AC – Achieving Enterprises AgilityPosted: November 5, 2004
Shai Agassi – exec board member SAP. In SAP, 1/2 the people in the company are sales people. Talking about a mega billion market (enterprise applications). About 1/2 of the transactions in the industrialized world go through SAP systems. Mentioned about the HBS article “IT Doesn’t Matter” He de-bunked the article (of course :-). Before client/server, you couldn’t run a company in a 100 places at once. SAP was one of the first companies to move from the mainframe to client/server. One of the toughest things about bigger companies is change–it’s hard to change. Example, a merger will typically take 18 months. Used to have eco systems that changed once or twice every decade. Now they happen much faster.
Jeffery Moore’s definition for company’s processing (Living on the fault line). Differentiation life-cycle: invention > innovation > standardization > commoditization. Example was airline check-in kiosks, ’99, 02, 04, ? commditization. IT capabilities required for this evolution life-cycle.
The presenter had a tablet PC. Started writing on his PowerPoint. Very effect to communicate extra detail in a preso. Then MS PowerPoint crashes. Unbelievable!
Talked about SAP extending their stuff via web-services. $100B market in the consumption of information. Expect to see some incredible innovation.
- “Event Centricity.” In the future, things are going to be event driven. No more opening up an app and the app waiting for something to happen.
- SAP’s problem, how to watch and react to millions/billions of separate events.
- Every company is now putting RFID into their business==and example of “billions of events.”
- Exception handlers are people that handle events that the automated cycle can’t.
- Mobility is going to be big. 1/2 terabit. Cache most of the objects you deal with will be on your hip at all times. That will be the center of the universe. When you get close to the desktop, then the desktop takes over. The desktop will be secondary to that device.
- Moving away from maintaining code and towards models. You can watch events as they flow through a system/model and make changes. Moving away from Java to domain specific languages (like modeling languages). In hardware, going from server farms to grids.