Big Data is one of the hottest buzzwords of today, and for good reason. The
possibility of massive data sets from many sources previously unthinkable
quantities to be processed, it is now possible for companies in every industry,
not only supercomputers labs. Wal-Mart
and Coca-Cola Freestyle Labs next-generation beverage dispenser, both the use
of new database and analytics platform for real-time data to predict customer
demand and to help shape new products to analyze.
CEOs starting to wonder if large data for real and if it can deliver a significant competitive advantage to their businesses. It is crucial to the risks and benefits of previously eligible because large data can get expensive fast. Here are the five questions CEOs should ask to help ensure the return of those gigantic analytics efforts.
1. How much data will help my business?
To the CIO, large data is about exploiting the size, speed or variability of data from many sources to support business objectives. For the CEO, it's about the insights that occur when a company can merge data from operational systems - such as customer and inventory levels - with less traditional data, such as a listing on the website, Twitter feeds and e-mail text . Analyzing large amounts of data in these new combinations is the place where a company can move forward. CEOs not matter where the information comes from or what technology is used - but they need to know if big data will force them answers about the future or new ideas for helping their companies better.
For example, a car company to determine the online ad keywords lead to the biggest new customers for a particular model. Then the company can alter its online marketing plans on the fly. CEOs must be able to provide IT direction on the top questions in the next quarter or the next 12 months, the biggest impact on operating results will have.
2. How much will it cost?
Big data applications are not easy to build and manage, requiring specific technical expertise for the development and maintenance of storage and analytical platforms and specialized analysts called data scientists who work with large amounts of raw data.
The "large data stack" normally consists of heavyweight processing and analysis systems, data cleansing tools, visualization tools to the reports and the servers and storage systems that house and manage all data. The cost of building and integrating the large pile of data, including staff, can easily run into millions for a large company. For example, the U.S. government recently forked over $ 13.4 million for a National Science Foundation data mining project.
Depending on the complexity and budget of the project, there are three choices for the implementation of a large data project:
• Best of breed: A company can be any technological component of the best supplier in each category and build large data applications themselves-an expensive, but very flexible and powerful option.
• Packaged applications: A company can pre-integrated large data package to purchase from a single vendor technology - an excellent and cost-effective choice, especially if you're planning your big dates to concentrate in one area, such as analyzing online customer behavior.
• Software as a Service: SaaS vendors will gather and store the data, the home of the infrastructure and do all the heavy lifting analytics and data management for you, avoiding the significant investment in hardware, software and personnel.
3. How risky is it?
Investing in the right people for your big data project that is running is probably the best way a company can minimize the risks. Without data scientists and analysts accustomed to working with large unstructured data sets and draw conclusions from, your project will not have far to go. These specialists are not cheap, however, and they are still a scarce commodity.
CEOs need to understand what it takes to be a big project data over time to maintain. When the stored data is growing, and the company depends more on the insights that generate large data, the system difficult to manage. A company will need to plan for continued investments in technology infrastructure over time, and work directly with the CIO will assist in determining the smartest way to economically scale your IT systems from the beginning. Finally, applications and services that the modern, distributed applications can monitor the internal and external networks, such as Amazon and Google's cloud services are needed to delay a large data project on the knees to avoid.
4. How do we measure efficiency?
CEOs want the ultimate business metrics the company tries to improve, such as reduced application downtime or an increase in online sales to follow. In the industry, large data analysis comparing the results of two different supply-chain strategies for the same product, to see which is cheaper. In the financial services industry, large data is used to target the optimum profiling for a new product, with respect to the traditional methods of cross-selling, with the purpose of higher conversions.
But CEOs must remember that while large data can provide powerful information to help in decision making, it does not directly affect business results. That's where the smart people in the company are those who know how to analyze and use the information to a process or product change. Metrics program must also measure what decisions and actions occurred when - and how those actions affect final results.
5. How long before I see results?
CEOs may have too great expectations of large data, based on what they read in the media. It is important that CEOs work closely with their colleagues to understand the potential, with a line of business managers who may be involved. A SaaS provider can possible outcomes for a company within a few minutes after receiving their data because they already have the infrastructure and processes in place. If your company manages a large data project internally, it may take longer to see the results. How a company defines the success of large data will obviously affect time-to-results, and with each new technology - it's always smart to start small and easy to plan.
Big Data is still in its infancy, with many unanswered questions. But CEOs still need to do their homework, that angry members come knocking on the doors when large parts of the money be spent with little to show.


