The WCM marketplace

Web analytics is hot right now, in part due to the rise of Google’s free analytics service. The problem for Marie and her publishing colleagues is that most Web analytics products require them to learn a new system that lives largely outside their WCM tool. WCM vendors have taken note and are haltingly moving toward analytics dashboards built right into editorial interfaces, so that traffic shortcomings and opportunities can get more readily surfaced and addressed. The technical challenges here are not trivial. Also, customers with two different analytics services (one in their WCM tool and one from a dedicated analytics package) who find the two systems reporting very different results are left to face the difficult truth that Web analytics remains a rather mysterious discipline.

Micro-site development (and multisite management more generally) presents a surprisingly difficult set of challenges for Web site authors, managers, designers and system administrators. European vendors that had to grow up accommodating multilingual publishing scenarios have conveniently found themselves supporting multisite environments better than many competitors. That helps account for their recent success even in the (mostly) uni-lingual U.S. market.

Social applications (like blogs, wikis, forums, etc.) and user-generated content present some dilemmas for WCM vendors. In the face of rising pure-play suppliers of those services, WCM vendors often invoke the same argument ECM vendors trotted out against them just a few years ago: "Isn’t it all just content?" Well, yes, a blog is indeed a Web content application. But it’s a very particular type of content application, especially with respect to commenting services, which introduce all the complications of user-generated content. Not all WCM product architectures can easily accommodate content flowing into their repositories from the visitors’ side of the Web site.

Many enterprises that employ their incumbent portal or WCM technologies to support social software on their intranets are turning to separate, best-of-breed blog/wiki/rating/etc. tools to support user-generated content on their public Web sites. It’s a trade-off: They obtain better functionality in exchange for yet another independent set of content services and repositories.

Hard problems, still unsolved
The relative youth of the WCM marketplace that has spawned so many successful vendors also reveals itself in the number of hard problems for which most tools on most days have a good answer. They include:

• scalability, of both management systems and Web site delivery applications;
• orchestration, of often dependent logic, content and design changes through a proper life cycle;
• word conversion, a perennial challenge that leaves authors dissatisfied, even if the problem really originates on the desktop;
• reuse management and multiple preview services, to get a real handle on where and how all the components accumulating in Marie’s system are getting published across different channels and for different devices; and
• retention and records management, concepts almost an anathema to the first generation of Webmasters, but now too important to ignore.

Marie finds all this quite annoying. But I haven’t yet touched on perhaps her greatest frustration: that creating a dynamic user experience remains largely a developer’s task. Nearly all WCM tools ship with one-off templating systems that require a developer to finish off and install (and later modify) page layouts created by graphic designers. Moreover, publishing the kind of highly interactive pages that Marie now envisions typically requires coder intervention, with interaction rules often set outside the content management system—and therefore just beyond her reach.

Of course, the advent of Asynchronous JavaScript & XML (AJAX) has Marie thinking beyond the page, to try to create interactive experiences that don’t require navigating from page to page. Many WCM vendors certainly had Marie in mind over the past year when they souped up their administrative dashboards with AJAX widgets.

However, few if any WCM products offer her useful tools to support publishing AJAX-fueled pages. Even WCM products that manage content as component snippets still assume Marie will publish to page-based output. Swapping out individual components without refreshing the page can produce myriad different user experiences (that’s the whole idea!), but how is an editor or manager to understand or test them all? Marie will look in vain to find simulation services in WCM tools like she could access in most business process management (BPM) products.

Open source landscape

No examination of the WCM marketplace would be complete without an assessment of open source solutions. Since the emergence of Web content management tools in the mid-1990s, open source platforms have played an important role on the landscape. Open source WCM tools have never reached the kind of prevalence we see of their brethren in the (more developer-intensive) application server and portal worlds, but they receive greater attention in the WCM marketplace with each passing year.

Open source WCM offerings divide roughly into two camps based on business and organizational model:

• commercial open source—those that are driven by a central commercial firm that sells support, perhaps consulting, and very often fee-based add-on modules. Examples include Alfresco, an ECM vendor with belated but ambitious WCM plans; eZ Systems; and Hippo.
• community open source—those that are fundamentally decentralized projects, with no single commercial center of gravity, where small consulting firms and active customers take the lead in governing the platform, often via a formal association. Examples include Joomla, Plone and TYPO3.