The American supermarket chain Target made headlines a few years ago in the New York Times with its prediction of customer pregnancies . A father had complained to the company about his daughter – who was still in high school – receiving coupons for baby clothes. His tone was more subdued a few days later, once he'd learned his daughter was in fact expecting a child. Using data analysis of buying patterns, Target discovered what the girl had kept secret even from her family.
This discovery was amazingly easy: The pattern of products purchased – such as specific nutrition supplements and unperfumed shampoo – were evidence of a possible pregnancy. When all purchases are tracked using customer loyalty cards, this is not particularly difficult. The key term here is data mining: Big Data is analyzed by using algorithms to connect specific items in new ways, thus revealing surprising insights. Such as what the customer is currently interested in, or what she will want to buy tomorrow. And there's more than enough data – it didn't take Edward Snowden's revelations to tell us that. The transparent customer is often already a reality.
The advantages of comprehensive customer data analysis are clear: It lets companies place targeted advertising and thus gain customer loyalty early – such as with discounts on certain pregnancy or baby items, or eCoupons . And when these are used, they provide even more lucrative information. Cookies used to track Internet surfing can also provide specific evaluation possibilities. Predictive analysis tools can derive intelligent correlations from this data heap, and anticipate future purchases with a high degree of accuracy. Product recommendations from web giant Amazon are a good example.
This process is certainly not limited to retail. The same approach is possible in the financial sector, as shown by the example of Commerzbank: This bank wants to propose perfectly individualized offers to its clients based on smart data evaluation.
Companies can already use data mining to determine which products we will need tomorrow – before we ourselves are even aware of it. American athletics company Under Armour, for example, offers its fitness-focussed customers apps for recording and tracking their performance. The benefit for the company is being keyed into exercise and eating behavior patterns as a window revealing individual customer needs, e.g. when a guy needs a new running shoe, what sports a woman practices, or what kinds of weather they need clothing for.
But this is only the beginning. Customers' online behavior, their Facebook friends, as well as their exercise activities, where they live, where they drive and current weather conditions can – when properly put together – reveal even more than a desired product, whether they are expecting a child or want to buy a new home. Data mining achieves unimaginable levels of behavior predictability . Information that seems to have little significance on its own is combined to generate new insights – such as which products are preferred over others, how much money is available, social status, oft-frequented locations, and even possible illnesses. This fact will significantly shape our future.
The opportunities are huge: not just for business, agriculture, politics, traffic and healthcare, but for every area of life. But with such revelatory data, how we use it is a critical issue. Much remains to be seen, also with respect to data protection and data security.
How Big Data might influence our future, and which trends are on the rise in this area, is a topic at this year's CeBIT. Experts in the field discuss possible future scenarios for Big Data and data security at the CeBIT Global Conferences . The use of data mining is a focus of many additional panels, discussions and exhibition areas .