In the advertisement industry, machine learning could be a great asset for a company. It helps determine patterns among consumers, which allows the marketing team to alter advertising in effective ways. However, when it comes to advertisements online, targeted ads are becoming a huge controversy. It means advertisement companies track what you do online in order to suggest certain websites or products. They use your past searches as a base for what you might click on in the next five minutes or next week. They also collect data about your demographics in order to determine how much you can afford and when you might be able to afford it. This information could come from any company that sells your information to third party companies or from social media accounts – Facebook, Twitter, Instagram, etc. With the abundance of data used in order to successfully draw you in, it is obvious a human is not behind it, but machine learning. It works by collecting a huge amount of data in order to make accurate predictions about what you are interested in. The more information the algorithm has, the better it learns and the more accurate it becomes.
By using machine learning, marketing teams and companies gain a huge advantage. They are able to learn more about who their consumers are and how they think. It helps, “marketers analyze countless signals in real time and reach consumers with more useful ads at the right moments.” Jeff Rajeck creates a systematic outline that shows how to integrate machine learning into a company. First, one must find the features of the ad – platform, the copy, photo, etc. Next, one has to identify the results of an ad. The third step is to gather the right data that will cover the features. Once one has the correct data, they can then pick a machine learning program. Then it would be beneficial to split the data in order to have one set for learning and the other for testing. The sixth step is to run the algorithm and see the magic happen. It will then show predictions for which ads tend to draw in the desired consumers. The system will be able to improve itself over time and become more accurate with the data it collects. This ultimately changes the whole marketing industry allowing for more audiences to be reached.
At times, machine learning can get too good at its job. A Target advertisement system was able to determine a teenager was pregnant by her web searches before her father found out. They sent coupons to her house in hopes she would shop there and become a lifelong customer. When a customer purchases from Target, they create a profile, which assigns a Guest ID number that includes their credit card, name, or email. It also collects demographic information that Target acquired from the consumer or bought from other sources. The young mother-to-be was most likely researching things related to her pregnancy. If she was using Google, then they presumably sold that information to Target. Everything we do online is tracked and the data is sold to other companies in order to produce more personalized advertisements. Target statistician, Andrew Pole, analyzed historical data of women who signed up for Target baby registries. By using those past purchases as a base in a test, patterns began to emerge of specific products pregnant women bought. Pole’s system was also able to estimate when their due dates were within a small window, which allowed Target to send out coupons tailored to the stages of the pregnancy. People became aware of how in depth Target was, and were taken aback at how much they knew. Now, the company mixes their targeted coupons and advertisements with random ones, therefore, it is not as obvious. The more information a system has, the more accurate their predictions become.
Though using machine learning in marketing could be extremely beneficial, there are a few ethical implications that arise. People usually love tailored material in order to make their shopping experience much easier, however, is it worth it if you are constantly being tracked? Is it worth it if your web history is being sold to third party companies? Is it ethical for companies to make a profit off consumer information? These are the types of questions people must ask themselves and if the answer is no, then they should take the necessary steps in order to protect themselves. Machine learning will continue to be a controversial topic and the ethical implications must be discussed.