Business intelligence (BI) has matured to such a point that, particularly for larger corporations, it is almost universally accepted as an important part of overall IT strategy. In the industry over the past several years, there has been consolidation and a commoditization of extraction, transformation, and loading (ETL) tools. The major enterprise resource planning (ERP) firms have attempted to integrate BI into their massive framework systems and, to a great extent, have succeeded. But the idea that a single vendor will provide a single, integrated, corporate-wide BI suite is beginning to fade, as companies recognize that different areas require different approaches. This has led to a decrease in ad hoc solutions, greater integration and improved standardization between systems, and growing popularity of “executive dashboard” or “enterprise portal” solutions for bringing data analysis from different sources to the users that require it in a manner suiting their understanding, capabilities, and requirements.
In addition to these trends, BI analysis continues to grow in sophistication and technical requirements and is moving further into the enterprise. Real-time analysis, including business performance management (BPM), is creating new technical challenges and also provides new opportunities. At the same time, enterprises are taking a closer look at costs and requiring clearer justification for the introduction of new BI tools. This has helped to create consolidation in the market.
Technical Developments
Several trends are of fundamental importance to the technical evolution of BI. The first is the move toward real-time analysis based on real-time input. This is bolstered by the move toward business process analysis, where analysis is used for direct improvement of ongoing processes, such as profiling buyers and providing immediate solutions for them. This puts pressure on older data warehousing-based solutions, in which simple OLAP analysis is applied in off-hours through a batch process. Although these older systems are still necessary to handle the diversity of data available in the enterprise, they must coexist with real-time analytic systems.
Another trend is the move toward predictive analysis using sophisticated algorithms based on artificial intelligence and neural nets. Real-time, predictive analysis can provide obvious and immediate advantages to firms, particularly in areas such as sales or order fulfillment. Such analysis, however, requires greater sophistication and is likely to be applied only to a small portion of data, using systems designed by niche firms, thus providing added support for a multiple BI system scenario.
The growing level of sophistication in analysis is raising infrastructure requirements. Increasing processing loads required by real-time analysis and geometrical progression of available data have led to a renewed emphasis on processing capabilities and efficiencies. Even with the continued yearly doubling of systems resources, new strategies are required to ensure that data can be handled in a timely and adequate manner. The latest range of database, analytic, and data warehousing solutions includes multithreading and grid computing, as evidenced by products such as Oracle’s Database 10g and SAS’s SAS9 Intelligence Platform.
Another technical development reflects a growing recognition that “analysts” are not the only candidates for BI use. As the technology continues to infiltrate the enterprise, an increasing array of different operators with different capabilities and requirements are using BI. This has led to improvements in user interfaces, simplification of choices, and provision of different levels of presentation. The result is development of executive dashboard or enterprise portal solutions.
Analysis Trends
Increasing use of BI combined with continued development in technical sophistication has led to an expanding choice of applications. In addition to movement into a greater array of vertical niches, the most important trend is a movement of BI into fundamental business areas. Performance management is the most important of these areas. BI is becoming increasingly aligned with BPM. BPM is a process that begins with defining strategic goals and objectives; plans are created for tactical execution, and all activities are monitored and assessed to readjust plans. BI provides a powerful tool in assessing plans and overall results. This requires, however, sophisticated, real-time analysis of data. Capability to define subtle adjustments in operations to meet objectives makes BI an important strategic asset and provides a competitive advantage.
BPM is becoming an important competitive strategy for corporations, and this is likely to result in further expansion of BI technology. A November 2004 Cutter Consortium survey of 110 end-user organizations found that 39% of surveyed organizations have a BPM initiative currently underway, and another 26% plan to launch one within the next six to 12 months [2].
As BI is applied more universally, it is also employed with greater depth. Simple analysis of trends based on limited data are giving way to rich analysis of a large amount of online data, leading to “rich profiles.” Rich profiles are most visible in online shopping applications, where Amazon.com has become the leader in usage of advanced analytic technologies on data drawn from extensive monitoring of real-time customer interactions.
