Companies can test whether big data analytics systems can improve their business performance using the hardware and resources of hosting service providers, says business intelligence company PBT Group business intelligence consultant Willem Conradie.

Companies, and even small businesses, are investigating big data analytics systems to determine whether these systems will add value to their operations.

Therefore, companies can use hosting services to explore the relevance of big data analytics systems without buying the hardware and software, which are costly and time consuming to implement.

“If companies determine that big data analytics systems are useful, then they can start the process of implementing such a system and can justify the spending based on operational experience.”

Further, a hosted big data analytics system can be increased in scale by paying for more resources in the hosted environment, if companies want to increase the scale of a big data analytics experiment.

“However, if companies determine that the systems are not relevant and do not add value, they can stop the experiment and conclude the service without spending money on hardware,” Conradie explains.

Meanwhile, most problems that companies face are not unique and there are standardised big data analytics systems that will provide companies with a range of business algorithms that measure the relationships between various datasets to provide insight into business performance.

Some big data analytics systems can provide limited insight into transactions in real time, while more process-intensive systems will provide more complex insights, but are not fast enough to run on a real-time basis, says Conradie.

Both analytics systems are appropriate in several ways to combat various problems. The fast monitoring and analysing of text on social media pages and on the company’s help desk can be done to detect and resolve client queries rapidly or provide a deeper analysis of the company’s profitability.

Further, there also are organisations that make data available free of charge and companies that sell public data and data that has been stripped of individual detail. The data can be used to augment the in-house data of a company to add more external variables that the system can process to produce results to big data analytics queries from users, he notes.

However, a key part of implementing analytics systems is to take a procedural approach and have a firm life cycle in mind to ensure that the systems are adapted as more data is used during analysis and as the use of the systems in corporate processes changes.

“Big data analytics is an iterative process. With more processing capacity, the system can run through query cycles faster to produce results that will be sent to dashboards on managers’ desks to inform company strategy faster, which, in turn, enables the company to adapt quicker to changes in its environment,” says Conradie.

For many companies, such as telecommunications service providers, retailers, manufacturers and niche smaller businesses, the business landscape is changing rapidly. Analytics systems enable these companies to use historical and real-time data to forecast performance and generate different scenarios in which prices rise or fall, or consumer behaviour changes. These scenarios can be used to determine corporate strategy to mitigate risks and adapt business processes to changes in client demand, he concludes.