[MUSIC] Hello, my name is Nicolas Glady. I'm professor at the Essec Business School and you may have seen me in the MOOC, Foundations of Strategic Business Analytics. Business analytics applications have been around for many, many years. As you have seen during the other modules, finding groups within data or doing regressions are [INAUDIBLE], that are well recommended. And from CRM to operation management, data has been used for improving the performance of many businesses since at least the 90s. So is big data really new? Is the revolution some people claim it is? I try to explain to you that there is something fundamentally new in the phenomena we are observe in this video. And that it's, therefore indeed, a revolution. Let's take a very concrete example with CRM, customer relationship management, in the context of a telecom service provider. Like AT&T, for instance. And let's imagine that you're interested in four individuals. The presidential couple, President Obama and the First Lady, and two others. Let's imagine Felix and his [INAUDIBLE]. These individuals use their phone and hence make calls or transactions. With a telecom company, for a long time, these companies have been collecting data about their customers and their transactions. And so, one of the most classic ways to apportion CRM would be then to collect a data set with three volumes. Recency, frequency, and monetary. One, recency means, how recent was the last transaction? [INAUDIBLE], two frequency means, how often the individual makes a transaction. And three monetary means, what the average transaction value is in dollars, for instance. It is the very well known RFM data stracture that will allow marketers to perform RFM segmentation. And then conduct some specific targeted market actions like course selling or customer contention for instance. Moreover, we can see that the data is created and collected and centralized by the company in a well structured way. In the years 2000, two phenomena appeared. What was then called the web 2.0 and social networks. With web 2.0, anyone could generate content and share it on the web. The data generation process was decentralized. It can be pictures, for instance, like the Four More Years picture of the presidential couple that was taken after the reelection. Or it can be YouTube videos from famous YouTubers like PewDiePie. Even your auntie or your grandmother can have a cooking recipe blog nowadays. And this explains the variety of the very famous 3Vs of big data. because the source of data can now be anything and in any format. This is explained by the fact that the data is user generated, and not only company generated anymore. This is typically unstructured data like pictures, videos, or text because it doesn't, for instance, have a predefined format or data model like an RFM database could. After web 2.0, the second phenomenon is the central role of online social networks in these dynamics. A picture, video, or blog content could be exchanged over networks like Facebook, Twitter, or Tumblr for instance. And this explains the velocity, the information is exchanged faster than before, and the volume of the 3Vs of big data. The volume may even be explained mathematically. If you're interested about storing information about an individual, the quantity of data stored will be proportional to n. But now if you're interested in the interactions of those individuals because you know it's better to target customers like PewDiePie for instance, you will generate data about the connections and not only the individuals anymore. And if you have n notes in a network, the number of possible connections is n times n minus one. So it's similar to n to the square. It's a quadratic relationship between the number of individuals in a network and the data generated about their exchanges. This [INAUDIBLE] calculation explains in part the volume of the 3Vsf. Before we had data generated by companies that was structured, and we could have thought to produce reports only once in a while. Like every week, a month, about our customers. Now, days of evolution happens. The data is generated by the users themselves, and in real time. And can be anything. Text, pictures, or videos. Strictly speaking, this is indeed a revolution. UGC, user generated content, places the individuals at the core of the data generating posts. Hence at the center of a business analytics [INAUDIBLE]. Now, that doesn't mean that all of the applications you encounter will be dealing with [INAUDIBLE]. As a matter of fact, in the last five years, most business applications are still using classic types of structured data. But this shift towards more customer centricity however, is a trend, that is more and more present. And when you're dealing with strategic business analytics, using user generated data or not, you should always place the individuals, customers, consumers, or even citizens are a the center of your analysis.