Johannesburg, 31 Jan 2013
Finding a competitive edge in enterprise data is evolving from data warehousing to big data analytics. While data grows every day, and gets backed up every day, much potential for insight is wasted. It's about applying new technologies to meet unfulfilled needs that (usually) can't be met by the traditional data warehouse architecture.
Mention 'big data' analytics and many company managers say there is no budget for it. Yet getting insight from big data is within reach these days, says Bernard Vertenten, Head of Innovation at solutions and systems integration provider, StorTech.
"Extracting value from traditional data warehouse sources, which consist of structured data internal to the organisation, is very different from using big data, which resides in different locations inside and outside the organisation, and needs to be processed until the value is derived. That said, however, traditional data warehouses and big data projects can work hand-in-hand, supporting each other."
Cost aside, getting a value from traditional data warehouses used to be a slow exercise - often needing months to obtain results from expert developers. If you were lucky, you didn't have to wait for years to get the value and ROI.
"The big change is that you can now extract what you're looking for, and derive real business value - driving down costs, adding revenue and delivering on the bottom line, thereby accelerating time to value," says Vertenten.
Look back to look forward
With big data analytics, you can start asking new types of questions of your organisation's historical data, and respond faster as a business, says Vertenten.
"As an example, say your organisation delivers consumer-level services, linked to loyalty programmes, some of which are bundled with tablet computers or smartphones. With big data analytics, you can start asking questions like, 'what are the demographics of the customers buying these services: more adults, or more teenagers?' And, 'how much data are the tablets generating?'
"From there you can adjust the loyalty programme; perhaps find a way to offer clients more free data or cheaper tablets. You can find out how many other services you have sold because of a tablet-driven campaign. You improve your understanding of the real return on a promotion. Because you get answers quickly, you can feed these insights back into new business decisions as you go along."
Unstructured consumer data
Big data projects, however, should be tailored to an industry or a particular company. Consumer data analysis will have very different objectives in the healthcare industry, for example, as opposed to in a financial institution, says Vertenten. And then there is the unstructured data peculiar to each company within a vertical.
"Though data doubles every two years, you keep it because of what is happening with tablets and smartphones; all these additional device interfaces pump a lot of unstructured data into organisations. Say a technical support person uses a mobile device to take a photo of equipment, and sends it to an expert to have a problem diagnosed. He has just introduced unstructured data into a structured data environment," he says.
"In a similar way, consumers responding on social media, Web sites or e-mail can introduce data that is not structured into your production environments, but that can be very useful for customer analysis."
StorTech's EMC Velocity Signature Solution Centre delivers the industry-leading big data analytics technology, Greenplum. This includes the Greenplum Massively Parallel Processing (MPP) Database, modular Data Computing Appliance (DCA) and Unified Analytics Platform (UAP), as well as analytics for SAS and Oracle.
A limitation on the data warehouse today is the lack of agility. Big data analytics provides faster answers and at far more value than before. Historical and current data, both structured and unstructured, from multiple sources, can now be queried without disrupting production environments, or breaking the bank.
Share