17th December 2012
In the past 12 months no technology in the IT industry has been hyped as much as big data. Every large enterprise vendor wants to ride the hype wave and capture both mindshare and potential marketshare around big data. The channel community is viewing the developments cautiously, and while some of the solution providers are eager to address the opportunity, few have a go-to-market strategy in place.
While some of the vendors have readied product lines to address the big data opportunity, few have a convincing channel delivery model in place that can help channels to generate sustainable revenue.
The data boom
Big data, by definition, is the process of deriving meaning from a large quantity of structured and unstructured data in a very short span of time. While this has been the goal of IT systems and database systems for the past three decades, it has become a reality following the advent of a number of new technologies in both software and hardware in the past half a decade.
And big data is riding a sharp growth curve. IDC predicts that big data technology and services will grow worldwide from $3.2 billion in 2010 to $16.9 billion in 2015. An IDC study commissioned by NetApp, Here Comes Big Data: Perspectives from Indian Enterprises, puts the market size at $153.1 million in India by 2014. A recent study by Informatica of the top 600 enterprises indicated that nearly 70 percent organizations are now considering, planning or running big data projects, with 44 percent considering, 22 percent planning, and rest running big data projects.
“A NASSCOM study reveals that 90 percent of Indian Fortune 500 companies are likely to have big data initiatives underway by the end of the year. However, we believe that big data implementations have just begun rolling out globally, with Indian firms in the consideration phase,“ says Venkatesh Krishnan, Head, Systems Business, Oracle India.
According to Krishnan, the biggest reason for an interest in big data has been the availability of big data. “According to analysts, data volume has grown over 900 percent in the last five years and is expected to grow at the rate of 40 percent year-on-year till 2020. Research indicates that people created 150 exabytes of data in 2005, and that the figure grew eight times to 1,200 exabytes by 2010. Four trends are driving this growth in data: the capture of detailed data more frequently across every customer interaction; the prevalent use of multimedia; the widespread adoption of social media such as Facebook and Twitter; and the roll-out of intelligent sensors embedded in physical devices that can sense, create and communicate data.”
Observing that customers are looking at ways to make better use of data, Ramendra Mandal, Country Manager, QlikTech India, says, “What makes big data captivatingly critical to organizations of all sizes is the competitive gap between enterprises that manage data effectively and those that do not.”
Speed matters, notes Jaskiran Bhatia, Country Manager, Information Management, SWG, IBM ISA. “What is different now is velocity—the rate at which data is growing and the speed at which analysis needs to happen, and variety—multiple unstructured formats in which they receive this data such as Twitter feeds, sensor readings and GPS feeds. Today, devices which we earlier never considered as computing devices are generating so much data every second. For example, energy meters.”
Another reason for solution providers to utilize big data better has been the availability of software tools that can handle both structured data and unstructured data better. Krishnan of Oracle estimates that around 120 open source tools NoSQL (non RDBMS) databases have evolved over the past few years. Says Rainer Hettinger, Principal, SAP Business Developer, Fujitsu Global,“You also have new techniques such as the Map Reduce algorithm and in-memory databases. SAP, for instance, has launched SAP Bana which takes advantage of low cost of main memory (RAM), the data processing abilities of multi-core processors and the fast data access of solid state drives to deliver better performance of analytical and transactional applications.”
There have been several innovations on the hardware front which are helping customers to analyze big data better. “Over the past few years we have seen remarkable technology innovations in terms of multi-core processors, faster bus speeds, and faster solid state drives including new innovations such as Fusion I/O and more cache memory. Today, servers perform the same amount of transactions as last decade’s supercomputers,” points out Sandeep Lodha, Director, Netweb Systems.
Getting more out of data
While for many years customers have been implementing data warehouses and business analytics to get solutions for most of their problems, vendors feel that big data could well be the panacea they have been seeking. They are quick to point out that while data warehousing and business analytics also try to provide similar answers, big data approaches the problem differently.
