Gartner uses false equivalency to warn against building custom content services

I guess it shouldn’t be too surprising that Gartner, a firm that gets paid by the software vendors it reviews, would be against the idea of custom content services. If companies stopped paying ridiculously high license fees and maintenance for bloated ECM platforms it could cut into Gartner’s bottom-line.

What is surprising is that they would use such a blatant false equivalency in describing why they think it is a bad idea to build your own content services platform. A recent blog post by Marko Sillanpaa starts out with this little gem:

You would not build a database from scratch…why would you ever accept the idea of building a content services platform from scratch?

Come on, Marko, you can do better.

Yes, at a very high level, one might describe a content management repository as a “database for documents”. But when the conversation shifts from needing to convey the basic concept of what a content repository is to why we might choose to custom build some or all of a content services platform, the comparison breaks down.

A relational database is a fundamental, commoditized element of a technical stack. For most uses, when someone needs to store rows and columns of data and then query for that data, any database will do. There may be differences in how databases perform, the platforms they support, or specific features they implement (clustering, replication), but, for the most part, they are the same.

Further, a relational database is very precise in the functionality it delivers. Only in the most niche applications would you hear someone say, “That database does too much for us–it delivers way more functionality than we need.”

So, for the vast majority of cases, it would never make sense to build a custom database for our larger application because so many are available that do exactly what we need and no more.

A content services platform, however, couldn’t be more different in these respects.

First, agreeing on what “content services platform” means is a challenge. Marko says, “the repository is more than file storage. It’s about version control, editor integrations, workflow, records management, etc.”. Really? I have several clients who require exactly none of those features but still need a content services platform.

Next is the issue of granularity. When you think about building a custom content-centric application, no one looks at FileNet, Documentum, or Alfresco and says, “Yeah, we’re just going to drop those in to our solution”. That’s because those platforms want to be the solution. Unlike a database, which is happy to focus on rows, columns, indices, and queries, legacy ECM platforms tried to do everything. My friends over at TSG do a good job of summarizing some of this in their blog post (which is how I came across the Gartner post in the first place). I love this quote:

modern content services needs to be rebuilt from the ground up focused on a new approach rather than rely on paradigms from the 1990’s that haven’t worked

I couldn’t agree more. And if the legacy ECM vendors aren’t going to modernize and rightsize their platforms, then service providers will continue to do so, and we’re going to continue to see content-as-a-service vendors like CloudCMS and Contentful continue to do well.

Now, like Gartner, I obviously have my own bias–it is more interesting to me to build custom content-centric solutions for clients than it is to try to turn aircraft carriers (ECM platforms) into speedboats (nimble, fit-to-purpose customized solutions). But maybe in the age-old buy-versus-build discussion we can avoid fallacious reasoning when advising customers which path is best for them?

If not, and Gartner wants some ideas, here’s a list of some good false equivalencies I came up with that they are welcome to use:

  • We would never build a custom web server, so why would we ever build a custom web application?
  • We would never build a custom enterprise service bus, so why would we ever build a custom order fulfillment system?
  • We would never build a custom desk, so why would we ever build a custom office building?
  • We would never build a custom keyboard, so why would we ever build a custom gaming rig?
  • We would never build a custom water heater, so why would we ever build a custom house?