“Where’s the disconnect between what’s possible and the too-often failure of CMS?”
Content management systems suck. Or so you would think from the strife heard from analysts and practitioners alike. And yet, many websites regularly publish vast amounts of information with superior control and ease compared to manually editing pages. So where’s the disconnect between what’s possible and the too-often failure of CMS?
Our experience with content management systems
To discover what kinds of problems people are having, I surveyed other practitioners. The survey was conducted in late January 2003.1 Participants of Sigia-l, AIfIA, and the ia-cms list were invited to participate. A total of 64 responses were collected. Figures 1 and 2 summarize the results. You can also view the complete survey results online.
As you can see, the issues are many, spanning strategy, design, content, technology, training and several others. One conclusion we can make is that content management has become a very large category—attempting to include content authoring, metadata authoring, database-backed websites, workflow management, and even thesaurus management—and instead of making CMS a goal you might start by focusing on which of these functions you need. Otherwise, the general complexity becomes the central problem facing any content management project.
Martin White, a CMS consultant and writer, writes in EContent magazine, “A CMS is probably the most complex rollout you and your IT colleagues are likely to have to manage.” This is because content management most likely requires contributions from many different skill sets and coordination across diverse departments and roles. A CMS project can cost hundreds of thousands or even millions of dollars and require months or years to design and implement. Because of the high planning, purchasing, and design costs, there is a need to effectively manage the complexity of CMS projects. I’ve seen some organizations do this well and others not so well.
Here are ten lessons in managing complexity gleaned from real-world, successful CMS projects. Ideally you’ll consider these at the beginning of a project when they can have the most impact:
- Keep the team small.
A big team usually requires a lot of coordination and communication, especially if it is spread across different departments, offices, or cities. This coordination increases the points of failure and quickly reaches a level of diminishing returns for systems that need close collaboration to be designed well.
To overcome this problem, one financial services firm formed a multidisciplinary team of only five experienced people to create their content management system. The team included people who both had skills to contribute and could make executive decisions. This team consulted with additional, specialized staff only when needed. In the end they succeeded in building a system in a few months in a company where other efforts typically spent several months and failed.
- Don’t try to fix everything at once.
A CMS alone is complex enough; combining that effort with a site redesign, new workflow, new content, and more may be asking for trouble.
An international retailer decided to expand their content management system in a way that required multiple new software packages. To reduce complexity, they swapped in the new CMS without changing the design of either their online or print catalogs.
- Only build what you need.
It’s important to remember the main benefit of content management systems is efficiency. Anything done with a CMS can be done without a CMS by people with the right skills, albeit in much less efficient ways. Websites often use content management when there is a large volume of content, frequent content updates, content distributed across several media, and so on, tasks that would be arduous when done manually. But if more effort is needed to implement a CMS than to manage the content manually, the return on investment is quickly lost.
The potential features of CM software spiral out in all directions, so discipline is needed to decide which features are needed most. At the beginning of a project we can examine which features bring us the most benefit compared with how difficult it is to implement those features, and choose the features with the most value.
A new media group at a bank took this approach and built the absolute simplest CMS that would serve their needs, and then gradually added one feature at a time as the need became clear.
- Create an efficient information architecture.
A content management system with a different template for every published page would not be very efficient. And if efficiency is the main benefit of a content management system, then it makes sense to use fewer templates. As designers, we must be very clever about how to arrange diverse information into a small number of templates while still retaining some flexibility in the presentation.
A large technology company achieved this efficiency by creating templates as well as reusable modules of information—such as a list of related links—that fit inside those templates. By creating rules that determined how templates could use certain modules the company struck a balance of CMS efficiency with display flexibility.
- Show your content some love.
Of all tasks in a content management project, the creation, editing, and migration of content are probably the most frequently underestimated on the project plan. The survey above reveals this void as the biggest problem with CMS. Amid much sexier design and technology issues, the creation and/or re-formatting of content can be delayed until this eventual necessity delays the project.
To counter this problem, one non-profit organization settled on an article layout at the beginning of a project so it could start preparing the content earlier, then continued the content work in parallel with the design and technology work.
- Hire bouncers as project managers.
Perhaps this is going a little too far, but you do need rigorous project managers that understand CM issues who will babysit the team to make sure every little task is getting done. These project managers must do more than make sure documents are delivered on time, they must help connect the work that all team members do.
One large retail firm took this to heart by using two project managers: one to oversee the business and user interface issues and another to oversee the technology issues.
- Tightly integrate design and technology.
Content management software involves certain components, such as content entry screens, that require a combination of interaction design, information architecture, writing, and database programming skills. Few people do all these things well, and having different people or groups design these components in isolation risks poor quality and consistency.
My smoothest experience designing these components was when my desk was located right next to the programmer’s desk and we constantly discussed the design as it evolved.
- Buy the right size.
In the survey cited above, the number one problem with software is the expense. You might think the solution to this problem is to buy a less expensive software package, but I think a better solution is the buy the right size software package.
Tips for choosing the right software:
- Buy small software if you’re a small organization. Organizations like Boxes and Arrows, the Asilomar Institute of Information Architecture, and Adaptive Path all use Movable Type to manage content, which was originally designed for weblogs. As content management software, it doesn’t provide many basic functions, but it simplifies the publishing process enough for occasional publishing needs.
- Buy big software if you’re a large organization. One big CMS can actually be more efficient than many different, smaller packages. One financial services firm employs a federated model of CMS by using one software platform to publish many websites, avoiding the extra training and technical work needed to work with several different software packages.
- Buy no software at all if you really don’t need it. In the decision to buy vs. build, we can also avoid software all together.
- Design faster than business can change.
You must design and implement your system faster than your organization can change.
For example, a large computer networking company found that it required three years to roll out a new website design across all its departments and websites. Before they could finish, the organization had undergone significant changes that needed to be reflected in a new site design. Designing fast may mean keeping the scope small, but it can also mean finding innovative approaches to problems rather than simply following conventional methods. Building a Metadata-Based Website is an example that speeds design of very large sites by focusing management on the business concepts instead of the content.
- Get a second opinion.
Content management is an elaborate, dynamic field and there are several solutions to any problem. Just as in medicine, it’s sometime necessary to get a second or third opinion to hear approaches that arise from different philosophies.
One international retail company brought in a CMS expert to consult to the team without doing any of the work herself. As an expert who didn’t work for the company or any of the contracted vendors, she was in a good position to provide impartial guidance. As a disinterested third-party, the expert can help smooth interaction within the team while leveraging experience from previous projects.
Now you have a list of problems others have had and ten ways to address the complexity of content management. Will following this advice solve your CMS problems? Not entirely. If you’ve only heard the hype from software vendors, you may have very high expectations that need to be reconciled with reality. Content management systems are not a silver bullet, but they can make your most onerous tasks more efficient if you actively manage the inherent complexity.
- Study: Content Management Tools Fail
- Sigia-l List Discussion
- AIfIA Survey Results
The Problems with CMS
- Managing Content Management System Selection
- CMS Watch
- CMS Review
- The Problems with CMS
- Buy Build Avoid
- Building a Metadata-Based Website
- This article is derived from The (Unfulfilled) Promise Of CMS, an AIfIA seminar presentation. I’d like to thank the organizers and participants of the survey and the seminar for contributing to this article and especially Vivian Bliss, Liz Danzico, and Chris Farnum for their assistance.
- The survey was conducted January 24–31, 2003