The problem with Content Management Systems
Posted on March 21, 2006 | 7 Comments | No TrackBacks
Apologies in advance to my friends in the industry who work for, with, or otherwise love content management systems (CMS). Feel free to refute this in the comments section.
Anyone that knows me knows that I'm not a fan of content management systems. Clients always seem to want them and I always want to ask... why? The majority of the time it's people who'd never dream of doing something like changing the oil in their car, but for some reason they have no problem editing the primary communications vehicle for a multi-million dollar corporation. Not to be cocky, but you hired me for a reason. If you don't know anything about the web, what makes you think you can manage your website?
Whenever I implement a CMS for a client, one of two things invariably happens:
- The client ends up destroying the site. - Broken links abound. Formatting and styles change radically from page to page. A 6-megabyte image on the homepage. I've seen it all.
- The client ends up calling me. - The CMS proves to be too much of a challenge for them (or doesn't offer the features they want) and they end up asking for help. Obviously I have no problem providing help to my clients but when a CMS is involved it ends up taking 5 times as long because I have to make all changes through a 12-step CMS wizard instead of just quickly editing the code.
From custom built database systems to enterprise solutions like Vignette, content management systems are usually expensive, difficult to setup, or both. For that reason I often advocate that clients simply setup a monthly retainer with their development team to perform maintenance and routine updates. More often than not this is much easier and more cost effective for the client.
EXACTLY! I've seen this time and time again. If it's not broken links, it's some crazy fonts, or it's an image that was 640x480 from their last luncheon that they shot on a 2 megapixil camera in bad lighting and is now being squeezed to 150x80 with width/height attributes. ARRRRGH! By the time they're through editing a normal paragraph there can be a ton of empty paragraph, span, breaks, and font tags oh my!
I'm all with you on clients being completely capable of ruining a perfectly good website, but what about when you want to have automatically refreshed content on pages throughout the site, such as your Recent Posts on the right? If there's a non-CMS alternative for that, I'm all in...
I think we should make a distinction between dynamic pages and CMS.
If you want automatically refreshed content on pages (i.e. press releases in a sidebar), then database driven pages are the way to go, and any decent web developer can create a system to update them.
The problem with CMS is that the client is adding the new content instead of the web developer so the system either has to be smart enough to prevent them from doing something wrong (i.e. expensive to buy or build), OR it'll let them do something wrong (i.e. garbled HTML, massive images, etc.)
Just as you'd use a good mechanic to maintain your car, you should use a good web developer to maintain your site.
I hate generalized CMS as well. Why try and do everything when usually 90% of the content changes on a site are managed by 10% of the processes created. It's those other areas where clients forget what they are doing and lack the repetition needed to get it right over time.
Yet here we are, submitting comments on an entry that was put into what has become one of the most common CMSs around: Movable Type. You're a skilled web developer, after all: why aren't you creating these entries manually?
I think Content Management Systems are the way to go even if the client never does updates. The right CMS (if you have even one page where, say, you post videos that are sent in every two weeks or so, and the first three are pulled in to a main page and the others to an archive page by really simple code, that's a CMS) adds order and reliability to tasks that can get out of hand if done by even good web developers who haven't systemitized their method.
One of the reasons things like MT are so beautiful is that your content is completely and fundamentally separate from persentation. It's a DB. It can be rebuilt into any framework. Like it or not, that's still far ahead of even the most standards-based XHTML/CSS these days.
I've used MT as a CMS for my personal homepage for over two years, and I can't imagine ever going back. I've used MT as a CMS for client sites where I will be the one doing every single update. I look forward to the day when we never think of content as residing primarily in files.
Nate makes some excellent points and for the most part I agree. Perhaps I should've named this article "The Problem with Clients Managing Content".
MovableType is great and I enjoy using it... but my point is that I know what I'm doing. I can upload a bunch of images to my server, then code the proper image tags to display them, maybe add a new CSS rule to align them the way I want, and then paste that HTML into MT. Most clients I've worked with couldn't do that.
My main problem with CMS is that they're most often used by people who know nothing of the web. That's a recipe for broken links, malformed code, missing accessibility attributes, etc, etc.
A CMS or blogging software like MT can be great as a tool to help developers elminate lots of tedious update work but there isn't any magic software that can turn Betty in Accounting into a professional web developer. Unfortunately, many companies think that's exactly what a CMS does.
Well... yes. But what about CMS use that's limited to just text? I do feel completely comfortable letting Betty in Accounting add/edit entries in MT as long as she knows she shouldn't ever use XHTML.
I like to see CMSs used that way because it allows for far more generation of content, and content makes a website more valuable to a client, to a client's users, and, therefore, ultimately more willing to come back to people like me to pay for more tools.
But maybe I'm getting away from the point again. I definitely agree that no automatic tool can turn someone who doesn't know web development into someone who does. But on the other hand, I really like things that enable people who aren't web developers to become publishers of content.