Graham Hill has a thoughtful blog post on Five Challenges that Keep CMOs Awake at Night with some insightful thoughts on the future of CMS software.

Some bullets:
"Marketing is inexorably becoming more decentralised"

I sometimes have to remind myself that the "web" is a reference to it's intrinsic structure and that applying centralized, hierarchical, and org chart models to the web is often the source of some people's stress.

We, as a CMS vendor, make no promises that we can help manage all web content. It's just not possible or practical today. I am, however, passionate about centrally managing all marketing investments and tracking/correlating their responses. CRM mixed with CMS is a good solution for this.

Customer-responsive model.... A model that biological systems have perfected billions of years ago. In this model, broad rules are developed centrally for universal use at touchpoints with customers, which are then interpreted locally in real-time using all the contextual information available.

This is exactly correct, but how does a CMS address this need? Our response is to create languages at a higher level of abstraction than basic HTML and let marketers seed page templates with rule intentions that allow for specific types of dialogues to emerge.

The CMS of tomorrow will be addictive, much like Google, Facebook, or Twitter. Marketers will actively monitor point solutions and rely on the CMS to spot trends in real-time and allow organizations to respond to comments, ratings, and tags.

The concept of 'reports' will gradually disappear and be replaced with real-time 'trends'.

Web Statistics 2.0 will emerge around social metrics. Legacy metrics like impressions, unique visitors, and browser type by region will have decreasing value.

Customer Co-creation is the next big thing.

When websites are effectively configured to support community input, then marketing's ability to "listen" becomes extremely important.

As CMS‘ have grown larger, more integrated, more unwieldy and more expensive, some organisations have responded by going for simpler, leaner, right-sized CMS tools

In designing i-Dialogue 9.0 I've come to several "forks in the road". Should we make a particular feature simpler or add more advanced functionality? Is there a compromise between the two?

For example, the concept of a "discussion forum" is one that should be as simple as adding a <dlog:Discussion /> tag to a page template that let's a thread ensue. But there will be those that criticize that simplicity. The old way of thinking is that you're 'supposed' to deploy a forum as a separate, and often expensive, solution.

There is such a huge gap between what analysts are telling people they need and what companies really want. In economically challenged times you see purchasing committees emerge that are trying to play it safe and manage all possible risk by purchasing a complex CMS system that has been around for several years.

Graham goes on to use Clayton Christensen's "Disruptive Technology" model as a means of understanding next-gen CMS's. If that is the case, then bottom-up adoption of faster, simpler, cheaper solutions are on the horizon.