Hype Vs Reality in Big   Data 

“We ourselves sometimes speak   about ‘Big Data solutions,’ but in   reality what our Stratecast Big Data   & Analytics practice is about are   data management solutions, some of   which are beginning to effectively   manage Big Data. And, contrary to   what is fast becoming conventional   wisdom, some organizations are in   fact deploying data management   solutions addressing Big Data.   However, especially noteworthy   given the benefits that harnessing   Big Data has to offer, the rate of   adoption is low. One reason is a lack   of clarity in the value proposition:   too many providers and hopeful   solution buyers are not making the   case effectively for why those who   sign the checks should invest.

“Deploying ‘the right stuff’ and not just‘more stuff’ alsoenabled the retailer to slash its IT footprint,and as a result, it is now saving more than $1.4 million annually”

The   benefits include the opportunity   to create an agile, data-driven   organization; leveraging data to   support KPIs; and transforming   fear of Big Data into employee   empowerment, offering the insights   people need, when they need them,   to make smarter decisions. All of   that translates directly into revenue   generation and retention, and while   vendors obviously hope to capitalize   on this by selling more products, we   see examples all the time where   companies actually save money by   deploying a Big Data solution. For   example, a bricks-and-mortar mega   retailer in Europe just markedly   enhanced its competitive position   with Big Data-driven real-time   analytics. Deploying ‘the right stuff’   and not just ‘more stuff’ also enabled   the retailer to slash its IT footprint,   and as a result, it is now saving more   than $1.4 million annually—while   understanding its business better   than ever before, and responding   faster to the needs of its customers.   “Stratecast estimates the market for   Big Data, analytics, and Business   Intelligence (BI) solutions at $25   billion for 2013, increasing at a   Compound Annual Growth Rate   (CAGR) of 12.7 percent to $40 billion   by 2017.”

Developing Big DataCapabilities In-House Vs.Outsourcing
“While some organizations are   certainly crafting their own Big   Data systems, they are often better   served by engaging a professional   team to either provide a Big Data   solution outright, or, at minimum,   work with the organization on the   initiative. This can easily cost the   organization less than doing the   job in-house. We urge buyers   to consider the following direct,   indirect, and opportunity costs in   constructing a business case:

 • “Failure to capture missioncritical   data–with a poorly planned   Big Data implementation, the   business will not capture all relevant   data to compete effectively in the   marketplace and retain existing   customers. This can render the   business vulnerable to competitors,   leading to missed opportunities, lost   revenue and higher churn. For the   business case analysis, calculate   the likely and potential top line and   opportunity costs of continuing to   operate in the absence of complete   data.

  • “Taking multiple ‘shots’ at Big   Data can delay implementation–   implementing a Big Data system is   a massive undertaking that touches   every area of the business; and   the downstream impact of any   delay is magnified. From decades   of experience optimizing IT   infrastructures, and more recent   experience implementing Big   Data solutions, expert consultants   understand where the vulnerabilities   and risks are, and plan accordingly.   With proper upfront planning comes   lower exposure to risk, decreased   chance of unexpected disruptions,   on-time project completion, and   an agile business that seizes new   opportunities instead of missing   market windows.

  • “Resource drain on internal team–   most IT and Data Science teams are under pressure to maintain daily   operations and incorporate new   business-enhancing technologies,   without increasing staff. By hiring   a team of experts, they can keep   internal staff doing the things they   need to do and avoid the costs of   delaying or deferring essential   tasks.   • “Implementing Big Data   architecture confines costs to   hardware–while merely adding   servers will not manage Big Data,   it is sometimes necessary to add   servers when implementing a Big   Data infrastructure. Even so, if the   organization has worked with an   experienced professional service   team to incorporate the Big Data   Architecture elements in Stratecast’s   Data Management Model, the cost of   the servers themselves should be   the only incremental cost.

 Making Big Data an  Integral Part of Business  Strategy
 SMBs and below should have their   heads in the clouds. Where ‘Big Data   for the rich’ may leave midsized   companies is in cloud-based Big   Data solutions. These solutions can   turn the embedded/legacy IT model   on its head, and help companies   do more with less. Reducing the   IT footprint of legacy technology   reduces both Capital Expenditures   (CAPEX: upfront   investment) and   ongoing Operating   Expenses (OPEX).   Some providers who   are providing more   affordable solutions,   some entirely cloudbased,   some premisesbased,   and some a   hybrid of technologies,   include Birst, GoodData,   Starcounter, GridGain,   Platfora, and GigaSpaces. Do they   have huge market share at present?   Compared to the market leaders—   No. Yet Birst has more than 1,000   customers including Citrix, YMCA,   and CBS Interactive; Starcounter   has 80 large customers including   Hyundai, Telia, Canal+, European   mega-retailer Gekas Ullared, and   Verizon Wireless; and the others   in this grouping also have robust   customer bases.

 Another challenge to the growth   of this market is the ‘IT glass walls’   phenomenon: at present, it still   requires too much specialized   expertise, in the form of the data   scientist or CIO and team, for   business users to quickly get the data   they need. The cloud-based data   management providers and other   innovators in this space are making   headway by providing solutions   that are not only affordable but also   geared to enabling business users   to get the insights they need without   having to beg IT for information.

 Seriously consider Hadoop—but   get all the Hadoop you need. We   hear too many companies in effect   saying, ‘I’d like to order one Hadoop,   please,’ as if that is going to   solve the Big Data riddle for them. It   is certainly not an exact analogy, but   I liken Hadoop to my Firefox browser:  I love it, but only with my favorite   add-ons like Tab Mix Plus, FireShot,   Google Translate, QuickTime, and   Adobe Shockwave. Similarly with   Hadoop, to get the most out of your   data management solution you need   modules that have been developed   by other Apache Software Foundation   working groups: modules such   as Pig to accommodate semistructured   data and simplify   various tasks. Hive to use Hadoop   as an Enterprise Data   Warehouse (EDW). Sqoop and   Flume to import Web log data   into a Hadoop Distributed   File System   (HDFS), and to integrate   structured   data into the mix.   Ambarri, Whirr,  Zookeeper, and  Ozzie to accelerate  deployment, simplify  ongoing management,  manage workflow,  and maintain data  synchronization.