Content Mapping

Sometimes I think I’m the only person who lies awake at night worrying about content. Well, I don’t literally do that, but it feels like I might be sometimes. I’m certainly gaining broken record status on the issue and thinking crying-in-the-wilderness thoughts at times.

Part of the problem is that it’s hard to articulate what the problem exactly is (well, I find it hard at least). It’s certainly made harder by the fact that according to the content management software industry it’s not a problem that exists if you use a CMS. How could it, since such software “manages” content! And who indeed could possibly have a problem with managing content after they’d spent half a million bucks on the latest enterprise XML format-agnostic end-to-end solution?

Not surprisingly, the project I’m on has just such a “solution” in place and it’s bringing the subject I love to hate back on the map for me again. To save me the bother of explaining why this is, read this piece on the subject (WARNING: shield your eyes from the photo). It hits the nail(s) on the head pretty much perfectly as far as I’m concerned.

But some problems on my current project are only indirectly related to a CMS. In the recent past I’ve been involved in some reasonably good solutions (up to a point) for getting around the more basic issues of dealing with “modular” builds, but that’s not going to be an option here. So we’re up against it again. The brief is to construct a better IA for the site, and migrate the existing content to that, culling, merging and re-writing as we go. An initial card sort has given us a good candidate structure that the client seems to be running with, even though it’s radically different from the current one. An initial audit indicates the site may have about 10,000 items of content, most of which is highly technical or at least assumes industry knowledge that we don’t have. It would be a big job even if we understood it all, but we’re going to have read, understand, and if we don’t understand, ask specialists about it. We have, erm, one day of two IAs in the project plan for this. Can you guess I wasn’t involved when they put that plan together? Moan, moan, moan (there, got that off my chest).

Let’s assume for a moment that we can get a better structure. That’s not hard – just time consuming. The problem we then have is how we communicate that new structure to a) those responsible for the content re-work, and b) those responsible for the content load using the CMS. The more gung-ho among you may say “Tough – it’s their site, they just asked you to re-design it.” But if the client can’t actually deploy the work you’ve done for them, who they gonna call? The gas board?

My first thought was to construct a huge spreadsheet. Each row is a “page” on the new (re-designed) site. I can then group existing content with each row (using Excel outlines) a bit like this:

Page Name X
     Intro text (to be written)
     Document 1
     Document 2
           Page Name Y
                Intro text
                Body text (to be written)
                Other text
Page Name Z
     Summary (to be shortened)
     Document 3

That would then be the “bible” when it comes to the new architecture. But how do I make sure that when I say “Page Name X” it’s the page to do with X on the current site? I could use its URL, but that’s very long (I note it’s even too long to be rendered by Excel as a clickable link) and they don’t all relate to the content in a one-to-one manner (long story, half understood). This is a CMS, remember?

So how can I expect somebody who doesn’t know the content intimately, nor much about my new architecture, to migrate the current stuff from one structure to another? Bear in mind the current structure is a really, really big mess as well, so it’s not the case that we can do things in chunks either. I’m currently looking at a unique “node ID” that the CMS generates for each content type, so that may help. But boy is it laborious to track down each existing node ID and associate it with a row in the above sheet. This is going to take weeks.

Perhaps I should just accept that just as in ten tousand projects before, it will all come down to the clipboard and a thousand monkeys. As it was in the beginning, is now, and ever shall be, CTRL+V without end. Amen.