Oracle's dependence on its database management software business as cash cow, although important, does not even make the top 5 of Oracle's top 25 IT Investment Research facts I have outlined in my analysis. What is more important to understand about Oracle revenue flow, as hidden in footnotes and cross references in the Oracle July 21, 2006 SEC 10K filing is that Oracle's real revenue growth for the past few fiscal years has been in the single digits despite (or maybe because of) its multiple acquisitions. This Oracle IT Investment Research Finding is ranked number one in importance in my analysis. When acquisition accounting rules no longer mask Oracle's below average performance (vs. the IT Top 12 as a group), the company is going to have to ratchet up performance with real “new" business or suffer among the IT investment community. 

Another factor for investors to consider is the implication of the number of application architectures Oracle supports (eight), compared to other relevant IT Top 12 suppliers (two for SAP and Intuit, and four for Microsoft). Because “Oracle has a highly diverse application product line architecturally” (Oracle IT Investment Research Finding 2), it faces much higher research and development (R&D) and ongoing product license update and support costs than these competitors. In addition, Oracle began to promote a two-architecture approach to deployment middleware in June 2006 before the release of their FY 2006 10K; I believe Oracle already (as of July 31, 2006) realizes that such a marketing message is self defeating vis a vis the whole Fusion message but if a two-architecture middleware code set is a technical as well as marketing reality (that is, one based on Oracle's Ironflare-heritage J2EE container and one based on its Sandia-Labs-heritage Rules Engine), Oracle will face increased R&D and support costs in this area as well. I don't consider the application architecture diversity a show-stopping issue because Oracle knows how to do integrate application functionality and infrastructure efficiently and effectively. Oracle bought Datalogix Gemms in 1996 and executed the same kind of integration it now needs to accomplish Vis a Vis PeopleSoft, Siebel and so forth over the period 1997-2001.  That it might take four years to accomplish this larger task is not surprising but Oracle's implying that the company is close to completion of the activity—makes this an issue in the market.