One thing we can do, however, is to enable the Do Not Track option in the settings of the browser. The Do Not Track option was one of the recommendations proposed in the FTC’s 2010 preliminary staff report. The Commission, however, has not voted yet on whether to support this idea, and it is still considering public comments on the report. Presently, unfrotunately even choosing this option will not eliminate all tracking and aggregation of data by websites and service providers. The more effective method, therefore, would be to build trust between technology companies and their users. To effectuate that trust, individual users should be more informed about the scope of their data collection and more in control of their data. One way to establish this result woud be to follow in the footsteps of the European Union’s General Data Protection Regulation and give logged in users access to see all raw data stored about themselves, as well as allow users to initiate “forget me” actions from site admins. Companies should also ensure that data is secure and properly managed once it is collected. Though these steps are just a start, they would ensure that our data driven world is headed in the right direction.