Information Today, Inc. Corporate Site KMWorld CRM Media Streaming Media Faulkner Speech Technology DBTA/Unisphere
Other ITI Websites
American Library Directory Boardwalk Empire Database Trends and Applications DestinationCRM Faulkner Information Services Fulltext Sources Online InfoToday Europe KMWorld Literary Market Place Plexus Publishing Smart Customer Service Speech Technology Streaming Media Streaming Media Europe Streaming Media Producer Unisphere Research

Magazines > Online > January/February 2003
Back Index Forward

Online Magazine
Vol. 27 No. 1 — Jan/Feb 2003
Content Management's New Realities
by Stephen E. Arnold

Content management touches virtually every aspect of an organization. The shift from proprietary network architectures to the Internet means that text, PowerPoint presentations, customer support records, and audio-visual data have to be managed. Many books have been written in an effort to define management. No single author has nailed down the concept. Like most abstract nouns, management means everything and anything associated with organizing people and their activities.

In the best tradition of synthetic phrase construction, stick content in front of management. The result is a bound phrase, in the parlance of linguists, that takes its meaning from the context in which it is used. To figure out what someone means when the content management phrase is tossed around in a business meeting, you have to understand technology, motives, experience, and what the problem is that makes a buzzword like content management look as if "it" is the solution.

When the needs and requirements are understood, a system to manage the creation, approval, and dissemination of text, images, and even streaming video can make life in today's fluid environment somewhat more orderly. When a system is matched to technical requirements, the likelihood of that system "working" goes up. Money, attention, and support are essential ingredients. Omit any of these prerequisites, and problems will abound. Of the three, attention is by far the most important. This might make you groan, but consider that attention may be the most expandable of the ingredients.


Content management is a relatively new software sector. Some companies in the CM business emerged from records management. Records management is usually associated with keeping track of financial, health, pharmaceutical, and engineering information. There are "destroy by" dates and retention policies that are often mandated by law, just as the U.S. Internal Revenue Service expects 5 years of records for individuals.

Another driver of CM has been the Web. Webmasters—often accused of being a bottleneck when it comes to moving content from the desk of the creator to the organization's Web site—want a system to get colleagues off their backs and onto a unified content creating and publishing system. Paper output and video streaming are of less importance than meeting the growing need to get content written, approved, and out to a Web site. Rollback functions, security, and enterprise integration come later in the process. Webmasters have a problem, and hundreds of software companies have come into being to solve this problem. A good list of content management companies is located in Google's directory [


Managing the facts, figures, dates, relationships among the data, and the business processes is a difficult and time-consuming job. Figuring out who, what, why, when, where, how, and under what circumstances is a procedural tar ball. There may not be a single way to figure out answers. The way organizations work is often mysterious. In some enterprises, employees will admit that it amazes them that their operation produces products and services at all. Life is often filled with irrationalities. As the popular writer Tom Clancy said to Larry King, "The difference between reality and fiction? Fiction has to make sense" []. Much of the discussion about content management and content management systems (CM or CMS, in the computer trade acronym) is real.

CM affects you, your colleagues, and those using the information you and your organization provide online, in print, on CD-ROMs, and as broadcasts or audio programs. The realities of CM can exact some interesting lessons in budgeting, technology, human relations, and organization dynamics. This brief article highlights six real-life lessons driven home in content management projects over the last year. Others' experiences may be different.


At first glance, computer users and programmers alike generally understand content. Data are digital. Anything in digital form is content. Content in digital form is "managed" in files, folders, directories, and other standard-issue information technology services.

What's the problem? In a word, boundaries. As long as the data are on one personal computer or one computing device, the management problem is keeping file names straight, avoiding version confusion, and having a backup. Until the issues of "ownership" and "responsibility" came up in trying to figure out who must approve what and when before it goes live on the intranet versus live on the public Internet site versus who can see the data on the extranet site, there was no need for fences, partner. CM makes fences mandatory. Cattle ranchers prefer wide-open ranges and no fences, thank you.

