The Future of BI – About the Same as Yesterday

Information Management Special Reports, March 2004

Gabriel Fuchs

Look through your old piles of documents in your office basement and try to find a report that was created about 10 years ago. Take it, dust it off and have a look at it. You will most likely be able to read and quickly understand its structure, even if the meaning of the figures themselves may be forgotten. Now take a new report created recently and look at it. Is there anything in the layout that you could not have done 10 years ago? Is there anything in the new layout or appearance that is substantially different?

Whereas the technology for business intelligence (BI) has developed in a way that almost no one foresaw a decade ago, the business users' needs have not evolved with the same rapidity. Ten years ago, data warehousing was hotter than ever, together with end-user tools such as ad hoc querying and emerging OLAP techniques. Five years ago, BI started to move toward a Web-based platform where ad hoc querying and OLAP analyses could be done via the Internet. BI may even be going out of fashion replaced by corporate performance management (CPM) - of which BI is a necessary component but no longer something that stands on its own to give answers to the organization.

On the other hand, the business users' needs have not changed much in 10 years, and they want the same output from BI today that they wanted 10 or even 20 years ago when BI was still called decision support.

Look again at some of the latest developments of BI: report engines that deliver standardized preformatted reports have become standard features in addition to CPM offerings. However, report engines are nothing new. On the contrary, they have been around since the 1960s. Report engines are still viable because few users want or are capable of doing their own reports in an ad hoc manner, even when there is a nice clean data warehouse to query (which is not always the case). Very few business users do OLAP analyses. Even fewer do data mining, a technique that many - including BI vendors in the reporting and OLAP market - thought would catch on in the mid-1990s. Most business users want standardized reports and, in some cases, analyses on their desk on a regular basis and according to their specific needs. Some business users analyze and query the underlying database, but they are relatively few. According to a Forrester research analyst, in some corporations as many as 90 percent of business intelligence (BI) software users need static or parameterized reports, not analytics.¹

Nothing is done in an organization without reports, whether it is the presentation of analysis, balanced scorecard or corporate performance management. The underlying technology that produces these reports may certainly have changed in the last 10 years, but the outcome looks very much the same today as it did then.

What Do Most BI Users Want?

People are rather easy to understand in the sense that most want an easy life, i.e., "more for less" or "more with less." Had we not been like that, we would not have evolved from the hardship of living in caves and hunting wild and dangerous animals with wooden spears. When it comes to BI, it seems that business users still want life as simple as it can be. Some of the following requirements bear this out.

Standardized reports and analytics created automatically. Most BI users try to standardize their reports. Once defined, the report can be automatically reproduced at regular intervals, with little or no effort by the business user. If there is no standard, comparisons between information in older and newer reports is nearly impossible. Without standardization, each time a comparison is needed, a completely new report is created instead of simply comparing the figures of an older report with the latest one.

Spreadsheets. Almost everyone likes to work with what is known, and what is known to BI users today is spreadsheets. Microsoft claims to have 150,000,000 Excel business users, making Excel the most widespread report and analysis tool in the world. No matter what one may think about all the spreadsheets in an organization and the impossible task of coordinating all the information, Excel is a reality. In practice, it is difficult to replace the spreadsheets with another tool. The good thing about the spreadsheet format is that it has actually set a standard that most business users use, no matter what other BI tools they may have as well. When implementing new BI solutions, Excel is very often a reference for the concerned business users.

Paper. Another thing that the business users seems to want is paper, i.e., the possibility to print their reports and analyses. They may be selective on what they print, only choosing their specific concerns, but they will print reports. A report tool that has poor printing capabilities will be disliked from the beginning. The idea of the paperless office has not materialized as forecasted years ago. On the contrary, we produce more paper than ever, and we continue to do so because it is easier to study and discuss paper reports during meetings than reports in electronic format.

Color and charts.What has become popular again lately is a focus on getting an overview through color indicators and charts. This was popular in the 1980s with the executive information system (EIS) solutions. These solutions used vivid colors and charts, but the underlying architecture feeding the system with data was in most cases difficult to maintain and they weren't widely accepted. Modifications in the EIS system to adapt to changing business needs became too time-consuming and resource-intensive. Furthermore, it was difficult to change the presentation without involving IT personnel. In any case, color-based indicators and charts have once again found their place because today's tools and underlying architectures are much more flexible than 15 years ago. Solutions and methods such as CPM and balanced scorecard are certainly capitalizing on visualization when communicating their results to the business community.

Detailed data. The one thing that hasn't changed is the desire for access to detailed data. A chart can give an excellent overview, but whenever there are anomalies or  interesting phenomenon, the underlying detail is asked for. Furthermore, this detail is wanted in a classical report layout, i.e., a table or spreadsheet format.

