Tom Wentworth, from Ektron, has written an inspirational article on Forbes called "http://www.forbes.com/sites/ciocentral/2011/08/30/context-will-drive-the-future-of-web-content-management/". The article describes a very legitimate strategy for achieving a goal that CMS customers all want: serving their audiences with the right content, at the right time, and in the right format. The strategy, which uses visitor context (location and device) to infer intent and other preferences, is something that a CMS can help with. Content management systems have been introducing increasingly sophisticated profiling and presentation functionality to implement elaborate display logic. Content management systems have always been able to capture the metadata and other configuration that drives this logic.
The right CMS will give you the capability. But that is only part (and a really small one) of what is needed to be successful in this strategy. The bigger, harder part is designing, implementing, and testing the rules behind this logic. How would you treat a visitor differently if you knew his context? The fact is that you still don't know the visitor and what he wants. Context may give you a clue or a hint, but you can't be sure. You can only make assumptions and these assumptions may be very wrong.
These assumptions are like any stereotype. Sometimes they are accurate and you look like you have some kind of sixth sense. But when they are wrong, you better be prepared for damage control because you can look like an idiot and frustrate people. Success with a context-based strategy, or any other personalization depends on 1) knowing the likelihood of being wrong, 2) knowing when you were actually wrong, and 3) recovering from a mistake.
Let's take these elements in a concrete example. Let's say you work for a movie theater company and have a theory that the mobile visitors to your website only want to see showtimes and directions. One of your customers sees a help wanted sign at one of the theaters with a link for information on how to apply. This young applicant's only computer is a smart phone (an increasingly common phenomenon). He goes to the website and is redirected to a mini-site that only has show times and directions. He can't get to the desktop version of the site with job information. You just lost a job applicant and embarrassed yourself.
Running through the tests.... You might think it was a reasonable assumption to only show certain information to mobile users, but if you were not absolutely sure, you could have put a link to the full site and suppressed the redirect logic. How would you know about this incident? Are you tracking clicks on that "desktop version" link? Was there a place to leave feedback? Or will that incident only be seen as a bounce in your traffic report? If you figured out what happened, what are you going to do about it? Do you have the resources to correct mistaken assumptions and continually improve the behavior of the site(s)?
I really like Tom and what he has done at Ektron since he joined. In particular, Tom has helped focus Ektron and elevate the discussion about serving audiences. Buyers should definitely listen to these ideas but be aware that supporting context-aware visitor experiences adds cost to the initial implementation and ongoing maintenance. There is more logic to test before you launch the new site; but the real effort is after deployment when you have to constantly test and tune the logic as you see it respond to real users. You have to accept that context-driven logic has the potential to worsen visitor experience and be prepared to make corrections. This means having a team with the time and skills to monitor and adjust the logic. If you have access to that talent, you should be able to use context information to help deliver great visitor experiences.