I recently attended an executive forum on business intelligence. The audience was mostly high-level IT managers and consultants ı decision makers for a variety of Fortune 1,000 companies ı who worked in some way with Cognos Inc., which sponsored the forum. The session began with the question, "How many of you are using business intelligence within your organization?" A feeble response at best. The moderator continued, "OK, how many of you understand what ıbusiness intelligenceı means?" No hands went up. One nervy man muttered, "Whatever I want it to."
This scenario epitomizes the crisis in terminology that our industry perpetuates. Like coffee houses and housing subdivisions, new terms are coined with little regard for market saturation or English grammar. Some of these terms are reasonable expressions of more complicated technological concepts, and for those terms I am grateful. Others are coined by vendors trying to make a splash on the market. The problem is that many of these terms mean the exact same thing or imply a difference so minute that only the vendor can describe it.
Take "OLAP" as a case in point. Even back as early as the 1970s, a reasonably advanced technology for data analysis called multidimensionality existed. It survived the 1980s intact, but by the 1990s the industry was growing as was competition among vendors. A young company called Arbor Software Corp. commissioned a white paper from E.F. Codd. He modeled the paper after his previously successful "12 Rules for Relational Databases," calling it instead, "12 Rules for OLAP." Not only did this paper put Arbor on the map, but Computerworld heard of this great Codd discovery and published an article based on his research. When Computerworld discovered that Coddıs original paper had been wholly sponsored by Arbor, the publication printed a retraction. But it was too late. The term OLAP had stuck, and nobody was going to go back to plain old "multidimensionality," even though it was a far more descriptive term. It seemed that nobody could agree on what "online" actually meant. This was years before the Web provided a logical definition.
Soon other vendors wanted a piece of the market. MicroStrategy Inc. was first, popularizing a term "ROLAP," or relational OLAP, that uses metadata to present a multidimensional view of relational tables. Then came a slew of other terms Iıll call "xOLAP." We have DOLAP, or desktop OLAP, for client-side data transformations and processing (often used as a front end to RDBMSs) using tools such as BusinessObjects and Cognos PowerPlay. MOLAP, or multidimensional OLAP, emerged as a response to ROLAP. MOLAP lets you query prebuilt data cubes that have rule-based calculations; Arborıs Essbase is the most popular example. For those who still arenıt satisfied, thereıs HOLAP, or hybrid OLAP, which combines the benefits of ROLAP and MOLAP. HOLAP products include Pilot Decision Support Suite and Speedware Media/MR. And yes, there has been some mention of of WOLAP, or Web OLAP, but most people still prefer to keep the word "Web" as a separate entity and just say "Web OLAP."
Another term I fear will go the way of OLAP is "business intelligence," and thatıs why I attended the Cognos event mentioned earlier. The guest speaker was Howard Dresner of the Gartner Group, which is where the term was born. Dresnerıs explanation didnıt provide me with too much insight, however: Business intelligence refers to the notion of end users leveraging high-level data to gain perspective to make better decisions. Well that sounds nice, but from a grammatical perspective it poses problems. First, people now are saying that they "do" business intelligence, which makes as much sense as "doing" French fluency (or "architecting" a system, but thatıs a whole other ball of yarn). Furthermore, businesses as inanimate objects canıt possess intelligence ı that is reserved for the people within the businesses. There must be a better way to describe this. After much thought I believe I have the answer: data under multidimensional processing, affectionately referred to as DUMP.
We are all guilty of using often illogical industry jargon our speech, and I certainly donıt propose that we stop cold turkey. I donıt think we could. There is too much of a premium on big words. You wonıt get that big promotion for writing a press release that says, "XYZ Corp. has a new product," but you may with, "XYZ Corp. is implementing the first in a series of integrated tool suites designed to leverage advanced technology with mission-critical functionality to facilitate a paradigm shift in enterprisewide computing."
If we arenıt careful, however, we may get so swept up in this battle of verbosity that weıll become incapable of clear, concise communication. I want to help. I want to show you what this language really means, what it implies, what it covers up, and who is behind it. I also want to give you alternatives so that you arenıt forced to use these terms. I want to put you back in the driverıs seat of your own vocabulary. If you encounter terms you donıt understand, donıt be afraid to ask for clarification. Better yet, share them with me (cparkes@mfi.com) and Iıll explore them in future columns. Remember: Words have power. Learn how to use them, and use them wisely.