The tension between ranchers and sheepherders in the Old West resurfaces in cubicles throughout an organization. Adding in the idea that a new product news release must appear on the Web makes content management jump to another dimension. The phrase "content management" is a dirt street where different organizational warriors fire memoranda at one another. They pause long enough to reload their keyboard and Blackberry. Financial officers and Webmasters duck for cover. Some of those memoranda can do serious damage.


Let's shift from the Web to a more controlled content domain. Customer service, or what is becoming euphemized as a "contact center," is becoming Web-centric. The idea is that a customer can use a Web browser to ask a question about a problem. To some visionaries, voice-over Internet protocol (VOIP) allows the Web to blend voice, browsing, and e-mail to create a "presence" for the customer. Some customers will click to a Web page, then look at a Yahoo!-style listing of topics or enter one or two words in a search box. Others will make a voice call that will link to a Web-centric system that handles multi-mode content. (This means voice, e-mail, Web page content, and facsimile messages. Very visionary, very grandiose, and, as it turns out, very expensive, whether the system works or not.)

On the surface, this seems like a simple job. Customer service is contained, although accounting and marketing are involved with some issues. The content is usually available in the form of printed material generated from word- processing files. Some companies have created online systems to hold frequently asked questions, answers to common problems, and various bits and pieces of customer information.


Data has begun to come to light from such firms as Nucleus Research Inc. (a consulting firm in Wellesley, Massachusetts), user group meetings, and online discussion forums about customer support software. One startling fact reported in Computer World [September 30, 2002, page 7] is that the average project cost is about $6.6 million. This figure brought to mind the aphorism "The early worm gets eaten by the bird." The bird in this case is the vendor of contact center software and systems.

Experts in "pure" content management have reported similar financial "news" when top-tier systems are deployed in an organization. Jupiter Research said in June 2002, "Some companies spend a hefty $25,000 per non-technical employee per year to manage simple content on a Web site. In many instances, customization, integration, and deployment costs can rise as high as six times the basic licensing fees." A company with 50 people in the customer support chain would translate to $1.25 million, excluding software license fees.

Despite the costs, Jupiter opines, "According to an April 2002 Jupiter Executive Survey, 53 percent of companies will have deployed new document, content, or media asset management systems by the end of 2002. Moreover, almost one-fifth of Web site managers (19 percent) said they will be involved in content management consolidation projects as they unify systems to manage multiple Web properties."

A quick review of the marketing collateral available on the Web sites of the content management leaders suggests a very different story. The payoff from content management can be a savings of 20 percent or more compared to pre-content management systems. Customer satisfaction rises. Even turnover in among what are often low-pay jobs is reduced.

The license fees for content management systems range from a few hundred dollars per month ( for a hosted content management system to more than $1 million for a license from Documentum, Vignette, MediaSurface, or similar company. Atomz, like most ASPs, gave up on the low-end market and now sells a mid-market product for about $2,000 a month.


When the silver-tongued marketers of content management systems find their credibility tarnished, the pitch is shifted. Recent examples may be found in the software marketed for portal building. Often these products are described as "portal toolkits" or "application servers." Poke around among the options offered by such firms as IBM (WebSphere), BEA Systems (WebLogic), and Sun Microsystems (iPlanet), among many, many others. These toolkits and servers are really modules of software that allow a programming team to build a content management, customer relationship management (contact center), or knowledge management application.

The firms selling enterprise software are playing a bridge game. Each new "word" used to describe a function performed by the software is really intended to trump what other vendors' have put on the table. A company with a gone-wrong CRM system will find that the vendors of WebSphere, Documentum, or Oracle will pitch content management or some other buzzword to solve the problem. With every new buzzword and compound noun containing the word management, the number of things most professionals know nothing about increases. The object of the software game seems to be sales through confusion.


