Integration Matters
In his post IT Doesn’t Matter – Integration Does John Schmidt makes an interesting point in respect of Nick Carr’s first book "Does IT Matter?", which caused shockwaves through the IT industry. Mr Schmidt argues that even if IT becomes a commodity there will always be the issue of integration to attend to. It is not just a case of ‘plug and go’. He focuses on business integration as opposed to systems integration, but his argument holds for both. Thus even if all hardware and software eventually become commoditised there will still be competitive advantage gained through one’s ability to integrate disparate technologies together.
To some extent using Carr’s book as a basis for argument is a little unfair as it was written for a previous biotechnological era (the post-dotcom nuclear winter). Though I admittedly am guilty of this as well in my new book. His book’s value was not so much in the logic but in the fact that it gave the IT industry and user community a big slap in the face. As a result today’s IT discussions are increasingly based around business value.
Schmidt may possibly be a technologist and this can see deeper into the issues highlighted. However as an ex technologist who is looking at trends in the marketplace, I see the whole area of system integration becoming a non-issue. Hence the death of the phrase system integrator in the description of what today are better known as IT service companies. Technologies and standards such as XML, Soap and middleware in general have taken the fun out of integration from a technologist’s perspective.
The issue of business integration is less evolved and so Schmidt’s perspective is very valid today and will be for a while. As we have seen, and will see, courtesy of the recent credit crunch, swathes of people will lose their livelihoods to be replaced by technology. As the ratio of people to technology changes the issue of business integration will go the same path as system integration.
@
Thanks for picking up on my blog post. Now that I’ve discovered your writings, I’ll try to keep up.
I feel compelled to address your comment about “system integration becoming a non-issue”. While “the death of the phrase system integrator” may in fact happen, it will be for reasons other than integration becoming a non-issue. The phrase may go away simply due to our habit in this industry to constantly come up with new buzzwords as we mature.
From my perspective, I actually see the reverse happening. That is, integration becoming a formal discipline. Universities and Colleges are starting to include integration topics in their computer science and MBA curriculums. In 10-20 years, we may be able to hire a university graduate with a BSc in Integration.
I maintain that integration is a distinct discipline and should not be confused with enterprise architecture, software engineering or program management (although a good integrator has elements of all three). Some architects claim they do systems integration – but only from the perspective of defining and governing standards – they don’t actually implement, operate and maintain the integration elements. Some people apply software engineering disciplines to integration but this only goes so far – the reality is that an integrator needs to deal with issues resulting from components that were engineered independently using incompatible technologies and inconsistent data not to mention a mis-match of process context – many of these are political issues and have little to do with technology. And program managers do indeed perform systems integration – but as a “project” and not from the perspective of the integrated solution having a life-cycle that needs to be sustained indefinitely.
There are two key reasons why integration won’t become a non-issue in our life-time: complexity and semantics. The tremendous complexity of modern-day systems-of-systems are such that they cannot be “engineered” – there is a human limit as to how much one mind can comprehend so we must engineer components within manageable problem domains and integrate them as a separate discrete activity. And the problem of semantics or “meaning of data” will not go away until computers literally become smarter than humans.
While debate generated by articles like IT Doesn’t Matter does have a positive side, on the negative side it discourages young people from entering the field. This sentiment is at least one of the factors behind the decline in computer science graduates in the U.S. in recent years and our need to import talent from overseas. We should do the reverse and encourage our brightest minds and most talented individuals to enter the field of integration since this is where many of the toughest IT challenges are that need to be tackled.
@
Blogs are so interactive where we get lots of informative on any topics nice job keep it up !!