Can CIOs afford to ignore big data?

Writing about big data these days has its own set of challenges. There is so much data out there, but getting what you want out of that is indeed a challenge. Adding to the woes, some of the CIOs we spoke to pleaded to leave them out of the story, saying: "I do not have anything to say any more."

Other CIOs urged not to quote them. Another category, which we personally believe sounded more like ‘Wikipedia entries' ( we are not quoting them as you can get that in Wiki anyway), gave long lectures on the evolution of big data and the possible ramifications. And yes, some CIOs did say they are all geared up to big data challenges.

Broadly speaking, the tone of the enterprise IT community is one of skepticism to cautious optimism to one of ‘you cannot ignore it' kind of a mindset.

Meanwhile, the vendors and the analyst community demonstrate a kind of technology jingoism and any conversation these days obviously end with big data and what it is doing and often laced with jargons, which makes many to silently look out for those big data for dummies' books so that ‘he/she can also be a part of those conversations'-with great comfort at the expense of the other person's discomfort.

The 2 Sides of the Coin

Jokes and skepticism apart-like it or not big data is here to stay. And it's up to the CIOs to decide whether to be a part of that bandwagon or not. Yes, often it is hyped and a India centric view at this point in time is in the ‘debate' stage but probably one year down the line we will see much more proof points and success stories on big data.

Says Venkat Iyer, CIO, Wockhardt: "Currently, all these big data conversations is air to me, nothing new about it". Other veteran CIOs like Cipla's Arun Gupta too agrees that there is ‘too much hype' surrounding big data.

Recently for an interaction, Dataquest did for our annual-CIO Handbook, Gupta said, "The propositions and stories on big data are all hypothesis of what can be done, no real business problems have been solved as yet, and neither have any companies found real use. Because big data is all about unstructured data while structured data has models to generate conventional wisdom and where is the empirical model for ROI."

 

Vendors-are you listening? Can you all present in simple terms what big data can do for Indian enterprises and what is the real bottom line impact the enterprises will accrue out of big data? Many CIOs cite that most of the vendors present information that is ambiguous and makes for demystification.

A Study on Big Data

Let's here look at how data among Indian enterprises has grown in the last few years. According to the study ‘Here Comes Big Data:

Perspectives from Indian Enterprises', commissioned by NetApp to IDC, "Over the last 2-3 years, with the increased technology adoption, and emergence of newer form factors, along with the rise of enterprise mobility, cloud, and social media, the data growth has been tremendous.

About a third of Indian organizations covered in the survey have indicated more than 60% year-on-year (y-o-y) data growth."

The study also pointed out that there is a high prevalence of unstructured data in Indian organizations; with more than a third of Indian enterprises having mostly unstructured data. And a major chunk of organizations do not have a formal big data strategy yet. More than a half of the organizations do not distinguish between data and big data. Big data management still figures as a part of the overall data management strategy for a major chunk of enterprises.

Quips Venkatesh Natarajan, special director, IT, Ashok Leyland, "Organizations have suddenly started realizing that they have so much of data captured from a variety of sources. Structured data from ERP, CRM, DMS, PLM systems and those residing in the test beds, data captured through telematics systems, etc. Then we have the unstructured / textual data from social media, blogs, etc. Mixing the data from these sources and extracting value from the information is going to be the challenge for CIOs in the future. How do you translate / convert big data to big information? Faster systems unleashing tremendous power in computation backed up by ‘user customizable analytics' is the key for exploiting the power of big data. Large enterprises have now started identifying the sources of data within the organization. In the next 2 years, big data will be all pervasive."

Clearly organizations are just about scratching the surface and it will take time to get a form and desired function and outcome. Even if we go by the IDC study it says: "The big data solutions market in India is currently at a nascent stage, with only 5% of the organizations in the large and very large segment (with more than 1,000 employees) having embraced this technology."