If we narrow our focus to the problem of a three-person department trying to create and post new content on a single Web site, we can identify the basic functions a content management system must have. These functions are, at a minimum, the following:

Check in and check out features. This allows an authorized user to create a document or a graphic and put in the system. Once in the system, a mechanism is activated that keeps a copy, allows or denies access to some users, and makes the document available to the Web updating system.

A graphical editor. This is software, often Dreamweaver, FrontPage, or a similar tool, that authors and designers use to create templates, Web pages, forms, or other elements of a Web site.

A library. Templates, graphic objects, and other bits and pieces of a Web site are stored for reuse.

An uploading mechanism. Once a single Web page or a complete site is ready to go live, an easy-to-use function performs the file transfer process.

These basic functions are often suitable for a small Web site and a small number of users. This type of functionality is available in products and services available from desktop authoring systems (Macromedia, Microsoft), "blogging" packages (blogger, manila) and low-end, server-based Web CMS packages (Ektron—and literally hundreds of other companies—for an annual cost ranging from $300 to about $7,000. Types of variations in pricing are associated with each customer's security, storage, and system requirements.

If we take the $7,000 range and triple it, we have a target budget for a basic content management system in the $20,000 range. This begs the question, "How does one go from $20,000 to $1.25 million, or the $6.5 million figures reported by Nucleus, Jupiter, and annoyed newsgroup posters." These are jumps of 62.5 percent and 325 percent respectively. Numbers with variances like these carry the Web team into Enron and Global Crossing mathematics—not a comfortable place for people trying to bring order to a Web site.


The realities of content management are just now becoming visible. Readers embarking on a content management system for their organizations can avoid most of the financial "shocks" associated with enterprise software by looking for and avoiding some pitfalls. The balance of this article identifies what I have described as "new realities." The old reality, of course, is that digital content is difficult to corral when more than one or two people have access to the documents, Web site, and software.

We Are an Information Company

Everyone in a "with-it," today organization is involved with information. This means that many people can write, change, authorize, de-authorize, design, update, and influence others with regard to content. When three people are involved with a Web site, the maximum number of interactions possible is six. What happens when there are eight involved, for example, a content creator in marketing, the content creator's manager who must sign off on the copy, a designer, a product manager, a lawyer responsible for product announcement reviews, accounting responsible for pricing statements, a programmer, and a technician at the hosting firm where the site resides? The total number of interactions is 40,320. Not all of these interactions take place, but enough do to make the analysis of who does what to whom and under what circumstances time consuming and expensive to figure out. If these interactions have to be converted to the type of programs that drive workflow processes in the content management system, get out your checkbook.

The License Fee Is Only Part of the Cost

A content management system is a collection of functions. In fact, the software that makes up a content management system must be set up. Like most set-ups, the system administrator must have answers to basic questions. Some of the questions are technical and focus on the IP addresses for the Web server, the specific security settings for specific folders, and the location of specific files. Other questions are fuzzier. "Fuzzy" questions include:

• For Approval Process X, what people are in the approval chain? What amount of time is allowed for Person G to act on a document created by Person F?

• When a rollback notification comes, what priority does the rollback have? If the priority is High, what processes are stopped to permit the rollback? If the priority is Emergency, what authorization is required to take the server offline and restore the previous version of the content? How does the operator on duty determine if the rolled-back version of the content is acceptable? Who is in the rollback verification approval chain?

• When a price list is to be repurposed for a trade show, what is the update cycle for the Web site? Reprints of the price list if there is a change? Who has the authority to change prices during a trade show if that person is not in the normal approval chain?

These are process questions, and they are very expensive to answer. The reason is that most approval processes are decided on the fly. Most content management systems with workflow or work routing systems are driven by rules. The job of converting ad hoc processes into rules can be a tough one. Once the rules are set up, some specialized expertise in programming is useful. Tweaking rules can be done by experienced professionals. Tweaked code must be quality checked or the automated process will not work.

