There are two key questions all the leading information technology (IT) providers are asking themselves in the current industry environment:
- Does the company want to be a technology provider or a services provider (using an auto industry analogy, does it want to make parts like Meridian, or does it want to assemble and sell cars wholesale like Toyota?)
- If a services provider, does it want to service IT departments or end-users?
Oracle had previously been inconsistent in its intentions relative to the technology provider vs. services provider question, but in my opinion Oracle President Chuck Phillipsâ€™ recent presentation (April 17 about â€œOn Demandâ€) gave a strong tip as to where Oracle will come down. He said Oracle was already delivering database, middleware, and application products via Software as a Service (SaaS). Most analysis refers only of applications when it mentions SaaS. Oracleâ€™s delivery via SaaS takes strong advantage of Oracleâ€™s grid computing technology, which lets Oracle deliver multiple users (multi-tenant) support for multiple segregated databases (to avoid commingled data among tenants). Some of Oracleâ€™s grid computing and database features let Oracle do that more efficiently and, presumably, more profitably than a services provider using other server-farm topologies and less parallelized databases. I take issue with Oracleâ€™s claim that no one else can keep the data segregated but Oracle has a technical advantage. In addition, Oracle is making changes to its applications to permit them to be segregated as well. Currently we estimate that the On Demand portion of the business is fairly small and, because most of it stems from the Siebel acquisition, very little prior to this quarter was grid-based. The April 17 announcement signals a shift and should lead to a stronger future SaaS revenue flow. Another thing investors should watch for, in addition to the absolute growth in that revenue flow, is how the revenue is accounted for. As Oracleâ€™s SaaS revenue flow grows, Oracle will go through a process, such as the one CA is just completing, where new-model revenue becomes a larger and larger percentage of Oracleâ€™s revenue total. That new model will cause revenue flow to be smoothed out rather than recognized in the upfront manner that the supplier uses to account for â€œold-modelâ€ perpetual right to use license revenue. This will temporarily depress Oracle growth rates vs. what they otherwise would have been beginning as soon as its next fiscal year (which starts on June 1, 2007) given the same amount of sales activity. But analyzing that accounting issue is a good thing given that it means Oracle wants to make cars and not mufflers. A more detailed analysis of Oracleâ€™s April 17 financial-analyst meeting is included in the Research 2.0 April monthly.Â — Dennis Byron