Advanced Analytics is a term often used, but what does it mean and what is the difference between big data, descriptive-, predictive- and prescriptive analysis?
You could be forgiven for feeling overwhelmed by all the buzz surrounding Advanced Analytics – but what does the term mean and how can you leverage it in the context of your supply chain? A common reaction is “this isn’t for us, it’s only companies like Google, Amazon, Netflix and Uber that can leverage from it.” This couldn’t be further from the truth. The sooner you realize that you already have data available, the better. Correctly used, data can give your company a strong competitive advantage. If they don’t already do so, it’s only a matter of time before your competitors will benefit from Advanced Analytics.
To begin with, let’s define three types of analytics: Descriptive, Predictive and Prescriptive. The latter of these two are often referred to as Advanced Analytics. Optilon can support you in all three types of analytics by using the most appropriate methodologies and applications.
Descriptive Analytics is all about visualizing historical data to support better decision making. A business user may have a hypothesis about a relationship in a given dataset and use bar charts, tree views and other means to substantiate it. This is generally referred to as Business Intelligence, although the word “intelligence” is perhaps not the best description of this method. An example of a typical question in this context is: “We believe that profit by product category has a strong correlation to geographical region.” Optilon offers several alternatives within descriptive analytics, all gathered in the area of supply chain visualization.
Predictive Analytics is the term used for analytics that predicts what is most likely to happen based on collected data. This is one of the most sophisticated types of advanced analytics and contains methods such as machine learning and deep learning.
Prescriptive Analytics, as the name implies, is about obtaining a “prescription” on how to solve a specific problem. Here are a few typical examples: “How should we source products during high season to keep up with demand and minimize total supply chain cost?” “What is the optimal safety stock level to guarantee service level objectives and minimize stock investment?” and “How can manufacturing orders on the bottleneck be sequenced to maximize efficiency and minimize set-up time?” Optimization, heuristics, and machine learning methods, can be applied to these types of issues depending on the industry and the specific challenge.
We haven’t mentioned Big Data at all so far. Is Big Data something completely different? The term Big Data is often somewhat misleading and takes thinking mainly into the area of capturing and storing large volumes of data. The Four Vs is one way of describing Big Data: Volume (scale of data), Velocity (analysis of data-flow), Variety (forms of data), and Veracity (uncertainty of data). Optilon will not guide you how to capture and store large amounts of data. We focus our efforts on the fifth V: Value, i.e. Data Monetization – gaining measurable financial value from stored and available data.
It’s not hard to understand how people easily confuse the concept of various analytics because Advanced Analytics includes so many different analytical methods. Hopefully this overview gives you some key insights to help you make sense of Advanced Analytics.