"People will try to circumvent workflow systems that don't help them do their work properly," notes Tony Byrne, founder of CMSWatch [], "and sometimes they succeed." Byrne recounted how one of his clients, a major U.S. cable television company, discovered that staff had been FTPing new content directly to a production server 6 months after the implementation of a new CMS. Byrne quickly discovered what the frustrated conspiracy of staffers had recognized on the first day of the new system—the workflow encapsulated by the technical administrators didn't faithfully reflect how the editorial team actually needed to work. So they found a way to bypass it. Most such attempts fail, but that's not the point. Organizations are indeed "organic" and highly unique creatures; CM systems need to reflect this, which can take more money and attention than first seems apparent.

CM Systems Collide with Other Enterprise Software

At some point in the future, software will output Extensible Markup Language. Agents will make intelligent decisions about metadata. Database systems will seamlessly read the tables created by database systems from other vendors.

Meanwhile, different enterprise software vendors are fighting for market share and doing their best to "lock in" customers. In the days of proprietary software and hardware, "lock in" was automatic. The phrase "no one gets fired for buying IBM" has yielded to Open Source. Governments in China, Germany, and South Korea are embracing Linux. The mantra is open source, standards, and technologies with cryptic acronyms. The goal is to wrest control of hardware, software, and systems from those with proprietary systems to gain maximum flexibility.

The goal is cost reduction and competitive bidding. Reusable code replaces customized code. Just as a shirt from Wal-Mart is less expensive than a shirt from an Oxford Street tailor, open source promises to slash costs.

A small-scale content management system from eMojo (London, England) using Cold Fusion and low-cost Intel-based servers may meet the need of a department, a mid-sized hospital, or a trade association. Even if eMojo would scale to handle a company the size of Toyota or a government department on the order of the Department of Agriculture, it would be headed for a battle. eMojo would probably have a difficult time winning if IBM, Oracle, FileNet, SAP, or a similar enterprise software vendor wanted to capture eMojo's redoubt.

The financial stakes are too high for the major companies to concede market share and revenue to small companies gaining momentum. The brutal fact of Darwinian marketing is that the big boys will get "lock in" one way or another. Once the small and annoying beasties are eradicated or neutralized, the big boys will have to fight or buy one another to keep revenues flowing.

Enterprise sales offer promises of big economies of scale. One way to get these economies is to roll up the smaller software, systems, and hardware into the larger system. This type of efficiency allegedly reduces costs. The idea behind "shared services" is one that opens the door to a higher-stakes poker party in which the pot is revenues from companies that standardize on WebSphere, WebLogic, PeopleSoft, or some other enterprise solution.

CM Means Organizational "Contention"

There are in most organizations two or more software systems that are able to provide content management services. One software environment is the database engine running the major back office systems of the company. Another is the framework used to build the company's information technology architecture. Most purchasers of content management systems want to get a handle on a Web site. But the database providers (IBM with DB2, Oracle with Oracle, and Microsoft with SQL Server) want to make certain their database engine drives as many data services in the organization as possible. Therefore, there will be database "forces" getting involved in CM systems that reach beyond one small department. Companies offer complete software toolkits and engineered architectures that "snap together" and "talk" to one another without any extra programming. Vendors riding this train include IBM (Web-Sphere), Microsoft (Dot Net), Sun Microsystems (Sun ONE, iPlanet), BEA Systems (WebLogic), and recently SAP (R/3 and

CM is essentially a subfunction in a framework. Many content management teams come to work after a pleasant evening at home and discover that their departmental project has become an enterprise initiative. Control of the project's scope, the budget, the hardware—indeed, everything associated with the project—has become larger, more important, and more complex. In one administrative eye blink, a manageable CM project has become a footnote to Parkinson's Law. The costs soar by orders of magnitude instantly.

