Microsoftıs entry into the OLAP market is good news for end users. For developers and managers, it signals a new growth phase for OLAP technology. And itıs yet another example of how Microsoft can change any market it enters ı with good and not-so-good effects.
In an industry where vendors scream about being the first product to butter both sides of your bread simultaneously, innovation is not always Microsoftıs path to profits. Despite a large and growing research and development effort, Microsoft often patiently waits for smaller bleeding-edge companies to open and groom a market. Microsoft then swoops in and makes the technology a mass-market application. OLAP is certainly not new ı or even leading edge. Multidimensional products date back more than 20 years. Itıs a proven technology that so far has been mostly a niche application used in large companies with sophisticated staffs. Microsoft will undoubtedly change that by popularizing OLAP.
Microsoft can enlarge the size and scope of a market in several ways. Some companies, especially smaller and midsized companies (but also some divisions within larger companies), will think that if Microsoft is selling a technology, itıs something they should evaluate and adopt (or wait for version 3, as the old joke goes). In addition to Microsoftıs considerable sales and marketing forces, the legion of Microsoft-centric developers will soon add Microsoftıs OLAP Server to their arsenals, and this will fuel a bottom-up expansion of the overall market. Once OLAP opens the door, the rest of the data warehouse industry will rush in. Competition is a good thing.
To the degree that a rising tide lifts all boats, an expanding OLAP market may actually help other vendors and indirectly help users too. Microsoft will consume a large portion of the market, but not 100 percent. Vendors directly in Microsoftıs path ı and thatıs a wide swath ı will face the most serious threats. But other vendors will find supporting roles Microsoft is not yet gobbling up. Yet others will try to compete based on features and innovations. Over time, this will raise the overall level of capabilities as Microsoft and other vendors match and leapfrog features.
Pricing is another critical factor. Most OLAP servers are not cheap. Whether Microsoft bundles its OLAP Server into BackOffice at no extra charge or sells it as a separate product, it will most likely drive down the average price of OLAP servers and clients. While Excel has long been an OLAP client, Microsoft will begin heavily promoting Excel as its OLAP client. As users gain one more reason to stay in Excel, it becomes harder for other vendors to wrestle them away. Microsoft can easily ride out a price war. Many smaller vendors have trouble sustaining development, support, and sales efforts when previously high prices come down to earth.
Microsoftıs moves will also accelerate the rise of Windows NT as an OLAP platform. Very large OLAP systems will continue to run on Unix, but NTıs momentum is also attracting many decision-support developers, especially for data marts.
Beyond OLAP, itıs interesting to note what forms of specialized data management Microsoft is still waiting out. I can think of two that are probable next steps for Microsoft, although I have no idea when Redmond will reach out. The two datatypes are textbases (including documents) and geographic or spatial data. Both text and spatial data are complex datatypes that do not fit neatly into plain RDBMSs. Both are ubiquitous; every company must manage boatloads of text including HTML, and almost all systems include addresses, the heart of geographic information systems for business. Like OLAP, text and spatial DBMSs have been around for well over a decade, but their adoption remains the exception rather than the norm. Each market has well-established but relatively small vendors. And both are not so distant from RDBMSs that further extensions to SQL Server are out of the question. At some point, I think Microsoft will set its sights on one or both of these potentially lucrative database spheres.
Microsoftıs OLAP Server will cause casualties in the OLAP market. Some vendors may simply go under, leaving users stranded. Others will make tasty acquisitions for vendors that lack an OLAP offering. Much of this disruption will not be pretty. And a few vendors will reposition themselves into safer niches. But having used OLAP myself, I know itıs a valuable analytical technology. By boosting the overall OLAP market, those vendors agile enough to find a role will continue to prosper and serve users. And many more end users will gain the benefits that OLAP can deliver. Thatıs what weıre here for, right?