Thinksum

Political Turf

So Shai Agassi left SAP ... I'm sure he was disappointed in Netweaver's lack of success.  Sure SAP customers may want to use it, but relevance beyond that is nil.  Why do application vendors ever think they're going to make their own proprietary stuff into a hugely successful platform (hint hint salesforce.com)?  It's like when Microsoft thought it could take over programming languages with C#.  Self-delusion.

One interesting thing about SAP is the culture clash between the German parent and the platform team in California.  At my last company, about two years ago we entered into serious discussions with SAP about being acquired (we ended up being acquired by someone else).  I participated in due diligence with folks from SAP Germany and also SAP in Palo Alto.  It was clear to those of us who participated that there was an uncomfortable relationship between the groups.  The Palo Alto people considered themselves smarter (I'm not sure they were), but the folks from Germany were clearly in charge.  Every meeting was fraught with internal politics, all the SAP people being very careful about what they would say in front of others, and saying different things to us when we were alone in rooms.

I never met Shai Agassi, but from what I could tell he seemed more like a politician than a software person (which he kind of purported to be in leading the SAP platform vision).  I'm not surprised that political involvement is part of his future plan.

That's why I like start-ups and small companies.  Less politics, less hassles.

A New Kind of Software

Stephen Wolfram released his book A New Kind of Science a few years ago (as an introduction to this area of thinking you might want to read Steven Levy's accessible Artificial Life). Wolfram's premise is that while scientists have long looked to explain nature in terms of mathematical rules, in some if not all cases it might better be described as a set of "programs" akin to computer programs. In addition, these programs need not be complex but can in fact be extremely simple. A key belief Wolfram espouses, and attempts to prove, is that the vast complexity of nature and the universe can in fact be "generated" by a finite set of simple programs.

In the world of enterprise applications we are in the midst of a transformation not completely dissimilar to the one Wolfram describes in science. Enterprise applications have tried to solve complex business problems with complex solutions and large packages of functionality. But now that is shifting, and standardized integration and inter-process communications have created the opportunity to solve complex business problems by combining a finite set of simple components.

For people who believe in the new kind of science that Wolfram describes, the quest is to "discover" the programs at work in nature. For people who believe in the new kind of software, the quest is to "discover" the individual components that can be used and reused to build a more powerful set of enterprise applications.

Enterprise Applications Are A'Changin'

Enterprise applications as we've known them are dying, they just don't know it yet. The age of large sets of tightly coupled functionality is coming to an end. The failures of this model have been identified: hard to customize, hard to change, and even harder to change as time goes on. So far, the accepted model for enterprise applications has been that you spend a lot of money on the application, you spend a lot of money and time on the customization, and then you hope and pray that you did it right because you won't be making many changes for a long time -- unless you want to keep spending a lot of money and time, which is what a lot of companies do.

The age of loosely coupled pieces of functionality and processes has begun. This is the end of behemoth applications, replaced by smaller modules that can be linked and combined in a multitude of ways. These modules are tools that can be used and applied creatively by business owners, much in the same way we use a word processor or a spreadsheet today. This is the beginning of putting process and change control into the hands of business owners, as Phil Wainewright discusses here, and removing IT as a costly bottleneck in process improvement. It won't all change in a day or a year, but word's out, the alternatives are appearing, and it's just a matter of time.