More on suites vs. best of breed

A reader writes that when it comes to open vs. proprietary technology, the choice is as much about what you do after as it is about the decision process.


More on suites vs. best of breed

In response to the Sept. 17 column by Chuck Phillips, "Suites vs. best of breed":

I have been in the IT industry for 20 years and have worked in the open systems area since the mid-'80s. I worked for IBM for many years and found myself asked to discuss open vs. proprietary back in the late '80s. A point I used to make then that I believe is equally as valid today is that the choice is as much about what you do after as it is about the decision process. Open or proprietary, suites or best of breed, you are making a choice of management discipline when you make the technology choice.

Why is this? Well, I have no formal data to prove this point, but I assert that the old business-school graphic that charted profitability vs. the value-add through to the low-cost business strategy and showed two profitable business models is at play. In the IT case, the ends of the spectrum are proprietary and open instead. Being in a half-and-half position is non-profitable. I know this is not literally true, but the principle of an optimal position is one most people would agree with.

My assertion is that these are an open stance or a proprietary stance. Both are effective, if managed with the appropriate technology management discipline.

For the open or integrator approach, you must observe the disciplines of openness. That is, clean interfaces, build your own end-to-end solution, design in interoperability, remove dependence on any single vendor, and so on. You have freedom of choice but the responsibility of integration.

For proprietary, follow the relevant disciplines for this pre-integrated approach. These include seeking and accepting vendor strategy advice, investing in understanding of the vendor strategy, following the vendor approach fully, and resisting the temptation to build your own when the vendor version is coming.

You have the advantage of pre-integration and risk sharing, but, as a result, have a lesser choice. The biggest problem is buying a proprietary technology and then taking an open management approach. You end up with neither choice nor pre-integration.

Mark Johnson
Melbourne, Australia