Mac vs Pc

So this strikes me as a bad idea for a post title, but it does succinctly describe what I’m about to write about…

Anyway…

Earlier today I was talking to a friend of mine who works at a web design company next to my office. Apparently, before I walked in, he and the head designer had be discussing the Mac vs. PC argument. In particular, the designer had found this article on Infoworld discussing a recent survey of IT professionals. The tl;dr version of the article is that according to some 60% of the survey’s respondents, Macs have a lower Total Cost of Ownership (TCO) than Windows PCs do (note that Linux, Solaris, or other options are never mentioned in the article). So my friend said, “Well, what do you think?”

At the time, I told him I disagreed with the article and that, in my experience, enterprise environments saved money by using PCs. After I left his office, I started thinking about why I answered like that. Why was I so certain that Macs were more expensive that PCs in the enterprise?

I knew that many PC OEMs as well as Apple do cut prices when selling to large organizations (particularly at education institutions) and I knew that, at that level of IT, a certain competency on both platforms is likely expected any more. Then, after reading about TCO online, I found it’s essentially a bullshit metric that marketing people throw about. The way almost every TCO study has been performed has inherent problems in either its data collection or some other important factor. The early academic studies referencing TCO “had serious flaws and problems with their accuracy and reliability.” (http://www.eaconsult.com/articles/RethinkingTCO.pdf [PDF])

But also, TCO studies (proper ones, at least) require two completely homogeneous environment of Microsoft and Apple products with similar user requirements, size, etc. to properly indicate differences. And then there are so many other issues (e.g. do the tech(s) working there have more experience with Microsoft or Apple, specialized requirements [there’s no CNC software for Macs…], etc.) that can affect the outcome that it’s basically hot air.

So we can see there are issues with TCO metrics. But what about the opinions and experience of the experts who were polled for the article? Well, it looks like Infoworld essentially ignored any insight beyond one metric in the poll they referenced. They freely admit that both their article and the poll they referenced don’t take any upfront costs into effect when determining TCO. What was it I said earlier about TCO studies having problems?

“Editor’s note: The price of business-level Windows and Mac laptops are equivalent, at about $2,000. Consumer-level PCs, which large businesses avoid, are sigificantly [sic] cheaper than consumer-level Macs.” Here we have Infoworld getting statistics significantly wrong. A cursory glance at Dell’s Business Desktop Computers shows that a Precision Desktop would start at $799 and a Quad Core machine at $999. Precisions are typically ordered for the users in an environment who are doing CAD work, graphics manipulation, or other much more complicated tasks. The Optiplex desktops (which are more appropriate for a simple office machine) start at $399. Looking at Apple’s Mac Pro page, a comparable desktop to the $999 Precision from Dell would be approximately $2500.

Now I’m not a huge fan of Dell computers. I’ve fixed too many over the years to enjoy even looking at them. But this is a extremely significant different upfront cost.

The article ignores so many provisioning costs associated with setting Macs up on a network. They’re again assuming the lack of an existing infrastructure that caters to one environment or the other.

I’m not trying to shoot Macs down. I think OS X is fantastic. I think for a large segment of the population, there’s nothing wrong with using Macs. But there’s a heavy laziness and ignorance that seems to go into a number of articles (and opinions) like this. Adding new devices to (or completely rebuilding) a network is ridiculously hard. The domain admins at my old job all ran Linux ‘cause all the backend servers where Linux or Unix and it’s easy to connect to them that way. All the Exchange admins ran Windows ‘cause (oddly enough) Exchange is a Microsoft product and Windows connects better. The Journalism department was all Mac-based and guess what? The IT guy in the Journalism department ran Mac servers on the back end. If any of these people were asked to switch they would likely say “fuck you!”. Completely reconfiguring their networks would take a ridiculous amount of work. If they were told to switch to something else, they would put up with it and hate whoever made the decision.

A well-planned change, upgrade, or conversion can work, but the costs of changing things over time (and in the Enterprise, changes like this take years) would likely be quite high.

To summarize (tl;dr), it doesn’t matter whether the TCO of Mac or PCs are lower or higher in a new and homogeneous environment, that situation doesn’t exist any more. An enterprise environment (at this point) has proper high-level service contracts with the OEMs and ISVs that supply them their hardware. They have a infrastructure already in-place. The cost of switching OEMs alone would likely dwarf the drop in TCO of competing products.

Wow. Apparently, this issue gets me going.

Edits galore: C&C software is actually CNC software, tagging posts is a good idea, formating paragraphs helps to read things, I’ll get the hang of this some day.