Back when I was first interested in pursuing a career in computers in the early to mid-nineties, the consensus on staffing an IT department was that it was much cheaper to outsource, or in other words, hire in contractors to fill your staffing needs on demand.  Not to be outdone, many a corporate accounting department canned their IT department, though usually by transferring them to the payroll of a contracting firm and paying significantly more per hour to have them on staff.

Over time the consensus on that mindset shifted.  Since I worked for a contracting firm when I first started out it was easy to see why: any work performed by computer consultants that is in the best interest of the company hosting them is strictly a coincidence.  With the IT staff concerns unmoored from those of the company indirectly paying their salaries, companies found that they weren’t really saving all that much money, especially when the product which was delivered was factored into the equation.

With Y2K consultant extortion still fresh in the minds of management, a different mindset came about in the post dotcom bust when IT pay rates came down and in-house staffing became more attractive.  The idea behind this strategy was that IT would be a driver for the business.  IT in this situation wouldn’t just be a mindless socket in the wall from which other departments would draw resources, but would be a partner in implementing new ideas of different business units and indeed, would be a driver for new ideas themselves.  This has a great deal of appeal to me since it makes sense that if a business has their own internal IT resources, they should justifiably expect that those people would be intimately familiar with how the business operates and that they should be able to apply technological efficiencies to everyday issues encountered within the business.  In other words, IT wouldn’t be about just fielding helpdesk calls, it would be about meeting with department heads and employees to find out their operations in order to make sure that IT isn’t an impedance, or IT could offer solutions to problems that are encountered in the differing department’s day-to-day jobs.

That strategy seemed to stick for a bit until the great recession bit in and businesses found that, although they valued IT, they were rather broke and the quality mattered less and less.  This ‘great cheapening’ didn’t only affect IT, but many IT services are distinctly sensitive to it since they can be easily farmed out beyond the geographical boundaries from where the service is required.  Which brings up the issue of the ‘cloud’.

‘Cloud’ computing was formally known as ‘hosted applications’ and are hardly new.  The business I work for was using a hosted form of a credit reporting app around Y2K, and it wasn’t much after that when we moved to a hosted payroll service.  In fact the hosted payroll service brings up a great point about hosted services in general: maintaining a third party app in the hosted form is only mildly less intensive than maintaining that third party app in house.  As an aside, it can be rather troublesome as well since you rarely have direct access to the hosted application.  For instance, this makes report writing for the hosted payroll a rather large bother.  Overall, as the IT guy for my company, I can take or leave hosted services.  Much like outsourcing in the past, sometimes it’s a fit, and sometimes not, but rarely will any significant cost savings be realized (though I should point out that exceptions to this can exist for exceptionally large, or exceptionally small companies). 

What’s brought this to the fore rather recently is this article that my buddy e-mailed me, and in the author’s intro paragraph he gets many things wrong:

When people talk about their IT departments, they always talk about the things they’re not allowed to do, the applications they can’t run, and the long time it takes to get anything done. Rigid and inflexible policies that fill the air with animosity. Not to mention the frustrations of speaking different languages.

How many policies are because of HR or Legal, or because some third party app on which the business is wholly dependent requires some flavor of ‘tech voodoo’?  The answer would be “most if not all”.  When I saw his line about ‘speaking different language’ I couldn’t decide if he was making a case for or against outsourcing, but anyway:

IT job security is often dependent on making things hard, slow, and complex. If the Exchange Server didn’t require two people to babysit it at all times…

I’ve never known a management team that tolerated an internal IT team that provided intentionally “hard, slow, and complex” services (outsourced services on the other hand…).  It’s also worth pointing out that if a business is large enough to require two full time exchange administrators, it’s highly doubtful that they would require less people to manage the ‘comings and goings’ on an outsourced e-mail product.

He then goes on to spout many of the same old arguments I’d heard a little over fifteen years ago, and they’re not necessarily bad, they’re just not quite as ubiquitous as he’d like to believe.  Just as in the past, insourcing, outsourcing, or a mix thereof is highly particular to a specific organization.  For instance, I currently host our web page in-house since I have to have a proxy server anyway and the in-house server buys a lot of flexibility (especially in terms of easing stack upgrades into place so as to test the web site code), however I host the DNS services to provide a better level of fault tolerance (but it’s worth pointing out that it’s actually more difficult to manage the hosted DNS due to the usual awkwardness of web interfaces).

In the comments section to the piece there is a line that sums up many of the issues with generic IT people and IT outsource adherents at the same time:

An IT department is there to keep employees in check who are not even qualified to use a mouse. If everyone had to bring their own computer to work, the IT department would be a separate entity designed to bring a profit.

Anyone not qualified to use computer would have support costs taken out of their wages or would be driven out of the company financially if grossly incompetent.

But as the IT department instead is seen a cost, the real IT incompetent dead weights are just left to cause havoc in a companies infrastructure.

The whole problem with this line of thinking (IT thinking users are ‘stupid’ and users thinking that IT is nothing but a bunch of babysitting killjoys) is that users outside of IT have a job to do apart from their computers.  I’ve had users frequently apologize over some system issue, to which I always reply with something along the lines of “that’s why I’m here, so you don’t have to worry about stuff like that”.  I don’t fancy myself an accounting expert, so why should accountants feel the onus of inferiority when they don’t know their computer inside and out?  Likewise, IT is required to implement corporate policies and manage the network in a manner that is fair to everyone (i.e., not letting a handful of users bring the network down by watching YouTube).

Beyond ever present bandwidth issues, there’s one other point that needs to be brought up in relation to ‘cloud computing’ /hosted services: the ‘eggs in one basket’ with a side of ‘vendor lock-in’.   One of my favorite lines in a cloud computing puff piece was a hosted ERP customer saying that the vendor had inferred that there would never be any drastic rate increases.  Good luck with that, at least with a stand alone system you can walk away from an overpriced maintenance pact.  And even with pricing aside, over the past several years there’s been several cloud services that have died off and not even because the company was some fly-by-night operation.  I used to use a Yahoo ‘suitcase’ operation to hold some of my files on their system, but it was killed off years ago.  Google video?  MSN’s baseball subscription service?  Microsoft’s Kin system? All dead.  Online services only last so long as the company providing has the cash and/or interest in maintaining them.  Taking advantage of services of modest importance is one thing, but putting business critical systems completely beyond a businesses control may be ill advised.  One of the poster children for ‘cloud computing’ is Salesforce.com, but the kooky, yet business savvy Karl Denninger points out that business may itself blow up in a bust so big that would put dotcom bankruptcies to shame.

Point being, sometimes business realities dictate that something must be outsourced, but if something is critically important to your business there’s nothing more reassuring than being able to put your hands on it.