After I posted this article last week discussing cost-cutting measures and briefly mentioning that CIOs who had structured their departments to make significant use of outsourced services had a much easier time cutting costs quickly or temporarily, someone mentioned that I should have been more specific on exactly what service to outsource for such results. To which I can only reply, good point. I should have addressed that more deeply, and will attempt to do so now.

The crux of making specific recommendations in broad contexts such as this, of course, is that everyone's department and demands are slightly different. While I'm confident that it's good practice in general to attempt to use outsourcing, I wouldn't care to recommend that everyone reading outsource, say, their telecommunications infrastructure. It's probably applicable for many of you, but then there will be some company for which it's a critical part of day to day operations and can't be reliably entrusted to hands outside the company.

So that would be my first rule for what you should consider outsourcing, then: services or functions which are both suitably generic and sufficiently distant from your core business that a third-party firm specializing in delivering them can probably do so reliably and more efficiently than your team in-house. Telecom is a good example; it's complex, but common, and there are a sufficient number of providers that good competition keeps the market efficient. Further, while most businesses need telecom, it's not one of their core competencies, anymore than generating electricity is (although electricity is just as necessary).

The second rule would be to look to those areas which the originally cited Hackett study notes as likely candidates for reducing costs in-house: application development and maintenance, user support, and the like. If those services are typically run so inefficiently internally that they have so much cruft to attract Hackett's attention, then it's likely that they can be outsourced effectively to specialty providers.

There are also those who suggest that outsourcing is not simply about cutting costs, but also about establishing lasting business partnerships, and I also support approaching the whole matter with that in mind... corporations which outsource solely to hack away at costs rarely come out of the endeavor with much satisfaction.

Determining what to farm out and what to keep is an age-old question that can only be adequately determined by careful consideration in each corporation, but the key thing that I find when approaching the matter is to divest the discussion of all the irrelevant clutter that is sure to come up and look solely at the costs and benefits. Few IT departments are anxious to lose IT missions, and many are largely populated with extremely technical people who enjoy doing complex, technical things. When those things, however, lie properly well outside the purview of the business, it's time to look at letting them go. It's fun to solder your own memory chips together, sure, but it's cheaper and faster to let Kingston do it for you.