Accurate, accessible, timely and comprehensible information. Obviously, all the information asked for by the users has to be reasonably accurate (where the definition of "reasonably" is decided by the given circumstances). The reports and analysis should also be easy to access in a timely manner. It is obvious that the Web-based BI platform has been tremendously successful in improving these aspects. Finally, the presented information should also be comprehensible. Forget about flashy 3-D charts presenting six data dimensions unless your users are highly accustomed to such things. In most cases, tables and spreadsheets, and bar, line and pie charts are the methods used to display information by popular demand.

Today's underlying BI technology is more adaptable than ever to allow for distributing the reports and analyses, with better accessibility and with more frequent updates. However, do not be surprised by what is often a basic layout of the distributed reports - which often seems to be stuck in the past and done with a spreadsheet program.

The Future of Users' BI Needs

Few predictions about BI and strategic IT usage, in general, tend to hit the mark. Even if they succeed in foreseeing something that will actually happen, these predictions tend to be too optimistic when it comes to the timing of the fulfillment. However, given the surprisingly few developments in BI users' needs in the last 10-15 years, the following forecasts about their future BI needs can be made:

• Spreadsheets, notably Microsoft's Excel, will stay with us for BI purposes for the foreseeable future. Many BI vendors seem to agree on this as the integration between Excel and many specialized BI tools is improving rather than diminishing.
• The continued usage of Excel means that the classic table or spreadsheet layout with its exact figures and decimals will still be in demand, no matter how advanced cross-tables or charts a BI tool can deliver. Reports will continue to look very much the same.
• The usage of colors and charts will advance. Access to data, the performance of personal computers and the flexibility of today's BI tools have all improved. Therefore, a practical usage of such interfaces is possible, and the fate of the old EIS applications can be avoided. Major development for specialized data visualization tools on the market will cause an increase in use. The balanced scorecard is already basing its presentations on charts and color indicators, as are CPM interfaces.
• An improved overview of data also puts pressure on the link between the high-level overview and the always-in-demand underlying detail. Therefore, it is likely that drill-down and drill-up functionalities in BI solutions will be further developed in order to allow a quicker, more intuitive and better-integrated link between overview and detail.
• Even though the business users' BI needs are likely to remain relatively stable when it comes to actual usage, the underlying data management providing information in a correct and timely manner is likely to develop further. Data management is becoming more complex due to constantly growing volumes of data. This puts an added burden on data quality, data structure and data access issues when implementing BI solutions. As a result, constant and further improvements in database technology and middleware with the purpose of providing more accurate information more often - sometimes in real time - is needed.
• Finally, the success of BI comes down to the actual usage, a major factor that is often neglected when distributing a BI solution within an organization. Few, if any, BI investments will give the expected return of investment unless the application is actually used. Every unused license is a cost that gives no return. An increased awareness is necessary to have a positive return on investment that will push and improve the actual BI usage, if not necessarily increasing the number of sold licenses. Auditing actual usage has not become more complex or difficult over the years.

As noted, these are just predictions, but the look of reports as well as their usage have not changed in a revolutionary way in the last 10 or 20 years. Furthermore, the creation of these reports should be as easy and quick as possible in order to please the majority of business users, something that is normally achieved through automation and standardization. Ad hoc reporting is still for the power users, which even some BI vendors will acknowledge these days. Business Objects did so in a press release when they announced their acquisition of Crystal Decisions saying, "Business Objects has long served the 'power users' in organizations..."²

Furthermore, most business users do still not know what OLAP stands for, even though it has been around for more than 10 years, no matter how useful and powerful multidimensional analyses are to the business.

 Ad hoc reporting, OLAP and data mining certainly have their place in most organizations, but not necessarily on a scale that was forecasted by many analysts and that may still be suggested by vendors. Even if user demands may seem simple, providing what they want can mean substantial time savings in their work coupled with better decisions that will lead to a positive return on investment.

Simply put, business users' needs do not become more complicated or advanced just because the end-user tools do.

References:

1 Russom, Philip.  Forrester Research. Intelligent Enterprise. August 12, 2002.

2 "Business Objects Announces Agreement to Acquire Crystal Decisions."  San Jose and Palo Alto, California and Paris, France. July 18, 2003.

Gabriel Fuchs is a senior consultant and business intelligence expert. His column Reality IT takes an ironic look at what real-world IT solutions often look like - for better or for worse. The ideas and thoughts expressed in this column are based on Fuchs' own personal experience and imagination and do not reflect the situation at any particular company. His book, Dealing with Nasty Colleagues: The Art of Winning in Office Politics While Still Getting the Job Done, can be ordered at www.amazon.co.uk. He can be reached at sgfuchs@bluewin.ch.

For more information on related topics, visit the following channels:

• Business Intelligence (BI)