[Sigia-l] "Study: Content Management Tools Fail"

[Thomas Vander Wal]

On 2/28/03 8:48 PM, "Beth Mazur" <bowseat at mazur.com> wrote:

> My theory is that the biggest problem with CMS implementations is
> that they fall under the category of what has been called wicked
> problems. Wicked problems are those where "each attempt to create
> a solution changes the understanding of the problem." See
> http://www.cognexus.org/id29.htm for more ...in particular, I
> love this: 
> 
>      One factor that increases the wickedness of a problem is the
>      number and diversity of stakeholders involved.
> 

CMS global enterprise (beginning-to-end organization-wide) implementations
can have serious crosscutting implications.  Each time a stakeholder is
identified (beyond 3 to 5) it always feels like the level of effort
increases by 20% to 50%.  I have found where I have the opportunity to set a
core group of stakeholders (3 to 5) that cut across the various parties
involved (communication/marketing, technology, executive offices, content
creators, etc.) that help set the requirements and only grant others
"wishes" as input to the process things can move along fairly well and there
can be some decent success.

CMS implementations can cause difficulties as they enforce cohesion and
structure, which many organizations are not prepared for.  Many people see a
CMS as putting an uniform on their information and many organizations are
made up of people who do not like wearing uniforms and take pride in their
sub-organization's individuality.  Providing the flexibility to work around
these issues is, at times, where CMS applications and implementations start
getting glitchy and begin to get stress fractures that can lead to failures.

My current job is focused on moving the client toward CMS.  The client has
40,000 to 50,000 hand-built and largely hand-maintained HTML pages.  After I
arrived the budget to purchase and implement a CMS was cut for that fiscal
year and the following.  There are a lot of cultural changes that should
take place prior to implementing a CMS, such as focusing on content creation
in the sub-client offices and not design, bringing user-centered design and
information architecture into the new sub-site design phases, setting
minimum standards, moving toward a controlled vocabulary, etc.  The biggest
problem is their is no central authority to enforce much of anything, the
Webmaster gets over ruled if there are any complaints to the current
development.

> In our third try (which was obviously very informed by our first two
> catastrophes), we felt we had really stated the requirements fairly
> well. In addition, we had decided to implement only the smallest
> set of features needed to try and avoid the dreaded feature creep.
> And best yet, we kept the stakeholders to a bare minimum.
> 
> But even with the effort we put into it, it was only after we had
> a prototype and spent some time using it that *new* requirements
> became apparent. Was this feature creep? Maybe. But it may also
> have been an inability of the client or the vendor to abstract into
> the future what this complex application might really need to be.
> And from my perspective, CMS applications for large organizations
> (mine anyways) are fairly complex, for all sorts of reasons.

These experience-based requirements are somewhat common.  Some of this can
be foreseen through scenarios and role-playing with the goal being to find
incongruence and failure points.  In a CMS everything is connected and
changes to one variable in the system will impact others.  This can be a
time consuming process, but it helps decrease the experience-based
requirements.

This can also be done in the requirements and CM selection phase by building
interactive and logical models based on the CM tool's capabilities.  This
can be done at a high-level or more detailed model, with the more detailed
being better but also being much harder to establish true representations of
the actual processes (detailed modeling can lead to discussions fine details
of a tools reactions rather than focusing on the possible results and
implications -- much like persona are imaginary people not real subjects for
this reason). Role play with the model.  This will often lead to asking good
questions to the vendors and their technical representatives.

I have found Bob Bioko's "Content Management Bible" (ISBN: 0-7645-4862-X) to
be the one book that echoes the experiences I have had and the some of the
solutions I have found to be helpful in the CMS implementation world.  I
have also learned a lot from the Bioko book.

All the best,
Thomas

-- 
www.vanderwal.net

The future is mine, not Microsoft's