Predictive Analytics and Big Data in Real World

In 2012, national retailer Target sent some customers pregnancy coupons and one of the recipients was an 18 year old teenage girl. Her father received the coupons and arguably was angered. He contacted Target’s management and rebuked them for sending the money saving offers to his daughter especially when she was not pregnant.

Although Target apologized, the story did not end there. After a couple of months it was the girl’s parents who were apologizing because they eventually found out that their daughter was in fact pregnant. They were more shocked by the fact that a retail store found out about their daughter’s pregnancy well before they did. So did Target have an Oracle goddess or just happened to make a successful guess? The answer is neither the Oracle nor the guesstimate but their statistical engineer Andrew Pole.

Andrew had identified 25 products that when purchased together indicated a higher probability of pregnancy. He applied this formula to Target’s data warehouse and compiled a list of customers who fell into the pregnancy category. The reason this example is important is because it shows that not only did Target figured out how to calculate their data but also what actions to take on it.

Yahoo’s Remote Work

In 2013, Marissa Mayer of Yahoo caused a firestorm by announcing a no working from home policy for Yahoo employees. The announcement caused a stir and critics jumped to the conclusion that she was too narrow-minded and comments both for and against the move filled the cyberspace. However, only a few stopped to think about how she made that decision.

From day one Mayer noticed that most of the Yahoo’s parking lots were empty and began to question the work from home program. Being a Google alumnus, she was groomed to “when in doubt, consult and trust the data”.

Mayer ordered the VPN logs and analyzed it to see if enough hours were being spent by the remotely working employees. Her calculations showed that not enough time was being spent and hence her decision to ban the program was made.

Yahoo is a golden example of Big Data being utilized for tough decisions.

Obama used Big Data in his favor

One of my favorite CTOs is Harper Reed. He was hired in 2012 to run the IT side of the Obama campaign. Harper made use of Big Data in the Obama Campaign and choreographed all of their moves based on the data he was receiving.

His team enabled voters to connect to the campaign page via Facebook and then used their data to give them specific points of the Obama spiel. For instance, if your Facebook profile mentioned you have interest in healthcare, the campaign would send you videos on how Obama intends to address healthcare.

Word of mouth marketing
There was a research study that came out suggesting that the number one influence on voters is the feedback they get from within their known circles. Harper used this study to his benefit by tapping into Facebook profiles looking for “best friends” and then asking users to send voting reminders to those contacts. This created a surge in free and effective advertisement for the campaign.

Targeting the right customer
The door to door marketing was also different under Harper. His team generated a specific list of households who matched a certain criteria. These lists were then sent to volunteers who in turn only targeted the listed families, instead of knocking on every single door in the neighborhood. This not only saved them lots of time and money but also created an additional voter base.

Facebook Data for Sochi Olympics

Data scientists at Facebook crunched some numbers based on who was talking about what kind of sport at Sochi Olympics. They then segmented this data and published a map outlining sports popularity by countries. The results were fascinating, for example the map showed that Europe loves the biathlon while figure skating is most popular in America and Russia.

Take Away

The common premise behind all of the cases presented here is that not only were these companies able to elicit trends from their data but also make sense of it. There is no benefit of Big Data if we don’t take proper actions from the analytical information.

One thought on “Predictive Analytics and Big Data in Real World

  1. You have certainly explained that warehouse management is the process of examining big data to uncover hidden patterns, unknown correlations and other useful information that can be used to make better decisions..The big data analytics is the major part to be understood regarding Hadoop Training in Chennai program. Via your quality content i get to know about that in deep.Thanks for sharing this here.

Leave a Reply

Your email address will not be published. Required fields are marked *