CMS products have come along way in the last five years in terms of functionality and usability. They now (mostly) deploy quickly, are reasonably easy to use, and are a lot cheaper. This is all good.

As I’ve discussed many times, the usability of CMS products is crucial, and there is a lot more to be done on this front. In particular, CMS products should match the “mental model” of authors and site managers, something that they often don’t do. The confusion of different terms, concepts and philosophies doesn’t help.

This is still not enough. I want to present an open challenge to all CMS vendors:

Content management systems need to much closer match the reality of how content is created, managed and published. They must also do more to help resolve the fundamental challenges that led organisations to purchase CMS products in the first place.

Despite the profusion of content management systems, each with its own ideas and models, there is surprising similarity in how they operate at a basic level. The concept of “workflow” is more-or-less universal across products, as is metadata, review and expiry dates, and the like. Underpinning this is the idea of a “publishing process” in a traditional sense, replicated from print to the online world.

The problem is that this doesn’t actually match reality. As I’ve said in the past, workflow doesn’t work, organisations still struggle to manage their content even with a CMS, and few manage to capture the metadata they require.

Indeed, it is questionable whether installing a content management system really resolves the problems it promised to. One of the reasons is that vendors haven’t really put in the effort to really design their products to best help authors and site managers. Rather, they are happy to simply follow the “way things are done”, even if these ways demonstrably don’t work.

My challenge to vendors is therefore to conduct user research to deeply understand how their clients are working with web content, and to revise their products to better match this reality. This will dramatically improve the success rates of CMS projects, which in turn helps the vendors.

Let me give some examples to put some shape around my rantings:

• Content reviews
  One of the fundamental challenges is to keep web content up-to-date, particularly in a decentralised authoring environment. The standard way CMS products endeavour to meet this need is via a “review date” for each page. This may be fixed to yearly (bad idea!), a chosen date, or a selection of possible periods (three months, six months, a year, never).When the review date is reached, an email is sent to to the author of the page, who may or may not be the actual content owner. The hope is that the author will then review and update the page (if required). Site managers may get a report listing pages overdue for review.

  The problem is that none of this really works in practice. Authors often ignore the reminders, and site managers need much better tools to track and manage the review process. How can we do this better?

• Metadata
  It’s notoriously difficult to get decentralised authors to enter metadata. It’s even harder to get them to enter good metadata. In practice, metadata quality is mixed, with basic problems such as a mix of singular and plural terms (for example).Site managers are generally given no tools to monitor, manage or improve metadata at a site level. There is no single list of metadata terms that allows global changes and cleanups to be made. Authors are given little help in entering useful metadata terms.

  Professional indexers and taxonomists have been asking more functionality for years. Why haven’t they got it?

• Workflow
  I’ve said a lot about this already. So all I’ll say: why haven’t we implemented a publishing model that actually matches how content is created and released?

(These are just a few simple examples, and other similar situations can be found across every aspect of content management systems.)

Note that I’m talking about improving the core design of content management systems. Yes, web 2.0 and enterprise 2.0 provide valuable new ways of looking at all of this, but the basics still need to be right.

I ask again of vendors: please do the work needed to make your products work better in the real world, even if it means challenging some of the ways “things are done”.

So, thoughts and comments? (As usual, in the continued absence of commenting, please email me and I’ll post the highlights.)