Once shopping moved online, however, the understanding of customers increased dramatically. Online retailers could track not only what customers bought, but also what else they looked at; how they navigated through the site; how much they were influenced by promotions, reviews, and page layouts; and similarities across individuals and groups. Before long, they developed algorithms to predict what books individual customers would like to read next—algorithms that performed better every time the customer responded to or ignored a recommendation."1
Describing the key difference between big data analytics and traditional versions of the practice, a report by the SAS Institute says that "The primary purpose behind traditional, 'small data' analytics was to support internal business decisions. What offers should be presented to a customer? Which customers are most likely to stop being customers soon? How much inventory should be held in the warehouse? How should we price our products?"2 But, as the report points out, traditional analytics requires structured data, whereas much of the information being produced online today is unstructured. Today’s big data technology, however, can make full use of both types of information. The ability to make use of unstructured data is particularly important for the vast amounts of "multi-channel" information now being produced. Enterprises today interact with customers in many more ways than in the past, including Web site visits and Internet chats as well as legacy methods like phone calls. And analytics can often be performed in real time, or a "stream." But real-time analytics is not yet fully mature and thus its applications remain somewhat limited.3
Broadly speaking, there are three types of analytics:
- Descriptive—Provides facts and information about past events without interpretation.
- Predictive—Forecasts future events and can be used with hypothetical scenarios.
- Prescriptive—Provides guidance about which actions to take.
Prescriptive uses of the technology are the newest applications; an understanding of how to fully use the concept is still being formed. Industries like finance and healthcare are leading the development of prescriptive analytics. Other industries are likely to adopt the concept more slowly, once clearer best practices emerge.4
Big Data Analytics in Practice
Big data appears to have grown from a specialized practice to a common, widely used idea. "Big data is no longer the hot buzzword it was a few years ago that people strained their brains to understand. It’s now entered the mainstream and can be viewed as an extension of traditional data crunching,"5 says Jonathan Vanian. Part of the transition to becoming mainstream is that many leading technology vendors offer some type of big data analytics product or service, including the following companies:
In addition, there are many newer, specialty companies that are having a significant impact on the market, including:6
A broad range of technologies are involved in big data analytics. Functionality that must be provided includes the following:
- Gathering data
- Comparing data
- Identifying patterns
- Storing and managing data
- Developing reports, visualizations, and dashboards
- Securing the data