IDC feels that this is a point of influx, and the market would experience an exponential growth in the next 2-3 years. By 2014, it expects the market penetration to reach 18.4% of the organizations in the large and very large segment across different verticals. A majority of the Indian organizations (52%) do not feel the need for implementing big data solutions yet. Around a third of the organizations (30%) feel that they lack information on the potential of big data and the available big data solutions."

Creating Data Scientists

Agreed there is hype. But, arriving coherence out of data clutter is the starting premise. Clearly most organizations need to change the perception as ‘data' as one ‘tied' to storage. A typical reactive measure was to increase the storage size as data grows. The real emphasis of leveraging the benefits of big data comes from analytics. Big data has 2 connotations-one data sizes can be humongous and hence you need to mine it for profit.

Telecom companies has been doing this for years, even though the terminology called big data per se did not exist in the past. For the rest of enterprises, data might not be huge but until unless they discover and define all sorts of data- structured to unstructured- a business strategy cannot be made. Here for smaller companies the term ‘big' in big data refers to the big impact it will have on their organizations.

So clearly the vendor community need to address this thought process that the big data is more about analytics and the emphasis has to be on ‘data seamlessness' out of what is residing inside the ‘storage silos'.

While we are not putting any vendor pitches here, and since EMC is one of the early movers in this space we asked Rajesh Janey, president, EMC India and SAARC, about the India specific opportunity. According to him: "The global demand of big data analytics services is slated to become a billion dollar industry by 2015. But it is not as simple as it seems. The only way for big data analytics services to take off is to have people skilled in data science, which is a rare breed of professional today."

"As per a couple of industry researches, its estimated that cloud will generate employment of 100,000 by 2015. Furthermore, big data which stood at $58.4 mn in 2011 will reach $153 mn by 2015. Big data has resulted in creating a professional streams of its own popularly known as big data scientists. As per a Mckinsey report, by 2018, there would be a shortage of 140,000-190,000 people with deep analytical skills globally. India stands a very good chance of creating these data scientists which involves a strong focus on Maths, Science, Statistics, and related spheres. The challenge is the shortage of appropriate curriculum to equip students of these emerging tech areas," adds Janey.

Clearly organizations need to invest in data scientists just like when security as a major concern came , one saw the emergence of CISO's complimenting with the CTO/CIO, and in similar fashion, enterprises first need to invest in right talent and then their big data play would become much more practical and feasible.

If we were to present a use case scenario, telecom is the best vertical to look at how data can be impacted for business strategies. Recently, Fredrik Jejdling, head for Ericsson's India arm, wrote in Dataquest : "Big data is an important pre-requisite for operators wanting to leverage value from the large volumes of data in their possession in a cost-efficient way. OSS/BSS tools can help operators leverage the value of the massive amounts of data contained in their networks."

A Big Data Plan?

Reflecting on why CIOs need to have a big data plan, Abhishek Bhattacharya, director, technology, Sapient Global Markets, tells Dataquest that: "Big data adoption in some industry segments is facing some headwinds. I feel there are 2 reasons behind it. The first reason is that big data technologies force people to ask different questions, it brings a new way of thinking, which is not always easy to adopt. For example, it forces a CIO to change by providing a better understanding of results from previous quarter and helps to model and predict the future quarter mindset. This challenges the status quo, changes the current decision process and at times questions the common wisdom."

"The second reason is that the results are not always certain with big data initiatives. There are several reasons for the uncertainty. For example, the technologies are not always easy, data is not always high quality, right skill sets maybe missing, wrong questions maybe asked, etc."

To sum up, as Bhattacharya says, "But I feel big data is the future and will be the critical strategic difference between 2 competing firms. It will revolutionize business models and how decisions are made. To be successful in this space, organizations will have to make a start, learn as they develop and keep an open mind. There will be mistakes but it is important to keep at it. Firms that are committed, will be successful. In that respect it is like being in a relationship, more committed you are, more fruitful is the relationship."