Ask a few people about what big data stands for, and you are likely to get as many definitions out of them.
“Like any relatively new topic, the industry definition hasn’t really settled down yet to a clear consensus. Some view it as a problem; most view it as an opportunity. The core idea is relatively simple: new ways to capitalise on the enormous amount of information we’re collectively generating. The “big” part of the term refers to the difficulty involved: new approaches are required to address these newer opportunities,” says Chuck Hollis, VP CTO for global marketing at EMC.
What they all do agree on is the fact that big data is here already in most organisations, and it is time that enterprises began to take an active interest in understanding and getting the most out of it.
“Knowledge is where the value is being created in business today. Big data is of strategic importance to organisations and usually the second most valuable asset after their people. The sooner an organisation realises its value, the better they are positioned to take advantage of today’s data driven business world,” says Marwan Mousfi, VP of business solutions at Technology Partners.
“Ninety percent of the data in the world today was created within the past two years and therefore big data can be defined in a number of ways, starting from datasets that become so large over time that they are difficult to work with and use to extremely large files. This includes being able to create, store, search, share, analyse, visualise, manage, report on or even gain meaningful intelligence on. By working with larger and larger datasets, organisations can turn a meaningful event to a revenue generating opportunity. The trend of big data will not stop as the cost of capturing, processing, storing and sharing data keeps falling, mobile devices, electronic transactions and social media consistently increase,” says Marco Gerazounis, senior VP for Middle East, North Africa and Turkey at Software AG.
“Enterprises must pay attention to big data due to the simple reason that if they don’t, their competition will, and the “first mover” advantage will give them that significant competitive edge. I say this because we see a lot of interest across industries and geographies where enterprises want to create innovative new business models using big data, gain market and customer insights, or bring in business process optimization and operational efficiencies they were unable to bring before,” adds N Veeraraghavan, senior VP and global head of enterprise information management and analytics at Cognizant.
Steve Bailey, regional operations director at Commvault says, “Big data is something that can creep-up on you and before you realise it, it can be causing huge expense and present new types of risk. What often starts as a project can become business critical and the bigger it is, the harder it is to keep it available or recover in an emergency.”
Harnesssing big data
Vendors believe that organisations across the globe need to take the rising importance of big data more seriously and should work towards encompassing it in organisational functioning for long-term benefits. However, most Middle East organisations continue to struggle even with the most basic of data management practices.
“Most companies are currently using traditional tools to store data, such as NAS and SAN technology. This leads to islands of data which is difficult to bridge without a big data solution. Businesses that want to gain insights into their customer and partner sentiment need to not only focus on their internal sources of data but throw a wider net and start investigating external sources of data. Data from social media sites, macro-economic analyst, supplier stock availability, and customer buying habits can help companies gain invaluable insights and fine tune their product development to better address changing market needs,” says Basil Ayass, enterprise product manager at Dell Middle East’s Commercial Business.
David Rajan, director of technology at Oracle says, “Big data is the other half of the data coin. The dense structured data we’ve been managing for the last 20-30 years has now been enhanced by the other half of that coin which allows business to complete a 360 degree view that until recently was outside of both our comprehension and grasp. Seeing both sides of the ‘data coin’ allows both the hindsight of the transactional relationships and events that we have captured but also the new capability to enhance our perspective by analysing the unstructured activities, events and relationships that provide insight into future events, wants, aspirations, desires and emerging trends.”
“Data management in the Middle East market is currently maturing from being operational to being analytical. IT predominantly takes decisions on data management to fulfill immediate business needs to answer the question of “what happened” rather than “what will happen”. Obviously, we see a gap here as compared to geographies such as the US and Europe. Big data management provides the opportunity to bridge this gap. ME enterprises can now move up the value chain and position business needs, business analytics, and customer experience as the driver for data management,” says Veeraraghavan.
According to EMC’s Hollis, big data are solutions in the region finding their way among certain sectors, including telecoms, finance, public sector and energy production. The second wave though, which would cover retail, manufacturing or pure information-based business models, is still in the coming.
