« Short-Term versus Long-Term Thinking: the Conflict between Self-Interest and Strategic Value Creation | Main | Visit to Chile »

August 21, 2006

Comments

Alex Osterwalder

Irving, this reminds me of an interesting article in the Economist on "simplicity" as the next big thing in technology. Engineers have indeed focused much to much on the artefact rather than the user... But with the re-emergence of design thinking in some corporate giants (e.g. Procter & Gamble) the user will become the center of attention again.

The Economist on simplicity & technology:
http://www.economist.com/displaystory.cfm?story_id=3307363

Warm regards from Switzerland, Alex

Jeff Osborne

Hi Irving, thanks for the interesting post. I have a couple questions.

First, do you think it's possible for future ERP systems to acknowledge the inherent complexity of the full range of business processes, without using such complicated representations? In other words, can integrated process management tools be complex without being complicated?

Some teams claim it, have done it, and from my experience their success is limited to one or two clever modules (nothing like an entire ERP ecosystem).

I'm curious to know if you believe sufficient interface standardization exists to attract innovative teams to develop those modular products. It seems like there might be in AppExchange, or others that may have limited visibility, but strong open technology and a fast growing community of users/developers.

Perhaps a related question is; what might motivate IBM, Microsoft, Oracle, and SAP to join forces around establishing a standardized open API for ERP modules?

Thanks!

Sam Kaufman

Recently, some universities have founded undergraduate Informatics programs. I'll be attending one in the fall. Hopefully we'll find some relief in that.

Irving Wladawsky-Berger

I think that one of the main advances in engineering and engineering tools is the ability to abstract and hide the complexity of designs until it is needed. This is true in mechanical design with CAD/CAM tools, for example, and in microprocessor design with EDA tools.

We need to do something similar in complex business applications. Is it easy and do we know how to do it? No, this requires quite a bit of research and real innovation but I honestly think that it is doable. It is an area where there is quite a bit of activity in IBM, both in our research labs and in our software development tools area.

I also fully expect that once we can start decomposing the business applications into the components that implement them, we will also start identifying which components are pretty standard and should be develop in a more open, shared way, and which are truly differentiated and each vendor will keep them proprietary. Most applications will likely be a mixture of standard and proprietary components, integrated together using tools that themselves will have standard and proprietary components.

Why will IT vendors support this? In the end, if the market demands it, that is, if the customers of our products want us to give them simpler to use, more standards based products, we have no choice but to do so. This is what happenned with the Internet, Linux, SOA, etc. Some vendors may be happy to do it, some may do it kicking and screaming, but I don't think there will be a choice.

Hopefully the information presented so far has been applicable. You might also want to consider the following. - William B. Doyle, http://www.wbdoyle.com/blog/

The comments to this entry are closed.