An important technical challenge that remains for BI is analysis of unstructured information. Unstructured information — documents and communications, such as e-mail — constitutes the final frontier of business data analysis. While only a few products provide analysis in this area, it is by no means prevalent. The combination of requirements for retaining data and liability legislation affecting e-mail has made this a growth area. Breakthrough products in this area are to be expected in the near future.
Implementation Trends
In implementation of BI, the value and profile of stored information has recently been raised by legislation such as the US Sarbanes-Oxley Act, requiring retention of information such as messages. The need to retain information leads to a desire to use it, and data mining moves quickly to data analysis and BI in the quest to leverage these information stores.
In August 2004, Unisys commissioned KRC Research to conduct a study of BI trends. The study surveyed decision makers at large US enterprises (those with a minimum $500 million in annual revenue). The following are some of the key findings of the study [1]:
- Almost all organizations with BI applications now seek to standardize reporting on a single platform.
- Companies are deploying multiple BI applications across their organizations, with one-half using seven or more and one-third using 11 or more.
- Risk management is arising as a major driver of BI implementation.
- Revenue maximization, cost reduction, and regulatory compliance remain key drivers.
- Half of surveyed organizations consider BI an important and integrated part of their IT strategy.
- Information security risks have become an important consideration in designing a BI strategy.
- BI has now stretched out across the enterprise, with no single purpose dominating the market. Major uses include management of corporate performance, monitoring business activity, reporting or regulatory compliance, and customer analysis.
Another important implementation trend is the growing need to establish ROI and cost justification for BI systems. This results from an atmosphere of increased concern over capital expenditure combined with a realization that today’s BI solutions can be expensive to deploy — and experimental systems or systems that do not provide valuable information, even more so.
Market Trends
Recent recessionary trends have led to reluctance of firms to spend capital, resulting in longer product replacement cycles. While BI has been less affected by economic slowdown than other areas, longer cycles have led to reduced revenues, resulting in market consolidation. Consolidation is also encouraged by the need to expand offerings to compete, both among the major framework vendors and among niche operators. The largest recent consolidations have been Business Objects’ acquisition of Crystal Decisions, Hyperion’s acquisition of Brio, IBM’s acquisition of AlphaBlox. Further consolidation is likely to occur over the next several years.
Another market trend is toward outsourcing, particularly in the ETL area, with most of the business going to Indian firms. As these firms develop in sophistication and experience, they are likely to move up the food chain toward more comprehensive BI solutions. To date, movement in this direction has been relatively slow, partly due to the fact that outsourcing still makes real-time analysis more difficult. However, improved communications, economies of scale, and growing sophistication of offshore providers are all likely to encourage this trend.
Conclusion
The BI market and technology are both maturing, and this has resulted in consolidation and some standardization among solutions. BI is moving forward in analytic capability and becoming further integrated with business processes. A trend is also emerging toward incorporating a diversity of BI solutions across the enterprise. A dashboard or portal that can be customized to meet the specific information needs and capabilities of the user will integrate these diverse solutions. As analysis grows in sophistication and becomes more pervasive, it will put more demands on infrastructure, particularly in the areas of networking and processing. Grid computing solutions are becoming more common, and the constant transfer of real-time data will also put pressure on networks — which are already being readied for multimedia and increased bandwidth applications.
As with most computing solutions, it is important to remember that the implementation environment for BI is extremely uneven. Few companies are able to employ the very latest solutions and technologies across the board. The majority of BI solutions implemented today provide only basic capability. This leaves plenty of room for growth, even as the technology itself matures.
References
1. “Business Intelligence Trends Survey.” KRC Research and Unisys, 16 August 2004.
2. Hall, Curt. “Business Performance Management: Trends and Directions. Cutter Consortium Business Intelligence Executive Update, Vol. 4, No. 20, 2004.
