What is Analytics?

Businesses today must be able to glean insight from data to improve their competitiveness and business performance. Analytics helps businesses to discover and interpret patterns in data across the business lifecycle and create actionable analytics that aim to improve the health of the business.

Today, a connected car can provides vital analytics on driver behavior (speeding, breaking, parking), car performance (mileage, speed), need for maintenance(oil change, tyre pressure, equipment failure) and even dial out emergency services in case of untoward incidents. Such is the power of data and analytics.

Data is the cornerstone of analytics. Historically data was processed in batch mode. Data was extracted from the source systems, staged, cleansed, conformed, transformed and loaded to a Datawarehouse for analytics. Businesses had to wait for a couple of days or more to obtain actionable analytics from data. Attempts to lower the latency usually meant investment in hardware and software infrastructure which wasn’t cost effective at that point in time. In parallel, requests from Lines of Business to reduce latency or improve turnaround time on change requests started to pose serious questions on IT’s role as a business enabler.

Rapid adoption of Cloud based infrastructure helped business procure inexpensive compute capacity to solving the latency problem but the explosion of digital technologies like mobile, IoT, API’s meant that challenge posed by the 4V’s of data namely Volume, Velocity, Variety, Veracity need to be dealt with. Today, Real time analytics is made possible due to Big Data processing and storage technologies based on the Hadoop eco-system.

Broadly, analytics can be classified as being Descriptive, Predictive and Prescriptive.

Descriptive analytics: Provides insight into the past. Uses reporting or visualisation tools.

E.g.: What was the trend in purchase of mobile phones by brand over the last 18 months in South-East Asia?

Descriptive analytics: Predict a future outcome based on past history. Uses Statistical tools

E.g.: What will be likely uptake of designer furniture during the Christmas season given that we have the past 7 year’s data?

Descriptive analytics: Advise on possible outcomes.

E.g.: Production and inventory optimization.

Advanced analytics typically covers Artificial Intelligence topics such as Machine Learning and Deep learning using Neural Networks.