We are nearing the end of a process of rolling up a number of flat Web sites into the corporate content management system. The initial migration happened 2 years ago and was largely automated. The team that led that project missed a number of opportunities to clean up the junk as the HTML was converted into database-driven content. The end result was that we continued to have 25,000 pages that had terrible metadata and mystery meat navigation.
The content management team activated a status quo intranet this year. We moved over current content and the design but left most of the content behind. As users need it, we are migrating it over so they can update it. It has given us some good opportunities to leverage the system’s automation. For example, a content owner places content in a folder and the site automatically displays the most recent items. This is ancient as far as CMS features go but it is a big step forward for our content culture.
This week I am moving over the last site. It’s a jumbled site that grew out of departmental silos and an attempt at audience focus that collapsed under its own weight. Content was funneled through a set of developers who did a great job of getting information up and into a home. But there was no editorial oversight and the navigation was just as execrable as the main site. The corporate site had an entrance menu of just around 140 links, and this site has nearly 150 on its left navigation menu.
The process of migrating the content over is partly to ensure we only bring over live content, not grabbing everything deposited at the file level of the Web server. As each piece is copied over, we’re placing it both in a corporate workflow – departments responsible for their content – and a navigation structure based on audience only. It’s a huge change culturally and should create a very small left navigation menu, on the order of 20 total items.
One of the things that struck me as I was moving across content was the personalities that were exhibited, both by the content owners and the developers who supported them. When you work for lawyers, you become inured to lengthy, windy documents. But the Web project has been an eye opener even for me.
First, you can see that certain departments and staff write in a particular way. Gerunds everywhere, for one thing: Filing your …, Contacting the …. Not only can you convert those to active – File, Contact – but you could even say what the document is about: How to File Your … or Contact Us About ….
Another common thread is the long document. I had no idea. I came across one today that had 20 subparts. It went on for miles and the developer had compressed the page by hiding each sub-part. The visitor scrolls down the page, clicks on an item in the list, is popped back up to the top of the page to see the content and scrolls back down for the next one. You can see why someone did this. It’s attractive to keep all the content in one place and use gimmicks to move people around.
The last thing that some groups seem to do is create frequently asked questions (FAQs) pages. I think these can be okay if (a) they’re very focused and (b) they actually have questions that people ask. Many FAQs seem to be (a) a way to make up for bad navigation and (b) questions they wish they could have been asked. I keep coming across FAQ pages, though. I almost expect to come across an FAQ about the FAQs.
Some of these I am trying to manage as they come across. The 20 sub-part document became 20 individual documents and a navigation page. The effect is the same, except that a searcher can pull up a single matching page instead of coming to the large page and having to hunt. The gerunds I’m leaving alone but it goes to creating some sort of style guide for the authors. The FAQs are being submerged within the content. If people want them, they’ll be able to access them within the context of the content they’re visiting – see page A and the FAQs related to page A – but the FAQs won’t be on the menu.
It’s an interesting process to go through. The information pro in me wants to fix more of it but that’s not my role. The content owners haven’t really touched the content since they sent it over to the developers. This is obvious because there are circular menus, or links that go to the wrong content. The web content team is trying to get out of the content creation and ownership role. It looks like, once we have this all moved over, it will be a good opportunity to sit down with the content owners and work on some standard practices. These would include:
- shorter documents;
- dumping PDF formats unless there’s a fundamental reason to use it, and keeping content as regular HTML;
- fewer documents, assigned to individuals, who periodically go through and update or delete older content;
- proper naming conventions for Web pages (no DOS-like file names or camelCase names)
I’m about 75% through the content shift and it’s been an enlightening experience. One of the repetitive issues is to undo the special styles applied to the content. Rather than use standards like <h1>, <h2>, I keep seeing things like <p class=Heading2>. That’s the developers giving free rein to their creativity. And why not? But it makes migration and decentralized content creation that much harder. Roll on the weekend!