Predictive analytics are the future of big data Posted on : Oct 10 - 2015

This analysis is described as the future for enterprises looking to gain insights into business operations and find patterns between sales and marketing activity against revenue.

Many organisations have used it to good effect. Camden Council uses IBM big data analytics to create a database that consolidates residents' data to reduce fraud and costs, while Expedia consumes big data to better understand what its customers are buying.

Open source frameworks like Hadoop make the storage of data more cost effective and, with numerous analytics tools on offer, the promised big data future is here.

But it is set to change. Much of the analysis of large data sets is currently a process of looking at what is happening or has happened across an organisation.

This data is analysed into insightful information that highlights sales opportunities or problems in a supply or manufacturing chain.

This is often used to make an organisation more effective, but cloud computing, machine learning and in-memory technologies are creating the foundations for a big data future where looking forward is the objective.

Predictive analytics is set to be the next trend in big data. Rather than react to insights gained through data analysis, enterprises will use a combination of real-time, historical and third-party data to build forecasts of what will happen in their business months, weeks or even just hours in advance.

Such an approach allows action to be taken to avoid predicted problems, such as equipment failure or depleted stock, or to capitalise on opportunities to market products to customers, such as targeting people in euphoric or deflated moods after a sporting event.

Forrester analysts Rowan Curran and Mike Gualtieri believe that predictive analytics have never been more relevant and easier to use, and offer ways for forward-thinking enterprises to succeed in competitive sectors.

"Big data, gobs of compute power, and modern tools are making predictive models more efficient, accurate and accessible to enterprises," they wrote in a Forrester Wave research paper entitled Big Data Predictive Analytics Solutions, Q2 2015 (PDF).

"Why do it? Because enterprises that predict will win, retain and serve customers better than those that don't."

However, the analysts added the caveat that predictive analytics needs to be used in a way that prompts actions rather than just forecasts.

"Predictive analytics has limited value unless the exposed insights can be deployed directly into software applications and business processes," they wrote. View More