Will big data be the big issue this year?
The term big data has dominated IT conversations for the past few years. Is it a new data opportunity or a new problem, or is this simply a case of age-old issues being rebranded and hyped up to make more money?
by Hans van de Groenendaal, EE Publishers
Neil Cosser, identity and data protection manager for Africa, Gemalto believes that data is the new oil, precisely because its value is growing phenomenally but remains relatively unexploited – and that data needs to be refined for value to be extracted.
“Many organisations are sitting on information assets of huge worth – just because they are not making the most of them doesn’t change that fact. Big data brings tremendous business potential but it also comes with challenges. This unstructured data which arrives in huge volumes have the potential to drive deep insights. On one side, it is clearly a new opportunity for companies which are now enabled to correlate and analyse massive data sets from disparate parts of the organisation and the business environment to determine trends, SWOT information, directions, strategic advantages, etc. Data about customers and processes is intrinsic to what makes a business tick and such data has far broader value particularly if it is analysed as a whole. On the other side, big data brings new challenges with large volumes of data being stored. These volumes of data require both management and security which have business implications.”
Ian Jansen van Rensburg, senior systems engineering manager, VMware, Southern Africa: “Technology continues to evolve at a rapid rate and is simultaneously increasing in intelligence. If you have a look at smart devices and what you can accomplish with just a single device you’ll see that with technological enhancements to smart devices comes a major increase in the amount of data generated. User expectations are changing as users want faster and smarter access to the data. Access to data has always been a part of IT but with the developments in new smart devices, smart access to larger amounts of data is necessary.”
Gustav Piater, marketing and sales director, AIGS does not believe that big data is anything new. “Companies have been gathering lots of data for a long time and this growth is part of the natural evolution. I believe that the biggest changes are the sources where these datasets are now being generated from and the frequency of updates. I believe that the industry has used this as a hype to upsell and create opportunities.”
Deon van Aardt, regional director Wonderware Southern Africa: “The world of industrial automation generates more data faster than any other. It’s all happening in real time down to the millisecond. To compound the problem, an increasing number of plant devices are getting smarter and providing more data while data-intensive graphics are now far more common than ever.”
In the South African context are companies ready to work with big data? In your opinion what does a company have to change in the way it operates to derive real benefits from big data?
Gavin Holme, head of business data and Rudraksh Bhawalkar, analytics and information management, Wipro: “Yes, the South African market has one of the most mature and growing customer bases in the world. The big financial, retail and consumer companies as well as healthcare institutions already have multi-channels to interact with their customers over platforms such as the internet, mobile apps, LinkedIn, Facebook etc., which is generating a huge amount of semi and unstructured data. The challenge is to integrate and leverage big data in conjunction with structured data generated from CRM, ERP, billing and sales applications.”
Gary Allemann, MD, Master Data Management: “Just as relational data bases revolutionised the way we stored and queried data approximately twenty odd years ago, big data requires companies to explore new and varied approaches and technologies. Companies need to recognise that old business intelligence (BI) approaches and technologies, such as structured query language (SQL), have a very limited role to play in exploiting big data. We have to recognise that, to extract value, we need to invest in skills and in new technologies that are designed to deliver analytics, not reporting.”
Warren Olivier, regional manager for Southern Africa, Veeam: “In South Africa, our data is exploding just as much as everybody else’s. We’re uploading more data, we’re storing repetitive copies of it and our bandwidth needs to increase. The majority of our time is spent managing copies of big data, not the actual data itself. Companies have to start using technology smarter and invest in tools that reduce their storage footprint, churn through data quicker and analyse it more effectively. Technology, such as deduplication is there to help them reduce their storage footprint and do more with what they have.”
Kholiwe Makhohliso, country MD and technology leader, Oracle South Africa: “Big data, by and large, will mandate inclusion of new types of datasets into companies’ analytics landscapes. This is usually where the challenge starts. Current analytics platforms are not geared to handle these new data types. These platforms are also not scalable and agile enough to handle real-time analytics which is one of the components of a robust big data platform.”
Neil Cosser: “As a gross generalisation, South African businesses are not ready yet to work with big data. The large majority still don’t take their business-critical data seriously enough and focus on confidentiality, integrity, availability and auditability of data is not sufficient to move to big data. Data first needs to become central to the organisation and South Africa still needs to adjust before taking the additional volumes of data required to get value out of big data solutions.
The benefit of big data is much dependent on real-time analytics. Does this mean that its application will be limited to only very large companies?
Chris Wood, executive head: emerging payments and strategy, Nedbank: “Not at all, most retailers don’t look at their stock or product information in real time, they follow trade cycles or stock purchasing periods. As long as the data is updated in line with the way the business operates, whether whether weekly, monthly or quarterly, it will add the requisite value. Data comparisons are also important, allowing the user to see how they are performing compared to previous cycles and to be able to effect changes to improve or fix what may not be optimal.”
