Analytics is one of the most disruptive and discussed technologies of today. Much has been written about the benefit of analytics, the complexity, its adoption etc. For the not so initiated, here is an attempt to encapsulate the different facets of Analytics.
What is Analytics?
Data Analytics is the science of examining data collected from various sources and representing it meaningfully to get insights, which support key business decisions. It entails identifying and segregating data pertaining to a set of objectives, and presenting them in a visually attractive manner to uncover patterns. Visualization techniques involve graphical representation in the form of graphs, trend charts etc. and can leverage statistical modelling techniques to gain key insights.
Most traditional systems have a reporting facility that help put a finger on WHAT happened or what one would call descriptive analytics. But exploring the WHY or HOW of a particular phenomenon involves analysing several aspects including historical trends. This forms the core of analytics and involves examining the data from multiple perspectives.
Where is Analytics used?
There is hardly an area of business that cannot benefit from analytics. At a high level, analytics can be applied to functions within the organization, to external facing websites or data that resides completely external to your organization such as social media. Web analytics, for instance, looks at the efficacy of the Web site by analysing the growth or decline of number of visitors visiting the web site, the web pages they visit and so on. Social Media analytics leverages data from social media sites such as Twitter, Pinterest, Google, or LinkedIn, to mine customer sentiment for new product design, improving customer service and so on. Organizational analytics, in itself, can vary depending upon the function within the organization it is applied to. It could include customer analytics, operational analytics, spend analytics, financial analytics, and enterprise performance among others. Operational analytics, in turn, could comprise of analytics related to inventory, equipment utilization, or logistics. The application of analytics is also industry specific. For instance, Customer Analytics, in the healthcare industry, analyses patient data related to diagnostics, medical claims, pharmacy usage etc. to better predict patient health, diagnose ailments and recommend better treatment. Whereas Customer Analytics in banking can help banks enhance customer acquisition, retain existing customers and cross-sell to increase revenue per customer. And, in the traditional manufacturing and services industries, it could help identify customer preferences, generate repeat business, increase customer loyalty and grow the revenues.
Big data analytics is yet another area of analytics. Data that resides within most applications is structured. At the same time, emails and other collaboration tools have free flowing data but may have classification headers thereby making it semi-structured. The advent of social media has led to exchanging data unstructured formats such as audio and video. Handling such variety of structured and unstructured data in huge volumes getting generated at high velocity, is at the core of Big Data analytics. Because the data could be unstructured, it has given rise to text analytics, sentiment analytics audio analytics, video analytics, and other similar areas.
Nowadays, yet another disruptive technology has emerged, viz. Internet of Things (IOT). IOT involves use of sensors and devices to collect data from vehicles, manufacturing equipment, appliances, electricity grids etc. Analysis of such data to understand usage patterns, failure patterns etc. is IOT Analytics.
Types of Analytics
There are different kinds of analytics, driven by organizational objectives. Investigative analytics, as the term suggests, helps in performing root cause analysis of business situations. In an information economy, one cannot stress enough on the importance of getting to the bottom of problems as quickly as possible, upon which success is so dependent. Predictive analytics, on the other hand, uses multi-year historical data to study a trend and forecast a future possibility. Prescriptive analytics, meanwhile, uses root cause analysis, digs into a repository of past incidents and the solutions identified, to recommend options to address a situation.
ERP systems have established themselves as a tool to capture day to day transactions within an organization, and are commonly referred to as an Online Transaction Processing System (OLTP). These systems lend themselves nicely to answering operational questions related to certain parameters. Examples include inventory status, accounts receivables, account payables, bank balances etc. However, when it comes to analysing causes or studying usage patterns over multiple years, the OLTP system falls short. Hence, the advent of Online Analytical Processing Systems (OLAP), which extract key data from the OLTP systems and reorganize it in a manner suitable for analytics. The tools used to extract data from the OLTP system or the source systems are called ETL tools. The system where the extracted data is stored and further processed is called a data warehouse. Query tools filter and extract key and relevant data from the data warehouse and present it to the visualization tools for representation. Visualization tools present this data in graphical and visually attractive formats to enable smarter decisions faster. Visualization tools have evolved to permit What-if analyses by allowing users to change certain parameters using sliders or dial gauges and getting the corresponding data reorganized on the fly.
As analytics provides strategic insights to drive key business decisions, it is a critical component of any organization’s strategy. It reduces the shoot-from-the-hip approach of CXOs and helps take rational and fact-based decisions to shape the future of the organization.
The author is associate vice president and ERP Head, Nihilent Technologies.
Here is the link to the article published: http://www.techgig.com/skill/dataanalytics/blog/Data-Analytics-A-Primer/6709957