Against "Big Data": five preparatory strategies for IT department

Large data is hailed as the key strategic business asset for the future. This means that in the near future, employees in other departments will also want to know the IT department's strategy for big data.

How do you explain that to them? To be sure, handling large amounts of data is not new to most IT departments, but analysts say big data is different from data warehousing, data mining, and business intelligence analysis.

Data is accumulating at unprecedented speed and variability, and unlike previous data, most of the data is unstructured and raw. Blogs, social networks, machine sensors, and geographic tools are generating a lot of unstructured data.

"We've produced a lot of data and we've been collecting data, but it's very limited and no one really knows how to deal with that data," says Paul Gustafson of Computer Sciences.

Industry watchers say it stands at the forefront of the data revolution.

Catalina Marketing manages a 2.5PB customer loyalty database, including data from more than 190 million U.S. grocery stores collected from the largest retail chain. The data is used to tailor marketing strategies to U.S. consumers using grocery cards, and to issue coupons based on Consumer records. In order to lead the enterprise to the real-time predictable intelligence age, Catalina marketing Eric Willams and other industry observers said that technical managers must develop their own enterprise information management architecture and culture to support advanced analysis of large data storage.

Mark Beyer, a Gartner analyst, says savvy IT leaders should be prepared to fight big data rather than sit on their own.

Here are five things the IT team needs to do to prepare for a new era of big data.

Inventory your data

Almost every enterprise faces a steady stream of unstructured data, from social networks or from sensors that monitor factory floors. But the fact that businesses are generating so much data doesn't mean there is a business need to save every byte.

"When you first start to face big data, people feel the need to know all the data from blogs or sensors," says Neil Raden, a constellation study analyst. ”

Part of the reason may come from suppliers and consultants eager to promote the next big thing in the field of computing, says Raden. ”

Smart It experiences do not attempt to process all the data, but rather find out which data is relevant to the enterprise and which is irrelevant.

Raden says the first step should be to take stock of data, see which data is internally created and which are external data resources, and this will help IT departments better understand the data and add insight into the business.

While defining the scope of the data, it should carry out targeted projects to show results rather than large data projects, "You don't need to spend millions of of dollars to start a project to see if it's worth investing in," Raden said.

Talk to business needs

Many of the initial large data projects began in areas other than it, for example, the marketing department has been using media streaming to better understand customer needs and buying trends.

From a business perspective, experts in specific areas may see the opportunity to make big data money, but it needs to be responsible for data sharing and data syndication concepts----Big Data strategy.

"It's hard to succeed if big data strategies don't align with business goals," said Dave Patton, a leading information management analyst at PricewaterhouseCoopers. ”

Catalina Marketing in developing a large data strategy, Williams asked the business manager and the Financial Planning and Analysis (FPA) team to discuss business solutions to the information architecture investment.

The business perspective can determine the value of new projects, such as purchasing items based on the products in the cart, and the FPA team can quantify the claims to increase productivity or increase sales.

Reassessment of infrastructure

Gartner's Beyer and other experts believe that a large data strategy will require significant changes to the server and storage architecture and information management architecture, and IT managers need to be ready to expand their systems to meet the ever-expanding, structured and unstructured data stores.

The IT team needs to find the best way to make the system scalable and to develop a roadmap to integrate all the different systems associated with the big data strategy.

"Now, most businesses have different island systems for different purposes, such as customer management, marketing, etc.," said Anjul Bhambhri, vice president of IBM's large data product, "The CIO needs to develop a strategy to integrate these decentralized island systems and build a central system." ”

Familiar with the relevant technology

The Big data field has allowed us to recognize a long list of new terminology and technologies that CIOs may never have heard of.

Among them, open source tools attract the attention of most people, such as Hadoop, MapReduce and NoSQL technology is helping the web giants (Google and Facebook, etc.) to handle its large data storage. Many of these technologies are still quite immature and require specific skills from the operator. Other important technologies related to large data areas include database analysis, columnar database and database warehouse equipment.

IT managers and their staff need to understand these new tools to make sure they can make the right big data decisions.

Get your employees ready

Whether a business requires a Hadoop expert or a data scientist, most IT companies lack the necessary technical talent to do the next big data strategy. Analytics is perhaps the most critical, and this is where most IT personnel have the biggest gap.

McKinsey expects that in the United States alone, by 2018, 140,000 to 190,000 experts in statistical methodologies and data analysis technologies will be required.

In addition, McKinsey expects to require 1.5 million data managers with formal predictive analysis and statistical training in the business or technical field.

For some companies, especially those in less populated areas, people will be the most challenging aspect of the big data strategy. True Textiles Rick Cowan, chief information officer of the company, said: "The big data field absolutely requires a different mindset and different skills." ”

"As a medium-sized enterprise, it is a challenge for staff to adapt to a changing environment," Cowan said, in order to meet the demand, he has begun to retrain programmers and database analysts to enhance their analytical capabilities.

In order to stand out in this new field, the head of the IT department needs to make some adjustments to themselves, Gartner Beyer said. Although the best technology leaders in the past are half the information managers and half the infrastructure engineers, future IT managers will be a combination of data scientists and business process engineers.

"CIOs have always managed infrastructure based on specific directives from the business sector, rather than discovering opportunities, and then driving information innovation," he says. "This is where the change is needed." ”

(Editor: Technology)