On its website, Oracle has posted the presentation (zipped, total of 71 sides) it made to the press at the Fusion update meeting last Wednesday. I took a look and here are some observations:
- As expected, lots of attention to Fusion architecture and middleware - application features finally start to show up in slide 52. Every contemporary buzzword is in the pitch from composite applications to process orchestration to activity monitoring
- Fusion Middleware brings together a number of Oracle products and concepts from grid computing to application server technology. A key point in a slide: Generally available - Now.
- Pretty good road map on what comes out by year, and fairly WIIFM for each of major customer bases (Oracle, JDE etc)
- Applications - based on best of what Oracle, PeopleSoft and JDE (and presumably Siebel). Presented some examples of features existing Oracle customers would see in new applications - rules based engine for AP matching, HR localizations for Brazil and Malaysia. Existing PeopleSoft customers would see sub-ledger accounting and HR localizations for Middle East and Korea.
- No examples of CRM, SCM or other non finance/hr features. Other than quick overview on Retail, little on any other vertical functionality. The consolidation and re-architecting of apps will mean they will primarily reflect portfolio of features circa 2005 (or earlier).
- Recommending slow, gradual migration. Step 1 - Protect investment by adopting latest non-Fusion releases. Step 2 - extend through better master data management, BI implementation and web services light applications, then Step 3 - Evolve to Fusion. Ironically, Oracle referred to Gartner chart which laments 80% of IT budgets go towards maintenance. If this gradualism (and related upgrade costs) is not "maintenance" not sure what is.
- No real discussion of migration tools or automation - though a few mentions of "Oracle and partner services" (related note: there were rumors in the Indian press that Oracle was close to acquiring an Indian services vendor)
- An ominous statement at the beginning "the following is intended for information purposes only, and may not be incorporated into any contract"
Long and short:
a) If you are competing against Oracle and think you can out SOA it, think again. Go around the edges in weak application areas - vertical functionality, emerging SCM, CRM, HRM areas. Or compete on a business model like SaaS and/or lower maintenance and implementation cost (tough to compete on license as Oracle will just about give that away).
b) If you are an existing Oracle customer seriously review the costs to get to Fusion apps - maintenance for every year till you feel apps are robust enough (post 2009) plus sizable migration costs given the demands on the services ecosystem. What you will end up with is a more current architecture, but likely little in business practice improvement (unless you consider sub-ledger accounting an improvement.)
c) if you are a new prospect evaluating Oracle v/s alternatives : Core ERP functionality (financial, hr, materials management) is solid. The architectural transition every vendor is planning will be safer with Oracle. On the negative side, Oracle will hand its hands full with migrating incumbent customers. Finally, if you are going with Oracle negotiate hard on the statement above "...may not be incorporated into any contract"
