The largest explosion of data is happening outside the borders of traditional corporate data - on the Internet itself.
Andy Hoiles, new workload leader for the Middle East and Africa at IBM Middle East, cites the rise of social media, which sees individuals posting data that could be relevant to all types of businesses on the Internet, as an example of this trend. "This data could inform them about the level of satisfaction with their products or services, their competitors' products or services, general trends or opinions across their industry or specialised area," he says.
Traditionally, IT infrastructures in many companies have grown organically, notes Hoiles, which means the data is sitting in silos on disparate systems. This means something as simple as changing a customer's home address could involve a complex series of steps, and with the addition of unstructured data, this is a potential recipe for disaster, he continues.
It may surprise some to hear that 80% of the world's corporate data resides or originates on mainframes, notes Hoiles. Companies that invested in mainframe technologies to consolidate their core business applications have their corporate data already centralised in a common format, making it less of a challenge to combine this with the relevant unstructured data from the Internet. "On this basis, it is possible to access, combine and manage a relevant mix of data centrally, with relative ease, and with the highest levels of data security and governance."
According to Hoiles, if production data has been centralised on a mainframe, businesses are able to run analytics on the most current data. Real-time analytics deals with the data as it stands, at the point when the analytics are run, and provides an up-to-date picture at the point of transaction.
He outlines an example of a customer placing a call to an agent to make a telephone banking payment. As this call and the subsequent transaction are taking place, analytics can be run to see if there are other products and services that may be of value to the customer, or if there are alternative options that could be favourable on the current transaction.
"This simple example demonstrates the power of real-time analytics using centralised production data to drive greater business opportunity and customer satisfaction."
Weather and Walmart
In the UK, the national weather service, the Met Office, deals with 10 million weather observations every day. According to Hoiles, accurate weather forecasts depend on complex predictive models that access huge amounts of historical data. The more data they have, the higher their levels of accuracy, because huge historical data sets are trawled to identify the most likely outcome based on the current weather patterns and what has happened in similar situations in the past, he says.
"This is big data and business analytics in action. Without an infrastructure that can support the huge volume of data and analyse it in real-time, they would not be able to produce such accuracy, 24/7, every day of the year."
Mainframe technology handles all of this data in a secure and resilient manner, says Hoiles, adding this platform has allowed them to set up vital services and provide weather warnings well in advance. For example, the Red Cross relies on the Met Office's flood alert warnings to prepare its support for those who could be in danger.
Similarly, Walmart operates in 25 countries and has over 11 000 retail outlets that process around 400 million transactions a day, 150 million of which are from online retail. "The simple logistics of delivering the right products to the right stores at the right time could provide significant challenges," says Hoiles, noting that Walmart tackles this by recording every single purchase made and storing this information on the mainframe.
It also offers self-service facilities, which are delivered from its centralised cloud services so that each outlet can manage their business in the most efficient manner, gaining access to all the information they need to keep their customers happy, he adds.
"There is no doubt that unlocking the potential associated with big data needs some careful analysis and planning, which includes building a vision of what the organisation would like to achieve at the end of this process."
Share