IBM recently announced that it will no longer sell IBM DB2 OLAP Server, an OEM version of Hyperion Solutions' Essbase product. This announcement heralds a potentially major change in the competitive landscape of the database world, regardless of whether you think an OLAP server is in your own short-term future.
Here's some background. IBM has been reselling Essbase as DB2 OLAP Server for about 10 years. Today, IBM has several relational business intelligence (BI) technologies built into IBM DB2 that don't come from Hyperion. Although DB2 doesn't offer a full-featured OLAP solution like those that Microsoft and Oracle offer directly to their customers, IBM's announcement means that DB2 customers must now shop outside of IBM if they're interested in buying a full-featured OLAP server.
That's not necessarily a major problem. After all, existing DB2 OLAP Server customers can simply move to the Essbase line that's still the flagship product of Hyperion. But for some time, the trend in the database space has been to embed more and more features and capabilities directly into the core database platform. For example, Microsoft has made significant investments in its extraction, transformation, and loading (ETL), reporting, and analytic tools in recent years. Oracle and IBM have both made significant investments in adding competing enabling and integration technologies to their respective database platforms. Gartner and other industry analysts argue that IBM is now at a competitive disadvantage to Microsoft and Oracle because IBM is the only "Big Three" database vendor not currently capable of offering a full OLAP solution as part of its core product. I agree.
What does this mean? Will DB2 users shun IBM if they need a powerful OLAP solution? That's not likely in the short term because the migration to Hyperion will be seamless. But let's be honest, when you buy your OLAP server from the vendor who provides your core relational database, your decision-making process is completely different from the process of choosing to buy from a third party, and users will wonder why IBM chose to sever the relationship. Hyperion's relationship with IBM currently accounts for a significant part of Hyperion's revenue, directly and indirectly. Oracle and Microsoft both have massive war chests to finance their OLAP engines. One has to wonder whether the smaller Hyperion has the cash flow and revenue to make R&D investments at the same level as Oracle and Microsoft. I suspect that IBM's announcement will lead many Essbase and DB2 OLAP Server customers to evaluate other OLAP solutions. Microsoft is particularly well positioned to attack this market because its price points for SQL Server are relatively low and Analysis Services can point to any relational engine seamlessly.
One also has to wonder about IBM's true long-term plans for OLAP. Current statements from IBM attempt to downplay the importance of having an integrated OLAP solution. However, many of Microsoft largest gains in the database market have been from customers who buy SQL Server for its complete feature set, not just the relational engine. Customers love the fact that Microsoft offers compelling reporting, OLAP, and ETL solutions in the box. Honestly, I can't imagine that IBM truly believes it can compete in the database market over the long term without a built-in OLAP solution. Will IBM build or buy? I wonder. Either way, IBM will be getting a late start behind OLAP leaders Microsoft and Oracle.
For more insight into the IBM decision, I encourage you to read the Enterprise Systems news story "IBM and Hyperion Call It Quits" by Stephen Swoyer at http://lists.sqlmag.com/t?ctl=11531:7B3DA . AMR Research provides an analytical article by John Hagerty, "IBM and Hyperion End Embedded OLAP Relationship," at http://lists.sqlmag.com/t?ctl=11534:7B3DA . You'll need to register to read the AMR report, but it's free.
