Forecast the six development direction of business Intelligence 2013

Wayne Eckerson is a senior data specialist with over 15 years of industry experience in the fields of data warehousing, business Intelligence (BI) and performance management. At the turn of 2012 and 2013, he forecasts the direction of the BI market for the New Year and provides six key phrases for the broad DBA's reference.

Hadoop changes the Data warehouse.

Internet companies that want to try Hadoop and cloud computing need to migrate all their data, including master data, into Hadoop and turn those relational databases into data marts. This is a step-by-step process. However, once they adapt to Hadoop, more and more data flow and analytics applications are put into Hadoop, and traditional data warehouses are left to applications that require relational processes.

The analyst turned into a salesman.

In a mature bi project, the ability to analyze does not represent the implementation of the project. Analysts can paint all the hype--excellent insights, reports, dashboards, models, and information maps; But if the enterprise does not take the appropriate decision, all is discreetly. As a result, analysts must sink in and not be complacent about the results of their analysis, but rather to exert more effort on the decisions of executives. In other words, the analyst must complete the transaction as the salesperson does. Only if an enterprise implements their project, does the analyst's job make sense and there will be bonuses.

Independent data governance projects may be abolished.

Many experienced BI administrators are frustrated by not being motivated enough to implement a formal data governance project. They argue that data governance is so important that it cannot be incorporated into regulations, but only from business needs. Although business people do not need to participate in data governance projects, it is necessary for them to attend a number of meetings to ensure the smooth implementation of the project. In view of this, it is better to abolish the independent Data Governance project and classify it as a project consulting method.

Seek data integration from federated data warehouses.

As the BI project matures, you may suddenly realize that multiple data warehouses and data marts in your enterprise are a small hurdle. In most cases, the enterprise will not have any time, effort and cost to invest in building the Super Data Warehouse. Then, after designing a series of scenarios, you'll find that federated data warehouses are the only way out. Fortunately, with the rapid advances in data virtualization, we can build a virtualized enterprise-class data Warehouse today, and generate a new view of different data without wasting the financial and material resources of an artificial example.

Cloud computing BI develops further.

SMEs, start-ups, and nascent bi companies have long been affected by traditional in-house IT operations and are easily attracted to Bi's novelty. As a result, they don't think too much about cloud security, putting the company's sales data, payroll, expense, and countless applications on the cloud. In fact, these companies are the pioneers in the development of BI in cloud computing, just like those who embraced E-commerce in the last century in the 90. In contrast, large enterprises tend to follow the trend, step-by-step to the risk factor of the BI application of low to the cloud.

Self-service bi is not easy to implement.

I sympathize with the new bi companies because they have so much to learn, and the first is the self-service bi. Although business users claim to be "self-service", most people just talk. 80% of them are temporary users, and they need a greater degree of interactivity; only the advanced users need to be in the true sense of varying degrees of control. However, it is difficult for these advanced users to master the BI editing tools. These are issues that are easily overlooked by the nascent BI enterprise.

(Responsible editor: Schpeppen)