John Eigelaar, MD, Keystone Electronics Solutions: ”Quite the opposite. With the recent advances in other IT fields (most notably IoT) it has become increasingly viable for smaller companies and even one-man operations to take advantage of the vast amount of information available from big data. In recent years there has been a push for industry players to adopt open standards and protocols, allowing big companies to drastically cut the costs of maintaining cumbersome frameworks.”
Warren Olivier: “Big data is not only for large companies and is not only about real-time analytics. It is about an historical context too, especially during forecasting and resource planning where companies want to look back at their performance during a specific cycle and fast-forward that to the present or even the future to provide them with a strategic edge to grow their business.”
Ian Jansen van Rensburg, senior systems engineering manager, VMware, Southern Africa: “As we move towards a more software-defined world (software being the overall IT intelligence) it is becoming increasingly cost-effective for smaller companies to benefit from real-time analytics. Companies, big or small will be able to benefit from intelligent software running on top of affordable x86 server hardware. This intelligent software supports big data, real-time analytics and business intelligence, that smaller businesses will be able to leverage to reap the benefits of big data analytics. ”
In today’s competitive business environment, companies not only have to find and analyse the relevant data they need to run their business, but they must find it quickly to be of any use. In your opinion how much public data is available and is it in an easily digestible format? And above all, is it freely available?
Gavin Holme and Rudraksh Bhawalkar: “There are organisations that provide trading data, pricing data, market data etc. to companies on a real-time basis. As the saying goes in industry circles ‘data is the new oil’ and everyone wants to make the most of it. Saying this, there is a price for every byte of data. Institutions like Platts, Dun & Bradstreet and Thomson Reuters are providers of such data sets. Some organisations like the World Health Orgnanisation provide public health data for free which opens up new opportunities to use these data sets to understand a situation and act on it, providing effective and efficient health care services.”
Kholiwe Makhohliso: ”Combining traditional data and other, possibly external data, adds additional context and provides the opportunity to deliver even greater insights. This is especially true in use cases that involve key data entities, such as customers and products. In the example of consumer sentiment analysis, capturing a positive or negative social media comment has some value, but associating it with your most or least profitable customer makes it far more valuable.”
John Eigelaar: “In recent years, the amount of publicly available data has increased drastically. It differs from industry to industry, but many companies have embraced the idea that sharing their data can benefit them in return. It is also encouraging that many IT companies not only offer data and information, but also provide internet services to process your own data. This mutual sharing of data has allowed companies to optimise their processes and ensure some robustness in their dealings with clients.”
Chris Wood: “Most data sources, at an aggregated, de-personalised level are available. This, added to survey data, geographic data and data to be shared by non-competitive merchants and retailers can have a big impact in building large data sets. Whether or not these are free is a different debate, meaning that proper big data platforms are probably restricted to being built by larger organisations and supplied to smaller entities or individuals.”
Volume is a part of a big data solution, but big data is more about unlocking the potential of structured and unstructured information. Are there enough experienced people in South Africa to analyse big data? Can companies handle this in-house or would it be better to outsource?
Deon van Aardt: ”Most of the larger companies with which we work in the mining and manufacturing industries are entirely capable of implementing big data solutions in house. Some supplement this with the assistance of system integrators for large projects while smaller companies tend to rely on system integrators completely – at least for the initial design and implementation of the system. Thereafter, the nature of the applications is such that end users are in control of the changes they might want to make. It is my opinion that, between self-reliant companies and the number of excellent system integrators available, we have enough experienced people in this country to handle industrial IT data levels.”
Anton Jacobsz: “More and more South African companies specialising in data analytics are starting to surface. However the banks are a good example of companies that have been collecting their own data for years and have used it to effectively position new products to their customers.”
Gustav Piater: ”I believe that the processing of big data used to be extremely complex and could only be done by data scientists with expensive tools. I know I might be stepping on some toes, but I believe that with the ‘new technology toolsets’ available, these capabilities are closer to the consumer.”
While some businesses produce and collect more data than others, the plethora of information now available on businesses and their customers is huge. What type of IT infrastructure will be required to handle this? Is it an option to buy time on the facilities created at the Centre of High Performance Computing (CHPC) in Cape Town or any other similar facility?
Gavin Holme and Rudraksh Bhawalkar: ”The best way forward is to implement an ‘as a service’ platform which can be charged as pay-per-use model which cuts down capex/opex drastically. This is similar to renting or buying timeshare from facilities similar to Centre of High Performance Computing where a storage facility is provided, making data available online on demand for ease of consumption.”
Garry Allemann: “The international trend is generally moving towards smaller, departmental big data implementations that may make use of cloud solutions. In South Africa we need to balance the costs associated with hosting big data in the cloud against the lack of skills that may stop us from proceeding – so it can be a good starting point. As data volumes increase at an exponential rate it may make more sense to bring data back in-house.”
Will big data be the big issue this year?
The answers were an overwhelming yes. No doubt big data, both from a storage and analytics point of view will remain on the top of the agenda. We can also expect to read and hear more about security, hardware and the cloud. They are all in the mix.