Apply today for a FREE subscription to CIO Magazine!
Wed, May 10, 2006 10:08 EDT
|
Posted by: Christopher Koch Blog: Koch's IT Strategy
Current Rating: |
There's a new front developing in the war between business and IT. Let's call it BPM (Business Process Management) versus SOA (Service Oriented Architecture).
To understand the point of conflict clearly, however, requires a look back at history.
You can go back as far as you want, but I think everything before the business process reengineering craze of the late 80s, early 90s should be considered prehistory. Here's why: The reengineering craze was the awakening of the business to the fact that their ways of doing things were buried inside functional computer systems that they didn't control.
Like many revolutionary concepts, however, reengineering had limited practical value. You can draw as many process diagrams as you want--and the consulting industry made millions leading these futile exercises in conference rooms all over the world--but if those diagrams don't address the complexity of accessing business and data logic in all those computer systems, you're not going to get anywhere.
These moribund reengineering efforts established the battle line that still exists today: "We could be a lot more efficient as a business today if we had systems that adequately supported the ways we want to do business and that could change when our business realities--and processes--needed to change. We should be able to take all those process diagrams we drew in those conference rooms and have our systems support them. Isn't that what IT is for?"
The biggest flaw in the reengineering concept (there were many) was that it was completely manual. You don't go after IT systems with a pen and paper. It wasn't so much that the business was completely naïve--after all, a pen and paper work fine when you're redesigning workflows on an automobile assembly line. Six Sigma and other process perfection efforts led to real productivity gains and quality improvements in manufacturing processes.
But Six Sigma hit a brick wall when functional computer systems appeared on the horizon. The black belts chopped madly at the blocks of big iron and came away with nothing but sore hands.
Frustration with this inability to link process change with systems change--along with the hysteria around Y2K--led to the next major conflict: enterprise software. CEOs reasoned that if they couldn't change processes with pen and paper, they'd use the canned processes inside ERP systems to do it. I've discussed this era in detail elsewhere in the blog. Suffice it to say that this approach worked in some companies and industries--mostly highly automated, centralized operations that had little variation in processes to begin with. But in many other companies and industries, it simply led to more frustration and resentment towards IT among business people.
IT's defense has been standardization and enterprise architecture. Both have been ineffective because they wind up being navel gazing exercises. Rationalizing the IT architecture, while certainly important, doesn't address the fundamental war cry of the business: We want to control and change our processes ourselves!
But then last week I realized that the new front in this process war is aligning behind two technology concepts: BPM and SOA. I spoke at Shared Insight's BPM conference last week in Phoenix and saw that the veterans of the business reengineering battle have ditched the pen and paper and adopted BPM as their weapon of choice in gaining control of processes--process mapping and workflow are the core legacies of these tools. Most of the people at the conference were either business people or IT analysts who work directly with the business.
Meanwhile, when I go to IT conferences, SOA is all people talk about. SOA is the way that IT will enable the business to change its processes. But of course, SOA almost completely focuses on IT and integration rather than process. It is the latest, best incarnation of the enterprise architecture efforts we've seen in IT for years.
But now that both sides have a technology weapon, I'm more hopeful than I have been in a long time. That's because for better or worse, the only advancements that have been made in enterprise process change (with emphasis on the enterprise part) have been technology based. The internet, integration middleware and enterprise software, for all their flaws, have built more cross-functional sharing of business processes than any six sigma process redesign effort. IT is better equipped and more motivated to support enterprise sharing than business people. A controversial assertion, I know, but research has shown that IT has had a more long-lasting effect on enterprise sharing than any of those organizational realignments that the business side tries every few years. There's just too much politics and competitiveness among business leaders and business units--as perhaps there should be--to build enterprise cooperation unless it is supported by IT.
So why am I hopeful about BPM and SOA? I know that they are buzzwords, but they both mean revenue for IT vendors and consultants and the pools of money grouped around those two buzzwords are converging. Traditional integration middleware vendors (the big moneymakers in SOA) are beginning to add BPM tools to their fold and BPM vendors are diving down into the IT infrastructure through acquisitions and alliances.
