A Tale from the Content Management Crypt

SeanceMovieLike all good horror stories, this one starts with a group of normal people engaging in a convivial discussion—one full of optimism and hope. It began as an online discussion about how all content management investments generate some good. There is no such thing, it was felt, as an absolute content management project failure.  Of course, this convergence of positive feeling could not be left to stand. Predictably, such cheerful dreams called forth a virtual Freddie Kruger who could not resist the temptation to intervene.

And thus a post was introduced into the story circle, called “The Perfect CMS Failure.” Perhaps because they assumed that the positive spirit of the campfire discussion would continue, the participants welcomed the contributor into the circle. They were gravely mistaken. To make things worse, the story was a very recent case study and undermining one of this group’s favourite beliefs, this case study saw the very best tools, or at least the most expensive ones, being deployed. It was immediately impossible to relegate this failure to the poor practices of an unenlightened past.

So the story begins harmlessly enough. A senior consultant was contacted by a large government organization that administers the judicial process within a specific jurisdiction. They wanted an independent assessment of their most recent foray into content management. This organization had invested heavily in an enterprise content management system and they disclosed that they had encountered a “few issues.”

This project had started with the usual honeymoon phase. Requirements-gathering sessions were convened and participation was good. This was impressive because many of the users were high-priced, and very busy, lawyers. And a team of high-priced consultants had been flown in from around the world to help with this effort. The requirements snowballed, and before long the team had an impressive matrix of requirements that would be used to select a content management system. Given the nature of the legal environment, many of these requirements revolved around ensuring that sufficient security controls were in place for the content.

content management cryptThis process happily culminated with a CMS selection that coincided with the parallel technical selection process, driven by criteria such as compatibility with the existing infrastructure, team skills, and architecture standards. At this point in their project, things were looking great. The choice of CMS platform would be one of the most expensive available, if not the most expensive, but it covered all the requirements and was especially strong on the security side. As when Icarus flew higher and higher, the project stakeholders started to congratulate themselves even before the roll-out began. In fact, they were so confident that they jettisoned plans for a pilot deployment so they could acquire a full enterprise license and charge straight through to implementation. Like Icarus, they probably didn’t notice the impact of the heat of the sun on the wax on their wings.

The implementation proceeded relatively smoothly, and a surprising number of the organization’s staff, including the high-priced lawyers, participated in workshops convened to refine the implementation details. Usability experts were engaged and worked earnestly to build in support for the users’ needs. External stakeholders were involved to ensure that the implementation was also implemented in a way that would conform to a raft of government standards and edicts on information management. The technology group established processes for performing a mass import of information from legacy systems scattered around the organization and that the new CMS was intended to replace. Training sessions were conducted and reference resources made available to all impacted staff. Although some who had been down this road before might have seen what was coming and given warning, the project stakeholders moved onto their second batch of Champagne as they approached the “go live” date with no indication of the darkness that lurked behind the project dashboard.

True to the horror story motif, it was only on the very day that the new system was to “go live” that problems were noticed. And although these problems had been initially positioned as a “few issues” the magnitude of the problems was in fact, breathtaking.

On the “go live” date, the new system was launched. It contained the master copies of all documents created in the department and that remained of active interest. The new system was the place where all future work within the organization was going to be transacted. It was a “Big Bang” cut over. “Big Bang”, it turns out was an apt choice of words.

legacy systemsOn one level, the failure was in the technical implementation. As high-priced lawyers sat down at their desks they realized that their machines were taking an unusually long time to start up. In most cases, their machines never did start up successfully. It turned out that the CMS client, when combined with the security software that it was bundled with, was simply too much for the desktop or laptop computers to load up. On the functional level, those few users who were able to gain access to the system found the combined effect of all the isolated efforts to optimize the environment for various needs, such as records management, in fact made the workspace a virtual train-wreck of complexity. Finding things, sharing things, using things proved to be as unbearably difficult as holding onto an umbrella in a raging thunderstorm.

Within a matter of minutes, the system was abandoned by its victims, the users, as completely unusable who retreated to the relative safety of legacy systems like villagers draping themselves in garlic. Suddenly the stakeholders, like the high-priced lawyers, raised objections that all their time that had been invested in good faith in the project had been wasted. The cost of this time alone ran to a figure that was comparable to the significant licensing cost for this enterprise tool.

Naturally, the horribly scarred project team contacted the vendor of the tool with a desperate plea for assistance. But the vendor, who might have been named Lector or Hyde in another time and place, said that they were about to retire this product and that the only real option was for the organization to migrate to a new, and completely different, offering. And unfortunately there would not be any credits provided for the acquisition of the new tool. In effect, none of the development investments that had been made would be portable to the new environment.

It was about this time that our intrepid, and innocent, senior consultant appeared on the scene to look into the “few issues” that the project had encountered. On the very first day of this engagement, the consultant, like an Egyptologist interpreting the hieroglyphic curse at the entrance of the tomb, had to inform the CIO and the senior business stakeholders that their project was a “perfect failure.” The investments made, on a per user basis, could not have been higher and the system logs proved that not a single user had performed any work in the environment. The stakeholder community who had invested so much effort and so much hope had, in the course of minutes, splintered into recalcitrant pockets of users using tools and processes of their own design. Because the “Big Bang” had coincided with the turning off of several legacy systems, these users were forced to come up with fall-back positions and post-apocalyptic manual workarounds.

As should have been expected, the CIO postured a lot and pretended to be unsurprised that this business-led initiative had fallen afoul. The project could still be seen as a success from the perspective of the technology group, because they had been able to expend their annual budget on schedule and the technology group team members had successful retired many legacy systems, a metric that the CIO was to be evaluated on.

trappedSo as the dust settled, the meaning behind this failure began to sink in and it was perhaps the most terrifying thing. On the surface of things, it looked like everything was done correctly. All the right specialists were involved. All the right stakeholders had been engaged. But the result was a massively expensive CMS repository solution with no active users and no active content – even though it technically held everything that was important. The failure had been perfect – even sublime. And, in the deep of the night, there is one thought that still awakens the consultant who found himself performing the autopsy and who grew silent as the terrible truth was laid bare. This perfect failure was a result not so much of doing things wrong but of doing too many things right. Like the tale of the body-snatchers, it was difficult distinguish were the guilty parties and the innocent victims (although it seemed pretty clear that the CIO had blood on his hands).

Our protagonist emerged from the experience with a changed view of the world. He had seen failure before but nothing quite so stark and unrelenting. There was simply no room left for optimism or hope. This consultant would forever after be unable to participate in discussion circles where this optimism and hope were the main currencies. And as expected, the introduction of his story into the original discussion circle led to complete silence. Suddenly cheerleading seemed out of place. And the walking dead moved on in search of new hope and perhaps new victims.

Joe Gollner

Joe Gollner is the Director of Gnostyx Research Inc., a company he founded to help organizations come to grips with their content in ways that are sustainable and scalable. He has led over 100 content management initiatives during the last twenty years, with several of these representing the high water mark in the industry for sophistication and scale. Joe also posts minor tomes on the nature of content through his blog, The Content Philosopher (www.gollner.ca).

Read more articles from Joe Gollner Twitter