“Why laugh and grow fat when you can experience anguish and success in a straight jacket?” -Marshall McLuhan
I wrote the passage below as a reply to the question “Open Source vs Proprietary [enterprise] CMS: why go either way?” in a discussion at LinkedIn on this topic. I’m re-posting it here after requests from some who do not belong to the group and have no interest in joining.
It has been my experience that in every vendor selection I’ve participated in at the enterprise level proprietary CMS have always won out over Open Source CMS. OS CMS tend to suffer from absent the robust sorts of functionality that are found in proprietary enterprise-level CMS that these customers seek: Easily and highly configurable workflows, group management, search, deployment, multichannel support, and scaling and performance. I’ll expand on each of these specifically, below. Before the OS fans rush to point out that there are add ons/modules/plug ins for their preferred OS CMS for each of these, let me acknowledge that yes, they exist and point out that they are not as integrated and/or configurable and feature-rich as enterprise customers expect and need.
Workflow and group management are two sides of the same coin: Enterprise customers need a robust and configurable workflow and configurable with meaningful group management to automate complex business processes across teams through the workflow.
Search: OS CMS search is good enough for their intended use, but not for the sort of searches and indexing across millions of records that is needed at the enterprise-level. TeamSite’s Idol search engine is considered state of the art at the moment at the enterprise level.
Deployment and multichannel support are also two sides of the same coin: Enterprise ecommerce sites often need to publish the same content, maintained as a single source, to different targets simultaneously in different formats based on business rules that are baked into the CMS. The transformations that need to take are not something easily accomplished with Drupal, Joomla, or even Alfresco, nor is publishing to multiple targets in different domains without some crafty tweaking. Whereas this is native functionality in an enterprise CMS like Ektron, SiteCore or TeamSite.
Scaling and performance are critical: My current employer has over 8 million records active in its CMS. Joomla is limited by it’s own code to around 70 thousand articles. Drupal is somewhat better, and only Alfresco can support numbers like ours we found. This brings up another OS limitation: IA. The two OS CMS you mentioned, Drupal and Joomla, were intended to manage content as ‘articles,’ such as a blog. Enterprise content tends to be far more rich and have complex relationships that demand substantial and meaningful information architectures. Something that those OS CMS are fundamentally unable to accommodate without extensive modification, yet again is present and native to enterprise CMS. Yes, you can customize your OS CMS to meet business requirements, but therein lies a huge risk: customizations can take your application off the upgrade path. And once off the upgrade path you’re forced either into very, very costly development and QA efforts in order to port your customizations to the new point releases or just muddling along with the version you have which inevitably will become unsupported without an additional support contract. Either way is costly and ultimately unnecessary in enterprise CMS. I’ve seen this issue happen many, many times, and it never ends well for those who made these decisions.
The trope ‘Free Like A Free Puppy’ sums up another concern that many large enterprises have that OS CMS may have a lower upfront cost, but greater total lifetime costs due to lack of adequate pre-release QA and post-release support. Large ecommerce businesses find it very compelling that proprietary CMS have dedicated support and engineers on staff whose sole role is to help you integrate and write code as well as an active QA team and a genuine, meaningful engineering cycle.
Again, I understand that OS CMS can be extended to cover these cases, but it’s risky and expensive. Large ecommerce businesses often ask why bother when you can buy a CMS that was designed with the same functionality baked in and is fully tested and better supported.
I noticed several interesting trends in the various responses. Most notable (and alarming) was a lack of understanding of the term ‘enterprise.’ Another was the total denial in partisans that the lack of enterprise features in their preferred OS CMS would have anything to do with it not making an enterprise’s short list. Benford’s law of controversy tells us that passion is inversely proportional to the amount of real information available, so perhaps the latter is driven by the former.
Link to the original discussion
One Comment
Very well thought out- thanks for sharing!