Content Management Q&A; with Matthew Berk

To help sort out the enigmatic world of content management, Intranet Journal picked the brain of Matthew Berk, a senior analyst with Jupiter Research who covers content management and site technologies. (Jupiter Research is a division of Jupitermedia, the publisher of this site.)

Berk’s expertise includes search technology, staffing and spending, Web services, usability, site analytics, content management, Web-enabled applications, and the history of the Internet.

Prior to joining Jupiter Research, Berk was the Chief Technology Officer for edu.com, an online retailer specializing in the higher education market.

Previously, he was a co-founder and Vice President of Technology for inc.com, an aggregator of best-of-breed content, information, products and services for the small business market.

On May 21, Berk will host a free Webinar “Web Content Management: Covering Essentials, Avoiding Overspending.” To register, or for more information, visit: jupiterwebinars.om.

Q: Among companies that plan to replace their existing systems in the coming year, Jupiter Research found that less than one-third are willing to look at another commercial package; the remaining 82 percent said they would build their own solution. Why have CM systems become the exemplars of bad technology investment?

Content management systems are not necessarily bad technology investments; rather, purchasing organizations all too often misunderstand what the technology can actually enable, and at what cost.

The central myth of content management is that it can free non-technical content owners from the constraints of IT.

This attempt to solve an organizational problem with a technical solution is often the root cause of content management dissatisfaction.

Jupiter recommends that companies always lead with organizational issues, rather than feature lists or technical specifications (what we call RFP “bloat”).

Again and again, Jupiter has seen scenarios where low-cost alternatives such as human HTML production; simple, Web-based forms; or simple products like Macromedia Contribute would have worked equally well, if not better — in terms of cost effectiveness and end-user empowerment — than high-end solutions that cost upwards of hundreds of thousands of dollars.

Q: What have leading vendors done to change the image of CM solutions as high-risk investments?

In our opinion, the leading vendors are neither the recognized content management name brands, nor the companies with the most buzzword-compliant technology.

Content management technology has basically become a commodity; this means that larger vendors, in their attempt to preserve market leadership among an ever-widening group of competitors, have accreted functionality that may have little to do with core content management needs.

The most positive influences in the market come from the strong, mid-market solution providers, who understand that the final value of a content management system is the extent to which it can enable non-technical content publishers.

These providers include companies like Atomz, Percussion, Red Dot, and Stellent. In our experience, these companies can help their clients hedge implementation risk through rational costs (generally sub $100k) and proven time-to-deployment measured in weeks.

Q: Content management has a very fluid definition, and now the buzz term “enterprise content management” has been added to the mix. Is there such a thing as an enterprise content management system?

Enterprise content management is a strategy, not a solution. It defines an approach to managing information resources to help enrich the relationships between employees, partners, and customers.

Contrary to what vendors may pitch, ECM does not mean bringing enterprise-wide content assets under unified, central management; nor does ECM merely represent the sum of WCM (Web content management) plus DM (document management) plus DAM (digital asset management).

Rather than talking about ECM as the centralization of content assets, Jupiter has defined Federated Content Management (FCM) as the ability to discover value between pre-existing, disparate content management systems.

For example, consider a media and entertainment conglomerate with many brands, lines of business, divisions, etc.

Using the federated approach, employees can discover and repurpose content from multiple sources: images stored in Canto, videos managed in an IBM system, product content living in Stellent, and reuse rights living in a custom database.

Q: Is effective enterprise content management really feasible given different workflow requirements and roles within departments at large organization?

There are two issues here: access to assets, and workflow. The FCM vendors who are best poised to enable the vision of enterprise content management have invested heavily in preserving access structures across multiple systems (who can see or use which assets), but often argue strongly against duplicating the workflows of the systems they unify.

True enterprise content management by definition exceeds the limited use cases an individual system has been designed to support.

Jupiter has spoken with many companies for whom the very structure of their many lines of business (competing record labels, for example) works against the grain of cross-divisional cooperation.

Q: The build vs. buy debate comes up often in content management because some organizations have very specific needs. If an organization has the people and resources, is building a content management system the best way to go?