No One Knows What Content Is

Content for a Web site used to be easy to define. It was text with markup language. Today, a Web site consists of text plus "objects." A Web site also contains scripts and code. Without the programming, the Web site may not exist at all. Welters Kluwer [] provides an example. Each Web page is created on the fly. There is no content in the 1994 sense of the word.

The definition of content comes slowly and breaks along the consciousness of those involved with the CM project. Content is a process and it generates objects, such as these:

• Versions of brochures, price lists, documentation for staff, and defined benefit plans.

• Pictures in a wide range of file formats with different copyright requirements, linked cutlines, and metadata about the date, the photographer or artist, and previous uses of the image (for example, used in the 2002 annual report and the briefing to Morgan Stanley on November 3rd).

• Graphic objects designed and stored in a proprietary file format as well as a file format suitable for Web use plus the metadata associated with the object as well as versions of the object.

• Audio and video files with metadata about copyright and previous use.

• Database files containing data. Some of the data are usable by certain employees. Some of the data are usable by the public. Some of the data are to be purged in accordance or not in accordance with regulatory or legal guidelines. (Think Enron and Andersen shredding actions applied to digital files and driven by scripts.)

The reality is that as soon as the concept of content becomes understood, significant time and energy are required to figure out what to do with data now stored, repurposed, and available to anyone with administrative privileges. What happens in most organizations is that there is a Berlin Wall between and among content. The Web site is the catalyst for a period of understanding the implications of managing data that are, well, evidence.

CM Is Dynamic

The most difficult reality for a content management team to grasp is the fluid nature of CM itself. Content management companies are not helping. Documentum is gobbling up companies that add enterprise software and workflow functions to what was a records retention code base. IBM is buying companies offering functions to extend WebSphere into every nook and cranny of business process functionality. Microsoft is using its installed base of desktop software to migrate to departmental servers and ultimately to enterprise servers, shifting from office applications to business processing applications in what is to Microsoft a natural evolution.

These organizational actions translate to CM companies being forced to move into other, unrelated software areas. RedDot [] said in a recent announcement, "March of Dimes implements RedDot CMS to manage online content and workflow." Some CM companies will merge. Others will be acquired by larger firms. Many will fail. Most surprising will be the companies known for other things (manufacturing servers, selling programming tools, or marketing desktop applications) that instantly become content management companies. Equally confusing are e-commerce companies such as BroadVision that become content management companies in a frantic effort to find a way to generate revenues. The mantra is, "Make sales. We'll figure out the software later."


Content management software is plagued by people like wearers of sunglasses who believe others cannot see them. In Africa, a professor repeated this Ashanti proverb to me: "Only when you have crossed the river can you say that the crocodile has a lump on his snout." For those wanting to embark on a CM project, follow these steps:

1. Do a thorough needs analysis even if there are only one or two people in your organization.

2. Write down your requirements and discuss them with at least two people who have built and used a CM system.

3. Start small and with a CM software that allows a free trial or has a low entry cost.

4. Assess your CM system after 30 days and make a list of what is right, wrong, needed, and unnecessary.

5. Repeat the process, stepping up in functionality and cost based on your "crossing the river."

When it comes to CM, cross a small river successfully. Crocodiles rarely exceed 6 meters. The snouts give them away. And mind the teeth.


Six New Realities of Content Management

1. We are an information company.

2. The license fee is only part of the cost.

3. Content management systems collide with other enterprise software.

4. Content management means organizational "contention."

5. No one knows what content is.

6. Content management is dynamic.

Author's Note: This article is extracted from the chapter about content management in my forthcoming book, Knowledge Management Sense and Nonsense. It will be published by Infonortics, Ltd. (Tetbury, U.K.) in early 2003.

Stephen E. Arnold [] is president of Arnold Information Technology (AIT), based in Harrod's Creek, Kentucky.

Comments? E-mail letters to the editor to
       Back to top