He adds, “We see one group of organisations that are going down a path of “analytically enabling” their workforce – making analytical tools and data sources available in a self-service easy-to-consume environment. We also see a smaller (and more aggressive) set of organisations investing heavily in the new breed of data scientists, and build very large and purpose-built environments around their unique needs. We also see a few more progressive IT organisations investing the time to inventory and catalogue their information assets, and aggressively market these assets back to the business in an effort to encourage increased consumption and utilisation in the organisation.”
The current situation being as it is, vendors state that enterprises in the region can start moving towards collating and obtaining intelligence out of big data by implementing simple solutions and processes as a starting step.
“At best, very large organisations use data warehouses. However, very few have an information governance framework that encourages the establishment of a structure for the creation, manipulation, storage, use, re-use and retirement of data. Most of them either don’t sufficiently understand the strategic implications, can’t find a way forward or don’t have the organisational willingness. A good start is to have an information governance framework that will help enable a data landscape based on content rather than data silos. Then comes the necessity to deploy enough useful big data capabilities into the hands of the largest number of employees,” says Mousfi.
Bailey states, “Using storage resource management (SRM) software is a good start. Agent-less SRM makes rollout a non-issue and it can really help you to organise data for the best access/cost compromise. Not only will the trending warn you that you’re heading down the Big data path it can help drive archive policies for effective long term storage and keep costs under control. Big data needs expert handling and getting the right help early can make a big difference, so while it’s not easy to spot a problem before it becomes acute doing so can save you a lot of pain and expense. The other point to make is that choosing the right technology is critical – and remember, getting it wrong can be catastrophic.”
Oracle’s Rajan points out there is a lifecycle to big data, which includes acquiring it, organising it and analysing it in order to make decisions, and organisations have to create a framework for every element of the same. Apart from that, he emphasises the need for putting in place ‘behavioural changes’ that can prepare and familiarise an organisation to take full advantage of big data.
“They have to be collaborative. Big data is very much a joint activity between the business and IT. It’s important to ensure that the team encompasses both elements. Be focused on value – find the use case for your organisation. Once you have that in place, be experimental. And finally, be strategic – consider how big data fits into your existing data and IT strategies,” says Rajan.
Many believe that these same behavioural changes, and the change management that it entails, is one of the biggest challenges facing Middle East enterprises in these early days of adopting and working with big data.
“Change management is the single biggest challenge that enterprises face in a big data initiative. A retail client of ours has created a big data crowd sourcing team with business and IT leaders, data analysts, and academicians to help foster a data-driven mindset within their organisation and bring about this change in an incremental manner rather than letting it be disruptive,” states Cognizant’s Veeraraghavan.
There are other challenges that enterprises are likely to encounter as they start out with their big data investments and changes. Management buy-in and the willingness to change their legacy business processes are key success criteria, says Ayass.
“The inherent challenge with recommendations for big data in the enterprise is that there is not a lot of published research on best practices for deploying and running a Big data solution in an enterprise environment. We therefore recommend that customer follow a time-proven systematic approach to implementation a success IT project. Dell leverages the WADI concept of Workshop, Assess, Design, and then Implement to tackle such complex projects and ensure the key success criteria are met,” adds Ayass.
Commvault’s Bailey points out, “Protection and disaster recovery can be staggeringly expensive. To mirror 1TB these days is relatively inexpensive, to mirror 500TB on high-end disk for scientific analysis is cost prohibitive for all but those with very big purses. Even backup to relatively cheap media like tape can be expensive due to the sheer volume. The other issue is time – while everything is faster today than it used to be, moving even modest amounts of terabytes around can take a while, let alone petabytes.”
“On a practical level, there could be a serious shortage of talent that is necessary for organisations to take advantage of big data. Also, several issues will have to be addressed to capture the full potential of big data such as policies related to privacy, security, intellectual property, and even liability,” says Mousfi.
Hype to productivity
In 2011 Gartner added big data to its technology hype cycle. Gartner uses hype cycle to assess where technologies lie through an entire lifecycle beginning from a technology trigger, to a peak of inflated expectations and then on to a plateau of productivity. According to the hype cycle for emerging technologies in 2011, big data is past the technology trigger and is on the rise to the peak of inflated expectations. This is also a time that is characterised by first generation products, high prices and lots of necessary customization, with early adopters investigating the technology. What that means in real terms is that we are likely to hear a lot more of big data in 2012.
