mapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmapsmaps

Business Intelligence Mega Trends 2008

Writing by Jorgen on Wednesday, 16 of January , 2008 at 8:56 am

Times they are changing. Last year has been tumultuous for BI at best. We have seen the BI market consolidate into a few remaining mega vendors. But there was also room for niche players especially the ones focusing on speed, visualization or collaboration. So what will 2008 bring? To quote from Star Wars: “Impossible to see the future is”. But let’s give a try anyway. There are five mega trends that I foresee. Some have been around for quite some time and some are new. The first one is called: (1) Increase: Business Intelligence must be able to cope with considerable increase in both data (size) and speed (real time). The second one is (2) Integration: Business Intelligence will continue to integrate with operational processes but will also be more aligned with strategy. The integration of (un)structured data will increase. Finally, Business Intelligence will integrate with other business applications or ERP systems. Third, and my personal favorite, is (3) Insight: Business Intelligence will continue to become more intuitive (less end user involvement) showing an increase in advanced visualizations but also more analytic (more end user involvement).The fourth is (4) Interactive: Business Intelligence will be more reciprocal by nature; creating a dialogue between system and end users as well as sharing information and collaborating within performance networks. Finally, number five will be (5) Industrialization: Business Intelligence development will be characterized by a more industrialized approach using standard development frameworks opening up the possibility to outsource.  Increase will be mostly characterized by the need for real time information (partly due to the fact that BI becomes more operational). This will be supported by tools like Netezza and Teradata. They can also support the growing process of data retrieval from vast amounts of data. Integration of unstructured data (text, search, documents, agents) for example in a document warehouse (using tags) sitting next to the data warehouse will be the wave of the future. I expect Google or Attunity to take the lead here but the original BI vendors (Cognos, BusinessObjects) have also been investing in this. Although the market is more or less divided among a few remaining vendors it is still possible that we will see some more integration. This will be partly triggered by the fact that BI will be used on all levels: strategic, tactical and operational. This means BI tooling must support strategy development as well as pervasive BI (it’s everywhere). The renewed interest in open source Business Intelligence (Pentaho, Jaspersoft) is a clear counter reaction against the market domination due to vendor consolidation. Creating Insight today will be the one mega trend where I expect the most innovation to come from. It is a hot topic and fun to watch evolve. Here you will find things like intuitive intelligence (Strategy companion) or decision engines supporting your daily work (Fair Isaac). Also advanced visualization will help the end user get more grip (fractal edge) on things. Triggered by the success of “competing on analytics” or “super crunchers” we will see an rise in the use of analytics, strongly combining informatics, statistics and business knowledge. Perhaps here lies the opportunity for the long awaited promise of data mining? A clear and much appraised trend is the sharing of information with other people either inside or outside your own company to create mutual benefit. But it is more than just sharing the information it is also about sharing the reasoning behind it. How have we come to this conclusion and do you agree? It is asking feedback or contributions to your problem solving by your stake holders. All the characteristics of the web 2.0 can also apply here. This can also be done by using a trusted third party to solve your BI wishes by providing BI as a service. As BI is getting more mature by the minute I expect that we will see a standardization of best practices resulting in an industrialized approach in (standard) BI development. BI factories, often based in typical outsourcing countries, can provide high quality, low cost ready-to-use BI solutions. Some people say that after all the consolidating this will be a boring year for Business Intelligence. But if I look at these five mega trends I can’t wait. Happy new Year.   

Category: Analytics, BI Thoughts, BI Tools, BI vendor consolidation, Business Intelligence consulting, Business Intelligence datawarehousing, Business Intelligence software, Business Intelligence solution, Business Intelligence strategy, Business Intelligence system, Business Intelligence tools

3 Comments

Comment by James Taylor

Made Wednesday, 16 of January , 2008 at 7:12 pm

An interesting set of trends. The Integration, Insight and Industrialization trends are going to be a challenge for those that think BI means reporting and OLAP. Integrating insight into high-performance operational systems on an industrial scale takes a different approach. You need to use business rules and executable analytics, not reporting tools and queries, to deliver on this promise.
JT

James Taylor
Author, with Neil Raden, of Smart (Enough) Systems
Blog at www.smartenoughsystems.com/wp

Pingback by links for 2008-01-17 | Smart (Enough) Systems, the blog

Made Thursday, 17 of January , 2008 at 9:32 am

[…] Business Intelligence Mega-Trends Interesting set of 5 trends all beginning with I. Several of them reinforce the value of extending an old-style BI approach to include EDM. (tags: BI 2.0 trends edm) […]

Comment by Dave Dixon

Made Thursday, 17 of January , 2008 at 8:01 pm

Nice analysis. Allow me to put my own spin:

1) The bottleneck in dealing with information (whether you’re talking about volume or speed) is not technology, but the person in the loop. People are poor at assessing a lot of info, instead they tend to pick a couple of things that seem like “big hitters” and place too much emphasis there. That choice is usually arbitrary, and the more choices available, the slower decisions get made. The solution is to allow the computer to value available choices given the available information, i.e. to build decision models. People are then freed to do what they do best (be creative) while the computer does what it is best at (quickly process lots of information). Blogged about it here: http://blog.provisdom.com/?p=32.

2) The only way BI can be more aligned with strategy is if you can value information as it impacts the strategy. The only way you can value the impact of information on strategy is if the strategy is chosen to maximize a goal that has a well-defined value. You might say your goal is to build the “best” product, but unless you can quantify “best” in dollars and cents, it’s hard to place a value on the associated information. Ultimately, the only goal of a public corporation is to maximize shareholder value, so such alignment requires that you can measure the change in shareholder value brought about by some information in the context of relevant decisions. Here’s a case study discussing the value of information in the context of petroleum production: http://blog.provisdom.com/?p=27.

3) I agree that BI needs to be more intuitive, but that intuition should operate at the strategic level, rather than at the level of “understanding” vast amounts of interrelated data. Visualizations and analytics have a place, but I believe it mostly to be one of informing decision-makers of possible choices, e.g. market segmentation provides you choices of segments which you can market too, information about what sort of marketing strategy will be effective in each, etc. But no analytics or visualizations will ever allow you to get your head around all of the available data without losing information (oversimplifying). Such information loss incurs a corresponding loss in shareholder value. And providing data in real-time doesn’t necessarily mean people can make effective decisions more quickly; it just swamps them in more information. Again, the actual process of valuing choices given information is best done via modeling (http://blog.provisdom.com/?p=26). The computer should be able to provide decisions in real time, given data supplied in real time. The role of the human is to ensure that the available choices and information continue to be relevant, make sense in an ever-changing business environment, etc.

(4) and (5) I pretty much agree with, though again both items must be done in the context of making decisions which maximize shareholder value.

Leave a comment

You must be logged in to post a comment.

Author

Jorgen Heizenberg is Principal Technology Officer for the Business Intelligence domain at Capgemini Netherlands. The views expressed in this blog accurately reflect his personal views about any or all of the subjects and is not part of the official Capgemini company view. PLEASE REACT TO HIS OPINIONS AND BECOME AN ONLINE BI GURU. See also: http://www.beyenetwork.nl/blogs/heizenberg/