I think you’re spot on regarding how companies should proceed. Too often they choose proposals based solely on the budget or the development house (both are important, of course), when they should also bear in mind the CMS/CRM that vendor will use to build their site.

For example, if an open source platform is used, it might leave the door open for the company to hire contractual or full-time developers of their own in the future to roll out new features. However, it will mean taking on a QA and PM burden as well, which has to be factored in…

It is true that no CMS is perfect, but the worst situation is to end up with one that never (or rarely) gets used. Sometimes it’s due to poor usability (engineers can build some pretty scary UI), sometimes because online content is not prioritized highly enough. But the key is to see them as we in the web business do: a means to an end. Only then can the causal relationship between a site’s purpose and functionality be adequately assessed.

BTW, if you liked that, check out: http://informationarchitects.jp/wtm4/

And XKCD sells poster versions of two popular comics/maps- Online Communities & Map of the Internet: http://store.xkcd.com/

The map really speaks to why businesses deciding on a CMS really need to create a complete plan before embarking on testing or finding a consultant to help them decide. Clear goals, timelines, and a solid budget are absolute necessities.

No matter what CMS you choose, you’re always going to find issues with it later, be they technical or aesthetic. Nothing is ever perfect given enough time. But if preparing manageable goals and timelines is done beforehand, businesses will also be able to find a CMS that is malleable enough to work with future changes.

(I’m printing this out by the way. Good catch.