“Five years ago, it was virtualisation. Two years ago, cloud became the new big thing. I predict that 2012 will be the year for Big data in the Middle East,” says Ayass.
“In the Middle East and across the globe, we see industry segments such as manufacturing, ecommerce, high tech, energy and utilities, banks, transportation and public administration showing a lot of interest in big data solutions. Early indications are that a good number of clients are willing to take a proof-of-concept approach to understanding the value of big data solutions,” says Veeraraghavan.
Being vendors who sell solutions, they remain positive that adoption is set to increase in 2012. As early adopters move to big data solutions, this is likely to happen, since big data itself is growing from a relatively small number. Hollis explains the model for enterprise adoption of big data thus.
“I use a three part model to describe enterprise adoption. The first group is already quite proficient at big data. They know what they need and want from technology vendors such as EMC, and we simply have to give them the tools to do more with their resources. But that isn’t a large group, especially in the Middle East.
“The second group really doesn’t have an appreciation for advanced analytics, or rich content models, or large-scale social, or any of the hallmarks of a big data-oriented enterprise. They’re just not interested in those topics, very often with very valid reasons,” Hollis says.
He adds, “The third group is much more interesting – they have some of the pieces, but they could be doing a lot more with what they already have. More importantly, we meet many forward-looking leaders who are considering taking the next step. These leaders want to learn – what are other people doing, what results are being seen, how does one go about organising for success, how to get an initiative off the ground, and so on. We’re doing what we can with these people to bring our experiences and insights to them to help them along their over journey.”
Enterprises should remember that these big data trends are not unique to the Middle East, and that they are in rather good company when they stay cautious about these technologies. The entire world is getting used to the idea of big data, and adoption is prone to be similar across continents for the new set of technologies.
As Hollis points out, “This isn’t just true in the Middle East – it’s largely true around the globe. Because when it comes to making new ideas real, it really is a global economy.”
Big Data – the definitions we heard
Ninety percent of the data in the world today was created within the past two years and therefore big data can be defined in a number of ways, starting from datasets that become so large over time that they are then difficult to work with and use to extremely large files. This includes being able to create, store, search, share, analyse, visualise, manage, report on or even gain meaningful intelligence on. By working with larger and larger datasets, organisations can turn a meaningful event to a revenue generating opportunity.
— Marco Gerazounis, senior VP for Middle East, North Africa and Turkey at Software AG
It is the accumulated information in a large company and though often not always appreciated, it is a companies top line asset for addressing business problems and implementing new business models.
— Marwan Mousfi, VP, Business Solutions, Technology Partners
It is the ability to analyze extremely large data sets, using processes, applications, and storage, and turn it into intelligent data that businesses can use to achieve better outcomes, make better decisions, and positively impact the bottom line.
— Basil Ayass, enterprise product manager at Dell Middle East’s Commercial Business
The core idea is relatively simple: new ways to capitalize on the enormous amount of information we’re collectively generating. The “big” part of the term refers to the difficulty involved: new approaches are required to address these newer opportunities.
— Chuck Hollis, VP CTO for global marketing at EMC
I would define big data as the science and “art” of managing data that is too large and complex to analyze with traditional data management tools and techniques, and doing so in the shortest possible time. The scope of big data management is much broader and deeper, expanding within the enterprise and extending outside of it. It can take various forms and shapes ranging from document files, storage devices, collaboration forums, log files, emissions, sensors, mobile devices, events, message streams, images, videos, and so on.
— N Veeraraghavan, senior VP and global head of enterprise information management and analytics at Cognizant
Big data is really about datasets that have become too large to be accessed or managed using traditional methods. They mostly occur in areas like life sciences, physics, geology, business informatics, trending and billing.
— Steve Bailey, regional operations director at Commvault
Today, we have greater volumes of unstructured data as defined above, such as web clicks, data that is not as dense as a sales transaction. This new data can now be captured, managed and analysed in an approach more appropriate to the business value / GB of that data. That is big data.
— David Rajan, director of technology at Oracle