“It’s not just about the data size. The methodologies, the techniques used, and the technologies behind big data are vastly different from how data warehouses approached a problem. Data warehouses worked in a world where it was possible to fit most data in a traditional RDBMS. You cannot create a big data solution by fermenting data warehouses, it’s not old wine in a new bottle,” explains Vishnu Bhavaraju, Regional Manager, Greenplum, EMC India.
Mandal of QlikTech adds, “For decades, companies have been making business decisions based on transactional data stored in relational databases. Beyond that critical data is a potential treasure trove of non-traditional, less/multi-structured data like Weblogs, social media, email, sensors, device logs and images that can be mined for useful information using big data.”
But how big is big data? Most vendors refuse to classify big data according to a database size limit. “It’s mostly about the business problem which you are trying to solve. Even a data size of 2 TB which cannot be mapped and fitted into a traditional RDBMS may fall into the realms of a problem that can be addressed by big data,” says Bhavaraju.
Rajesh Awasthi, Director, Telecom & Cloud Service Provider, NetApp India, agrees. “One way to think of it is that it’s so big that the tools we use today cannot cope with it. This is the reason why the core thinking around how we ingest data and transform it into insight and information is changing. Data by itself has no value unless it is translated into information which can help drive business results, and time-to-information is critical to derive maximum value from data.”
Is there a channel role?
The million rupee question is whether traditional Indian IT channels can effectively address this growing opportunity. The industry as well as the partner community are showing cautious optimism. Oracle is a vendor which seems to be optimistic. It has already launched programs around training and enabling its channels around products such as Oracle NoSQL Community and Enterprise editions. The software giant has already announced partner training and PoC initiatives around the ZFS Storage Appliance, Pillar Axiom Storage and the Oracle Big Data Appliance for its hardware channels.
“At Oracle we believe that channel partners play a very vital role in our larger go-to-market strategy. Channels can partner with Oracle to create solutions to manage the entire lifecycle of big data—to store, stream, acquire, organize, analyze and visualize big data,” says Venkatesh of Oracle.
EMC has been pushing its Scaleout NAS offering Isilon and Greenplum, the database and database appliance solutions, for big data, and over the past couple of quarters has been enabling channel partners. “Some of the top-tier partners have been enabled to identify opportunities, and we are game to start PoCs with channel partners,” says Bhavaraju.
However Praveen Sahai, Director, Channels & Alliances, EMC India, feels that big data adoption through channels will happen in two phases. “We expect channels to sell scalable storage and data management technologies, and help business manage data in 2013. You will see a much wider sector-specific analytics business taking off toward the end of next year with a huge channel play.”
IBM says they have been investing considerably in building channel competence to address big data. “We are extensively engaged with a large number of partners who are working with us as big data SIs and resellers. They possess skills in our big data platform and suite of products. Apart from our analytics portfolio, Cognos, we have recently announced the newest member of the PureSystems family, PureData System for Analytics, a simple data appliance for serious analytics,” says Bhatia.
S Sridhar, Director, Enterprise Solutions Business, Dell India, feels that channels can play a major role. “You can divide the big data opportunity into three aspects. The first is the supply of servers, storage and software components; the second is the systems integration; and the third is the business analytics. There are easily 200 good partners in the country who can do an effective job on the first two. The third element is the tricky one. But then the partner does have the choice to work either with a vendor like Dell or a third-party solution provider for the analytics bit.” Dell has announced big data in a box solutions with Cloudera, the commercial vendor behind Hadoop. Sridhar said that Dell will soon launch the product in India.
While agreeing that the triangulation of a partner and an ISV is the best model to enable channels to address the big data opportunity, Santanu Ghose, Director, Business Critical Systems, HP India, explains that for most vendors evolving a commoditized and predictable channel model is not easy. “We have hardware and solution stacks to address a customer’s requirement for a big data project. We have also enabled our channels to sell the products through comprehensive channel programs. However, big data requires considerable domain expertise and data analytic skills, and it’s up to the partners to build these skill-sets.”
