Want to see the future? This developing tech can help. And as it becomes more valuable to business, foretelling its applications will be key
Trite as it is to keep repeating, we live in a world where the only certainty is uncertainty – and executives would pay a lot of money for a working crystal ball.
That’s what makes predictive analytics – data mining that can make predictions about future events – so compelling, and the market for such tools so large. According to Zion Market Research, the global predictive analytics market hit $3.49bn in 2016 and is expected to balloon to $10.95bn by 2022.
Analytics is becoming increasingly embedded in certain sectors. Retailers, for example, use it to determine what to stock and how much, as well as to predict consumer behaviour. Office supplies retailer Staples saw a 137% return on investment when it bought into analytics.
Meanwhile, financial services organisations are using such tools to predict fraud, while health insurance firms can identify those most at risk of chronic diseases and plan the interventions most likely to work. Manufacturers can predict machine failures, allowing prevention rather than cure and reducing downtime.
How it works
In a nutshell, predictive analytics uses past data to predict future events. In a way, it’s what most of us do every day, using our past experiences to inform our next decisions, but with added rigour and science. It combines techniques from a number of areas, including data mining, statistics, text analytics, machine learning and artificial intelligence to produce models that can interpret big data to predict likely future outcomes.
According to analytics expert Thomas H Davenport, predictive analytics “isn’t magic”, and is achieved using three main building blocks: data (having a single source of truth), statistics (usually ‘regression models’, which predict a number, or ‘classification models’, which predict outcome), and assumptions (all predictive models are underpinned by assumptions, which need to be up to date).
The growth of the analytics industry and increasing levels of predictive capability present an exciting opportunity for procurement. EY’s paper, Empowered by Analytics: Procurement in 2025, predicts that by 2025, decision-making in procurement will be “significantly influenced by high quality data analytics”.
Robert Handfield of the University of North Carolina’s Supply Chain Resource Cooperative says predictive analytics is relatively new in procurement, but is “increasingly important, particularly in organisations that have already been through the cycle of spend analytics, supplier leveraging, segmentation, and consolidation.” He identifies the ability to forecast revenue, mitigate disruption and identify market opportunities as key uses.
Risk management is an area ripe for transformation. EY’s report predicts procurement functions in large corporates will soon have teams dedicated to quantitative risk management. Analytics will be able to track and predict third party risk, such as that of a cyber security breach, assessing vulnerabilities across a range of areas and enabling better decision-making.
However, with Deloitte’s latest global CPO survey finding just 3% of organisations have ‘fully deployed’ predictive analytics technologies, there’s a way to go before the breathless claims of technology suppliers become a reality.
Predictive to prescriptive
The next stage of the analytics journey, beyond predictive analytics, is prescriptive analytics. This even more advanced sector of analytics moves beyond telling you what is likely to happen to suggesting the best course of action to a situation, as well as potential implications. It will enable organisations to interrogate data and content to answer questions such as ‘what should be done’ or ‘what can we do to make X happen’, says Gartner.