In 3yrs, the term ‘big data’ will disappear: Teradata CTO
By Gabey Goh October 24, 2013
- ‘Big data’ nothing new, just a marketing term created to enable people to think out of the box
- In future, industry discussions will shift toward integration of relational and non-relational data
MANY organisations are getting ready for big data, yet it is not the volume of data that they should be mindful of, but rather the diversity of data.
“I think many still don’t fully understand what the notion of big data really entails,” said Teradata Corporation chief technology officer Stephen Brobst (pic), noting that when most people think of ‘big data’ and its challenges, they think of the high volumes of data generated but not the diversity.
Big data refers to the vast amounts of ‘unstructured’ data that comes from the Web, including social media, as well as forms of data that cannot fit into traditional databases.
However, Brobst also acknowledged that tools for analysing data with high diversity are still in their early stages, pointing to the company’s acquisition of data analytics provider Aster Data in 2011.
“That’s why we bought Aster, as it was years ahead for doing this kind of big data analysis,” he said at the recent Teradata Partners Conference in Dallas, Texas.
Teradata recently released the sixth version of its Aster Discovery Platform, and claims to be the first to offer cross-analytic engine optimisation and data processing innovations.
Brobst believes that despite big data getting the lion’s share of the spotlight right now, in three years the term will disappear.
“Big data is actually nothing new, it’s a marketing term created to enable people to think out of the box. It was needed to make that transition to thinking about how to better manage and analyse it.
“The term is deceiving because it makes people think of volume; in the end it’s all data,” he said, adding that moving forward, industry discussions will shift toward the integration of relational and non-relational data to mine better intelligence.
“I also think sensor data will be the next big topic, which is far more interesting and far more valuable than social media data,” he added.
A pillar of Teradata’s current drive is centred on the concept of data-driven marketing, which is touted to enable deeper consumer insights and return on investment for marketing campaigns.
The biggest demand for big data analytics and the potential benefits it brings, according to Brobst, is coming from marketing departments, with many marketers embracing cloud-based solutions – especially in situations where the organisation’s IT department is still dealing with legacy issues.
“Marketing departments are starved for information. Things are different now versus before, when the focus on analytics was on transactions and value.
“Now they want to understand the behaviours of consumers and that kind of interaction data is not something you can get from ERP (enterprise resource planning) systems, which is where IT is used to getting their data from,” he added.
However, amidst the consistently highlighted need for marketing and IT to work closer together, some sticking points remain.
“Marketing wants big data but there’s still an unwillingness to increase the budgets of the CIO (chief information officer). The complaint from them to IT is that they don’t move fast enough and is also way too expensive.
“Marketing is taking more control over budgets and making internal IT compete with external vendors,” he added.
However, Brobst noted that in many cases, it’s unfair to say that internal IT “wasn’t fast or cheap enough” because only the cost of software licences and hardware is looked at, and not other factors such as the cost of talent to manage solutions, as well as the cost of datacentres and floor space, among others.
“So there needs to be education on both sides, because it’s not just about the speed but also about the cost,” he said.
Research firm Gartner has predicted that by 2016, 20% of CIOs in regulated industries will lose their jobs for failing to implement the discipline of information governance successfully.
When asked for his thoughts on how CIOs should approach data compliance, Brobst said that the notion of good governance around data means ensuring that there are no data dumping grounds, with all data auditable, easily retrievable and summarised with all details intact.
“Regulation is tricky because it’s typically about doing the least possible amount of work to comply so that you don’t go to jail – as most regulation requirements cost overhead,” he said.
Brobst advised CIOs operating in highly regulated industries to exploit the concept of ‘regulatory dividends’ in their mandate to adhere to data compliance.
“CIOs need to make sure that the same data needed for regulatory compliance can also be used for other stuff. Make sure the dollars spent on it can be repurposed for other uses, don’t just go checking off the box to keep the CEO out of jail,” he said.
Gabey Goh reports from the Teradata Partners Conference in Dallas, Texas, at the invitation of the company. See also: Teradata targets unified data, better analytics and marketers
Intense competition seals IT-marketing partnerships
Bob Chua’s RM60mil big data bet
Despite ‘overhype,’ market still upbeat on big data vendors
Banking on Digital: The confluence of IT and marketing
For more technology news and the latest updates, follow @dnewsasia on Twitter or Like us on Facebook.