There are two key considerations when making a buy/build decision for content management.

First, a company with well-defined, highly granular requirements that do not change dramatically over time is likely better suited to build their own solution.

If, however, requirements are not well understood, or the requirements are likely to evolve in the near term, then commercial software, with its support for generic content management constructs, may be more suitable.

Second, organizations need to be honest about their own IT culture in making the buy/build decision.

If the IT group has a proven track record of creative, cost-effective problem solving that meets the requirements of business users, then a home-grown solution is likely to succeed.

On the other hand, antagonistic business/IT cultures, with more of a regimented development process are likelier environments for commercial software.

Unfortunately, these are also companies who tend to respond best to misleading vendor rhetoric around short-circuiting IT (the technical solution to the organizational problem, as mentioned earlier).

A Jupiter cost model demonstrates that an IT group not capable of significant traction within 8 to 10 weeks should instead be considering a solid mid-market solution.

Q: There are a number of open-source content management systems that will power everything from blogs to enterprise sites. What type of effect are they having on closed-source vendors?

The open source content management world is highly active; packages like Cocoon, Zope, and Bricolage solve many of the same problems as their commercial counterparts.

Jupiter has seen a far greater success-to-nightmare ratio among open source adopters than among users of commercial software.

Ultimately, open-source CM broadens the list of implementation options, contributing to downward price pressure in the commercial market.

More importantly, it will force savvy vendors to componentize their offerings to the point where the commercial offerings can compete — and win — against free alternatives. Userland, for example, has learned this lesson well.

Q: Someone made a comment at AIIM 2003 in New York that because people within an organization constantly make adjustments to workflow and have varying degrees of organization themselves, most people will find themselves at odd with many of the complex content management systems, and either not use many of the features, or not use the system at all. Is what the industry tells us is effective content management at odds with human nature?

Workflow is by and large a red herring. Process is everything. Attempting to create a working process by simply defining workflow within content management software will generally fail.

Workflow was originally intended to help manage and put the breaks on hyperactive content submission in the early days of the Web, and was — and remains — largely heir to software development metaphors like check-in, check-out, versioning, branching, state change, etc.

In a recent report on content management workflow, Jupiter found that companies who begin with highly scripted workflows invariably strip away the steps one by one until just three generic steps remain.

Most process can never actually be modeled because people always need the flexibility to step out, look left and right, and step back in.

Jupiter recommends developing simple, generic, roles-based workflows that support pre-existing (read: functioning) process.

Q: Content management is a crowded field, and when you add in the numerous open-source projects the number of solutions is nearly impossible to count. What do you think the market will look like five years down the road?

Today there are more than 200 vendors in the Web content management market alone.

Within the next three years, Jupiter expects this number to double due to the relatively low barriers to entry, and the fact that most new entrants are simply trying to “productize” what were previously service models (e.g., from small interactive agencies or development shops).

In five years, content management functionality will move in two directions: out to the desktop in the form of software like Office 11, and down to the infrastructure in the form of file systems that implement the essentials now seen in content management packages.

Q: What will the companies that survive have in common?

In the near term, the companies that thrive will be those that provide proven, componentized offerings with sub-$100k price points and time-to-deployment measured in weeks.

Jupiter also sees great longer-term opportunity for hosted content management providers like Atomz and CrownPeak, who insulate their clients from the details of managing complex software.

Q: Many of the names that have dominated software and computing, such as IBM, Microsoft and even Sun Microsystems, have recently been trying to position themselves as players in the CM market. Do you think this will have much of an effect on the market, or will products from such companies appeal primarily to their existing customers?

Jupiter believes that infrastructure providers like IBM, Microsoft, and Oracle will ultimately define the fate of content management, but not by providing specific solutions.

By promulgating the adoption of XML and Web services, pushing structured content to the employee desktop, and weaving content management functionality into the very fabric of the file and operating systems, they will shift the content management scene away from the risks and anxieties of implementation and towards its ultimate promise: leveraging content to enrich the relationships between customers, employees, and business partners.