The point is, however you want to link process change with systems change is fine with me. Just do it. Do it top down though BPM or bottom up through SOA. The business people at the conference seemed to think that BPM allows them to maintain their control over business process without throwing it over the wall to IT. Great. Wonderful. Takes away the powerless feeling they have with IT. Of course, I hear that BPM tools don't really have the power to link changeable process pictures with actual systems change at a deep level for major transactional processes yet. Likewise, SOA hasn't had much emphasis above the IT integration level yet.
But if the tools converge, the convergence could force the business and IT to finally address the core of the process conflict. An armistice could finally be at hand in the process wars. I know, I know, tools don't substitute for leadership and alignment and all that, but if you look back at the last 20 years, the tools have made more of a difference than the humans. As one conference attendee said to me, "The business has BPM and IT has SOA. But at least they both connect in some way to process."
What do you think?
my .02
/rant
This article is full of the most dangerous kinds of sweeping generalisms - based on pure conjecture and little fact. Is their any hard research to back up any of these claims?
Why is automation where the business doesn't get it? If you look at the three year revolving chair for many CIO posts I think that IT may not get it. Consistency of practice - then automation..or else you will automate failure (isn't there a shrinking auto-maker that once valued automation over consistency of practice? They spent more on robots than it would have taken to buy Toyota and several others that are giving them headaches now..oh yeah and they wound up throwing out the robots)
This article and the several others I have read on CIO.com today (esp the CTO - a dangerous title) are full of vendor/consultant-speak and catch phrases from analysts that they just can't drop ("driving the business" is the worst IMHO). Who wants someone like IT to drive when their driving record shows that they crash all of the time and 80% of the crashes are self-inflicted? When the rescue workers pull their bodies from the burning wreckage by firefighters they are found to be mumbling incoherently about "Strategy".
Many of these articles refer to "The Business" as a separate entity from IT. This makes ZERO sense. Who pays for IT? IT exists only in a context and that context is to remove or diminish business constraints hindering the execution of its stated goals. There is no need for "alignment" when IT understands its role.
/end rant
Where is the line between "business" and "IT"?
Business is about doing business, and IT is about handling Information Technology, but the division still lies between those who can direct smart people to do their part in the processes, and those who can formalize to the extent that a stupid computer can do its part in the very same processes.
It all boils down to "the art of formalizing processes enough", but aimed at either people or machines. And the frustration with IT is basically a frustration over the unexpected stupidity of the computers. They need an awful lot of formalization.
So, in a sense SOA is merely the opportunity for IT to create bricks that can be rearranged with less effort than if the whole process need to be formalized in one monolithic go. Bricks delivered to the business with the hope that enough complexity is hidden for the business to understand the formalism needed.
But the catch is that control over processes that are automated can only be achieved if you know how to formalize it well enough. No matter if it is realized in BPM-tools or a SOA. And that is - still - a field where people in IT has an advantage.
So the only way for business to get control over the processes is to learn IT. Not in the sense of learning the particular technologies involved, but a need to learn software economics, and to learn the level of formalism needed. The analytical part.
There is no short-cut when it comes to computers. They are still stupid and need explicit directing. As long as they are a part of our workforce, it is the responsibility of the directors to learn to direct them. Or step down to make room for those who can.
Ich can mich an dich uberhaupt nicht errinern.bmq
First issue: IT is manufacturing and favors stability and repeatability. BU is sales, always looking for the new edge. That tension will always exist. IT asks for better planning, BUs ask for more flexibility. Now and forever.
Second issue: IT infrastructure model is broken for new applications. Google serves millions of users, rolls out new applications every month, tweaks constantly, with nary a mainframe or a RAID array in sight. Instead of building reliability into devices, they build reliability around devices, that are cheap and endlessly replicated. IT is a prisoner of its infrastructure model.
See more at http://storagemojo.com/?p=66
The difference between Google and many other firms is that Google rolls out things developed according to their architecture.
When the business has a clear grasp of its own architecture (which itself is a result of previous business), it won't come up with frivolous ideas that crash the architecture, but smart ideas that utilize the architecture they already have.
Or, alternatively, orders the architecture to be changed, but this time explicit instead of implicit.
Business people are smart people. The only thing they need as IT owners, is a little bit of architecture awareness.