Most companies would like to see broader adoption of business intelligence (BI) by business users and thus achieve the strategic objective of becoming a more analytics-driven organization. There are several challenges and barriers to achieving to this, however.
Two of the main challenges are the increasing data volumes and the number of data types and data stores that organizations are experiencing. This trend, coupled with the move toward more advanced analytical processing, is requiring BI systems to support more complex and extreme analytic workloads, which drives up costs. At the same time, business users are faced with limited BI self-service, increasing tool complexity and extended IT development and deployment times because of restricted IT budgets.
Increasing costs and the slower time-to-value of BI deployments make it difficult for organizations to grow their use of BI and thereby exploit its value for improving business efficiency and revenues in today’s tough economic environment. BI user self-service is an important and effective way of overcoming these impediments to BI growth. To be effective, however, BI self-service must be done at scale, that is it must be able to cope with the constant challenges of growth and technology advances throughout the complete information supply chain from capturing and analyzing source data to delivering the final BI results to business users.
Software vendors have three main areas of development that promise improvements in BI functionality and adoption:
- Next generation solutions: Support for big data and advanced or smart analytics using analytic relational database management systems (DBMSs), non-relational (or so-called NoSQL) systems such as Hadoop and/or stream processing systems.
- Reducing costs: Optimized systems for investigative computing and built-for-purpose applications, exploitation of new hardware improvements and packaged hardware and software appliances.
- Faster time-to-value: Cloud computing, agile BI and data warehouse development, and self-service BI.
Although these improvements increase the power and adoption of BI, they also increase the complexity of the BI environment. This is why organizations need to extend their existing data warehousing systems to enable these technology developments, while also at the same time scale their self-service BI solutions to support this extended data warehousing environment.
The Extended Data Warehouse
A possible framework for extending the existing data warehouse environment is illustrated in Figure 1.
Figure 1: Extending the Enterprise Data Warehouse
This extended environment consists of three main components:
- Traditional enterprise data warehouse (EDW) environment where an EDW, dependent data marts and data cubes are used to store historical snapshots of structured data captured and transformed from operational systems.
- Investigative and built-for-purpose systems that employ analytic RDBMSs and non-relational systems to store, manage and analyze both structured and multistructured data.
- Stream processing systems that process in-motion event and sensor data as it flows through IT systems.
BI Self-Service at Scale
Self-service BI deployments need to support the four components of the extended data warehouse just outlined, while also providing business users with four key benefits:
- Make it easier to access data for analysis by providing data virtualization capabilities and connectors to big data systems such as analytic relational DBMSs and non-relational systems.
- Make a data warehouse fast to deploy and easy to manage through the use of public and private cloud computing platforms, BI software as a service and packaged hardware/software appliances.
- Make BI tools easy to use by features such as data mashups, customizable widgets, easier data mining, packaged analytic functions and sophisticated visualization that can be used in conjunction with any data store that is a part of the extended data warehousing environment.
- Make BI results easy to consume and enhance by supporting office product integration, a business glossary, data lineage reporting and actionable BI (alerts, recommendations, decision analysis workflows, etc.). Enabling mobile BI users and adding support in BI systems for collaborative interaction, information enhancement and collaborative decision making also enhance the usability of the BI system.
Choosing the Right Self-Service BI Solution
We can see that the direction of BI workloads is toward supporting multiple data systems and increasing data volumes, while also enabling more sophisticated analytical processing. To support these evolving BI application workloads, analytic solutions will be deployed using several different approaches, including analytic RDBMSs, non-relational systems, integrated hardware and software appliances, and cloud computing platforms. The results from BI processing will need to be delivered to multiple devices and in a variety of different formats.
Choosing the right self-service solution for the evolving and increasingly complex BI environment is not easy. Buying an integrated BI platform from a single software vendor no longer works in today’s modern BI world. One size does not fit all, and organizations must therefore develop an extended data-warehousing framework into which they can plug appropriate vendor products that enable scalable self-service BI solutions.
Copyright © BI Research, 2012