Application Architecture for the Enterprise

WCM: Bone In The Throat of ECM

A comment that I left on James' blog awhile back has kept me thinking on the topic of web content management (WCM) and its relationship with enterprise content management (ECM). Others will surely disagree, but I think ECM as a meme just about owes its existence to its subsumation of the WCM domain, however tenuous that is/was. In fact, it seemed like WCM had barely gained its foothold, and along comes ECM knocking it to the side - score 1 for the ECM meme wonks.

Large enterprises had the notion of traditional document management (DM) well embedded, where things like workflow and records management excel. But the outlier was always web content management - that curious mix of kind-of structured content, very tied to it's distribution mechanism. WCM boasted it's own market segment and product portfolio, but its no surprise that to fulfill their destinies, DM vendors had to step up.

But how well has that gone? Tenuous at best, I would say. It's one thing to fill out a portfolio with acquisitions and update the PPT deck. It's another thing to deliver on the promise of unified content management.

Have the ECM vendors really won over the original leaders in web domain - the web designers? When was the last time you heard a web design team stoked about the latest ECM platform? The fact is that WCM product integration seems to be largely about foisting DM concepts onto the web domain, rather than building in the value-add features that web designers and developers need. (Note: this is really a retrofit value-add, as they have these features outside the ECM world - they lose them on moving day.)

There is a paradigmatic gap between the non-ECM world of web content management and publishing, and the ECM world. Web designers spend most of the early phases of adoption trying to shed baggage that will burden them down in the ECM world, and trying to un-learn all of the techiniques they had previously used to templatize sites (thing like server-side includes, iframes, etc.) They often need to port these same concepts to a new implementation that is ECM-friendly, to gain the "benefits" of enterprise quality. And have you heard a web designer brag about how great it is to change HTML templates in their favorite ECM tool? Templating is a whole topic on its own.

The nature of web content means that many ECM (nee DM) concepts do not apply. The world of DM is organized around a single artifact - the document. Sure, you manage relationships, but they are secondary to the artifact, and often relate more to packaging a group of artifacts for management reasons. The world of WCM demands that you manage related assets for distribution reasons - a subtle but important distinction. Web content is all about interrelationships - linking, styles, etc.

A bow to the vendords - it's no small task to try to productize the WCM domain. The fact is that there is a great continuum of web sites and web content. You will not have the same requirements nor solution for a simple, static brochure-ware web site that you will have for a dynamic, personalized information portal. A tool that works well for one will likely not work well for the other, which is a fly in the ointment for the E in ECM.

But that notion itself is predicated on the flaw in typical enterprise thinking - that centralization is automatically a Good Thing. The ECM sell often has something to do with single-sourcing. When there is a business case behind single-sourcing, it is a huge boon. But without a business case, it's an enterprise boondoggle.

Many WCM vendors or offerings try to growup to be ECM, but don't catch up with compliance oriented features. Compliance is critical to many businesses, but there are often process solutions and bolt-ons that can address that - I say hold the course on WCM. There is no shame doing what you do, and doing it well! Tony notes that everyone has to compare themselves to Vignette... I say be yourself.

posted by scott @ 6:36 AM