Sriram S, the CEO of iValue Infosolutions, is more forthcoming. “Big data is not an opportunity which can be addressed through commoditized channel programs. It requires immense domain expertise, data science skill-sets, and above all the patience to deliver projects over a period of 6-12 months.”
However, Ghose says that it’s just a matter of time before vendors such as HP come out with channel initiatives to address the space. “Years back the business intelligence and data warehousing market went through such stages, and today we have fairly exhaustive channel initiatives to address the space.”
Awasthi of NetApp says that partners need to look at data and not just at storage, and that there are opportunities beyond analytics. “We at NetApp view big data a little differently. We categorize the big data solutions into the ABC of big data—Analytics, Bandwidth and Content. We believe our partners can address the opportunity to create storage for data-intensive workloads at really high speeds (Bandwidth), and boundless, secure, scalable data storage to write it, find it and keep it forever (Content).”
Deepak Jadhav, Director, VDA Infosolutions agrees with Awasthi’s view. Admitting that at present most vendors are keener to get global SIs on board for big data projects because of their reach and better skill-sets, Jadhav says, “We know that analytics is where the big money is, but we don’t have a first-mover advantage. We are therefore content to bet on the Scaleout NAS side of the business. We are presently readying a team with cross-domain skills to go after the analytics side of the big data business.”
Noting that presently there are no qualified big data projects which the company is addressing, B Shankar, Director, Ashtech Infotech says, ”Almost all the vendors including IBM, HP, Oracle and EMC have been trying to educate the enterprise channels. The hype is huge, but the potential is bigger. It’s just a question of time before the tier-2 channels will be activated.”
Other partners are more optimistic. “Today in India we have done close to half a dozen projects that can be classified under big data. This includes a project for an aviation giant for whom we implemented a solution which collects information from GPRS phone activities at the Delhi airport terminal and uses that to build a large information store to build a massive customer data study,” says Kamal Sharma, ITIM SBU Head, Mindlance, Bengaluru.
Another project Mindlance has recently executed has been for a leading quasi-government transportation entity in Bengaluru; a massive data store has been created to analyze and understand the customer base using the company’s services. Sharma advises that channels or customers should not be intimidated with the term ‘big data.’ “You need to look at it as handling a project which involves data which cannot be cataloged with known RDBMS tools.”
A number of startups which are eying the Indian market have also sprouted up. Comments Sridhar Dhulipala, Co-founder & Director, Solutions, Bizosys Technologies, which is focused on opportunities in big data, “When we started we were looking at the north American market which had companies with data stores that were really mature. However, we have recently bagged a project from a Bengaluru-based pharma vendor which wanted us to improve the performance of its Hadoop cluster. Another project we have bagged is for a Mumbai-based BFSI company.”
Most vendors believe that the telecom, retail and BFSI sectors hold the biggest potential for rolling out big data projects. “The telecom industry has generated petabytes of information, and there’s huge scope to mine this information and come up with answers to questions that were previously impossible to answer. Retail, especially e-commerce companies, will be offloading projects to collate visit reports, log reports and purchase data for big data projects,” says Bhavaraju of EMC. Ghose of HP feels that BFSI is considering big data seriously. “After implementing core banking projects, the BFSI segment has generated tons of data which are impossible to sieve through with the present tools, so this segment will start piloting big data projects next year.”
While big data is still perceived as hype, there is no skepticism regarding its potential. However, vendors need to figure out channel initiatives and business models that would allow partners to build sustainable revenue, build predictable funnels, and address real opportunities in the space. At the same time, channels cannot expect vendors to help them build skill-sets and domain expertise. The number of new channel partners is increasing, and several partners are building cross-domain skills to address the opportunities in the space. 2013 promises to be the year when the action is likely to start in the big data world