IBM and Sprint chat Big Data Project practices

&http://www.aliyun.com/zixun/aggregation/37954.html ">nbsp; Companies running large data management and analysis projects should focus on "customer-centric results", says Martin Wildberger, IBM's vice president for global information management.

On the IBM information on Demand meeting (IBM information on Demand) held a few days ago, Wildberger said that the goal of achieving customer-centric results requires understanding how to use large data to communicate more effectively with customers.

Sprint shares large data project instances

Among IBM customers, Sprint is a good example of using large data to improve customer service.

As one of the largest telecoms companies in the United States, Sprint has nearly 5.4 million customers and provides a large number of products for individuals, businesses and governments. Recently, the company began using IBM's Infosphere streams and some analysis tools to try to analyze the vast amounts of data generated by the company's network users on a daily basis.

"As an information medium, we have a lot of real time transactions every day," said von McConnell, executive director of Sprint's senior analysis. It may surprise you that we have a billions of transaction volume every day. ”

One of the goals of this project is to get new insights into how users use the Sprint network, and ultimately to find ways to improve the customer experience and make it more humane, based on all the information. In this regard, McConnell that the company still has a long way to go, but has made some progress.

"We have only recently begun to think about data governance; we want to figure out: What is data? How can we use them? What can they bring to us?" McConnell and his team are proposing different issues for the data to be able to predict network activity and circumvent the negative user experience.

"I really believe that the project will be good for telecom because we are not just optimizing the data, we are starting to have questions," he said. It is not easy to take this step, but we will certainly succeed. The value contained in the data is too great for us to see the dawn of the future. ”

IBM more suggestions for big data projects

In addition to the above, Wildberger provides a number of recommendations for a large data management strategy for the enterprise.

Wildberger said, on the one hand, at the beginning of the project more attention to customers, on the other hand, to develop a large enterprise data planning. In planning, you need to detail how a large data project should be applied to a department in the enterprise. The implementation of a new scheme must take into account the impact that these new insights will have on the enterprise as a whole.

A good way to do this is to start with a short-term goal, starting with the existing data (that is, the information stored internally). Wildberger said: "You do not have to start to overcome the project's difficulties, first from those you can harvest short-term results of the analysis, step-by-step in-depth." ”

The next step is to start building and acquiring the skills necessary to implement the project. You can recruit people or train existing teams. Remember, don't let technology deviate from business priorities.

The final step is to ensure that the application cases for large data management and analysis come from business results that are easily evaluated. "Large data cases must be set in the same direction, keep logic." ”