Thursday, September 25, 2014

One nerdy, yet cool technology – predictive analytics

It’s easy to see that technology is continuing to play a bigger and bigger role in marketing today. One nerdy, yet cool, technology is predictive analytics. Predictive analytics is a data analysis tool that takes data, learns from it, and then can predict decisions based on the data it has been given. Cool, huh?

For example, data can predict that a young child has high probability of touching a hot stove, but predictive analytics learns that once the child has touched the stove the probability of doing it again decreases.
Just as a child learns not to touch a hot stove, the same can be applied to consumer choices. For example, Target wanted to know which female customers are expecting a baby in the near future. To do this Target used predictive analytics to determine pregnancy based on purchasing patterns and behaviors.
Target started to “learn” purchasing patterns and behaviors from moms who have registered on the baby registry. Once the predictive model was generated, Target could then start applying it to Target customers who have not registered on the baby registry but have a significantly high probability of being pregnant based on purchasing decisions.
Target found 30% more female customers that were most likely pregnant and had not registered on the baby registry. With this data Target could then start promoting pregnancy-orientated promotions toward those women.
In 2012, the New York Times and Forbes Magazine reported about an angry father in Minneapolis enraged at Target for sending his high school daughter coupons for baby clothes and cribs. The unhappy father accused Target of trying to encourage his young daughter to get pregnant. A few days later the father apologized after learning his daughter was in fact pregnant.
This story depicts the accuracy of predicate analytic used by Target. Consumers should know, however, that Target carefully follows all privacy and data-use laws. The primary basis of Target’s modeling is from consumer purchases.
Pregnancy predictions are not the only way predictive analytics is being used, and Target is not the only company using predictive analytics either. In 2009, H1N1 was discovered and many public health agencies believed it would become a pandemic. At that time, no vaccine was available.
The goal of public health agencies was to reduce the spread of the virus. To do this the United States Center for Disease Control (CDC) asked doctors to report all flu cases, so the CDC could track the location of flu cases. Unfortunately this method had a two-week lag time.
Google shortly discovered that certain searches for flu-related information could show the spread of the virus. Out of the three billion search queries a day, Google was able to accurately identify areas infected by the virus. As you can see predictive analytics is highly versatile, effective and a little nerdy.
Target and Google are not the only masterminds of predictive analytics. A majority of major retailers, banks, grocery stores, etc. have all started implementing predictive analytics—even the Post Office uses predictive analytics. Technology is drastically affecting today’s marketing efforts and predictive analytics is just one nerdy, yet cool way.

(via)

No comments: