(1) Data analytics and the big data movement both seeks to use data collected to gain competitive business advantages over a company’s competitors. However, three major differences floats out. First, dig data recognizes the fact that the volume of data from gathered from numerous devices like the internet or mobile devices of companies increases exponentially everyday. With big data, the ability to measure the amount of data will pave way to data better data management. Second, it recognizes that velocity of data created is as important or even more important than the amount of data gathered. Because of the speed of creating data, making predictions is enhanced thereby paving way for better decision making for the management. Lastly, the big data movement recognizes the fact that data can be gathered from a variety of resources like social networks and other mobile applications. Because of these companies can come up with better interventions that are appropriate to specific groups of clients rather than by just making an educated guest. (McAffee, Andrew and Erik Brynjolfsson, 2012)
(2) A petabyte is estimated to be about 20 million filling cabinets. Walmart’s 2.5 petabytes of data in a day would be equivalent to 50 million filing cabinets. (McAffee, Andrew and Erik Brynjolfsson, 2012)
(3) MIT Media Lab estimated Macy’s sales of Black Friday by collecting location data from the mobile phones. Through this data collected, they were able to predict their sales during the day. (McAffee, Andrew and Erik Brynjolfsson, 2012)
(4) Erik and Lynn Wu used data they collected from virtual-real time search to predict the changes in housing pricing the Metropolitan Areas of the United States rather than by waiting for historical data and basing the predictions from it which is very slow. (McAffee, Andrew and Erik Brynjolfsson, 2012)
(5) Researchers at the John Hopkins School of Medicine used data collected from Google Flu Trends to predict urges in flue-related emergency room visits one week earlier before warnings arrive at the Centers of Disease Control. (McAffee, Andrew and Erik Brynjolfsson, 2012)
(6) One evidence showing the capability of the utilization of big data to improve business performance is through analysis of existing data to know the gaps in a company and utilized these gaps found to improve service mechanisms that are aimed in improving business performance. An example of such is what a major company did. It performed an internal study utilizing the data they knew that there is at least 10 minutes gap and 5 minutes gap in 10% and 30% of their flights, respectively. Computing the idle time of the employees in these gap, they concluded that they are losing millions of money because of these. In their effort to come-up with better business performance that would not allow idle time for their employees, they turned to PASSUR to come up with a system them would accurately predict the exact ETA’s of airplanes.
(7) ETA stands for the expected time of arrival of an aircraft. (http://en.mimi.hu)
(8) Sears reduced the cycle time to generate personalized promotions by creating a mechanism to integrate the data collected by Sears Holdings in one data center rather than the usual separate data warehouse which requires a lot of analysis time since data formats are different, the data volume is big, and is highly fragmented. This adds to the difficulty of combining the data and generating required information to be able to generate timely and precise personalized promotions. (McAffee, Andrew and Erik Brynjolfsson, 2012)
(9) Hippo stands for highest-paid persons opinion.
(10) Executives who want to transform to big data management, can start to determine what the data says before making any decision. Knowing what the data says is not enough because it could also help to dig deeper on the source of the data, how was it analyze and does the result reflect the current trends. Likewise, executives could disregard what the data says and based their decisions on intuition because the data says otherwise. (McAffee, Andrew and Erik Brynjolfsson, 2012)
(11) The lack of effective talent management can hinder the success of big data implementation. Big data implementation is not all about the collection of data and the analysis process that makes it possible to interpret these data. There must be talent beneath these. This means that the company must have talented employees who knows what data to collect, where to get it and what to do with it. These particular skills are not taught inside the classrooms thus, employees with these talent must be kept. For example, a company may choose to hire an undergrad employee who has the right talents to determine and look for important data rather than hiring an employee with several masters degree but doesn’t know cannot tell the difference between data that can make innovations and simply useless data.
Another possible barrier to the successful implementation of big data is technology. Big data requires the use of various technology and software although some may be freely available. However, the use of these technologies comes with talented employees who knows how to maximize these technologies. A company may want to spend several thousand dollars for technological infrastructure, but it there are no employees that knows how to manipulate such then they become useless.
(12) The HSBC, a US based midland bank with branches all over the world offers credit cards to its patrons as one of each services aside from the regular bank services. Currently, it offers promotions applicable to all clients and have no loyalty programs. They partner with several businesses that offers payments on a staggered basis or on full payment basis.
With the coverage they have and the number of clients they serve, the company should start thinking of utilizing big data to have competitive business advantages. Particularly, data on a credit card owners personal purchases must be analyze to see what things the client usually buy with her credit card and at what time of the year or the month. The client’s regular purchases as well as its yearly purchases must be analyzed separately.
Those with common purchases at specific day of the week or the month must be grouped together and be provided with promotional offers while purchases while other promotional offers will be sent to the client before she usually makes them. For example, it based on the data collected, the client is interested in buying IT gadgets, then that particular client must be given a personalized brochures and loyalty programs. This way, the company’s promotional efforts will be more specific, which in turn will have more sales turnout. Likewise, to enhance partnerships with businesses, big data can also be utilized. For example, if based on the data collected and analyzed, it sees that most of their card holders in a specific metropolitan location visits a salon once a week then this could be used by HSBC to create partnerships with salons located in the location. With this, then HSBC service will be available in almost all salons so wherever these group of clients go, they have the option to use their cards.
(13) The three conventional assumptions presented by the speaker in the video Hadoop and Big Data are: 1) Machines can be reliable 2) Machines have identities (3) Data set can fit on a single machine.
(14) Regular enterprises who currently uses Hadoop today includes Autodek, Fox Interactive Media, Carrier and Mailtrust
(15) The most common sources of data are simulations and scientific experiments, existing databases, user data (unstructured) and systems generated data.
(16) Hadoop doesn’t serve data in real-time because it absorbs data first before processing them. Hadoop is not actually a competing product of database software because it does give you a choice depending on your needs. It doesn’t let you read or write to your database because it is only capable of appending files. Likewise, it is considered as just a utility or commodity software that is available for free. However, unlike databases, it is capable of handling a thousand more data and supports both structured and unstructured data.
References:
http://en.mimi.hu (2013). ETA Page. Retrieved from http://en.mimi.hu/aviation/eta.html
Jorgwel (2011 August 28). Hadoop and Big Data 1/6 Challenging Old Assumptions [Video file]. Retrieved from http://www.youtube.com/watch?v=y8DRKd4SKWo
Jorgwel (2011 August 28). Hadoop and Big Data 2/6 Processing Petabytes [Video file]. Retrieved from http://www.youtube.com/watch?v=xQOKOl6lKJM&feature=relmfu
Jorgwel (2011 August 28). Hadoop and Big Data 5/6 Ferrari vs Freight Train [Video file]. Retrieved from http://www.youtube.com/watch?v=-QdCABPyu1k&feature=relmfu
McAffee, Andrew and Brynjolfsson, Erik (2012). Big Data: the Management Revolution. Retrieved from http://hbr.org/2012/10/big-data-the-management-revolution/ar/1