Data Mining – what do the numbers mean?

Ever wonder how Google, Facebook, YouTube selects ads to your interests and hobbies? For example, I like road cycling and when using the Google Search engine, Google will display ads on the right-hand side bar that will link me to online cycling stores such as Chain Reaction Cycles or Merlin Cycles. This is an example of Google Data Mining users web browser cookies and previously search items. Google names this service Google AdSense, and it is one of the main drivers of the company’s revenues. While this is an interesting topic, the main focus of this blog post will be on Data Mining conducted by Target, which effectively predicted when woman are pregnant, and also estimated the month of when the baby is due just based on the purchasing habits of their customer

So how did Target achieve such a feat? Let’s start from the beginning. Every time a new customer makes a purchase with a credit card, a purchasing profile is created for the customer and every future purchase will be added to the profile. As individuals shop more, Target starts to collect data on what each customer is buying. Then, a team of statistician are hired to make sense of the data and predict future buying habits of the company’s customers. Data mining is essential for marketing. It allows companies such as Target to create specific advertisements for products based on the customers buying habits, thus creating a unique and direct marketing communication stream for each individual costumer. Data mining makes marketing more effective for the company, as the company can cater to the unique needs of the target market, rather than trying to capture the market with a broader marketing strategy.

You can have a read of the Target article here:

If you want to know more about data mining processes and how companies learns their customer’s secrets through numbers, refer to this New York Times article which provides a comprehensive high level summary: