Documents versus Records

During 2002, I was involved in a project to implement electronic records management (not SOX, that act was still in development...) The subject has cropped up again and brought back some memories...

The biggest challenge the project faced was that electronic records had to be managed in an identical way to paper-based files. That experience resulted in point #4 in 'Why is KM so difficult'. I remember, in frustration, muttering 'you can't shred electronic records' when trying to explain why paper files and folders had fundamentally different properties to electronic ones. (Enron was headline news at the time, as was Microsoft's DoJ case)

The second challenge was that the project also wanted to improve document management (including collaboration) and, unsurprisingly, some thought it would be a great idea to roll out a single system to cover both. Now, in theory, it can be done but it needs to be recognised that managing the archiving of a record and managing the creation of a document are very different activities:

Records Management:

• Management of the record is more important than the content of the record
• The record never changes (although its properties might)
• Records require access controls, lots of them

Document Management/Collaboration:

• Without content there is no document
• The document changes a lot, that's the whole point of collaboration
• Access controls restrict and impede collaboration, the fewer there are the better

The ability to have one store manage the entire document lifecycle process depends on what's involved. If you require very granular records management capabilities, the database design will be different to what you need for most document management and collaboration activities. Mashing the two together could lead to performance and scalability problems. It's no different to any other situation where you can have dedicated solutions or combine them together. A sofa-bed is not as comfortable as a sofa or a bed, but if you can't have both, a sofa-bed may be better than only having a sofa or a bed... You have to decide if the trade-off is worth it.

If you insist on forcing an electronic system to replicate a paper-based system, you risk stretching the technology beyond reasonable design limits. Either the system will fail, or people will fail to use the system. As the LATCC learned, a change in tool sometimes requires a change to the process. Failure to do so can lead to escalating costs for a system that will never perform optimally - just because something is possible does not make it plausible.

My advice for companies who need to provide records management and who also want to improve document collaboration:

• Collaboration and records management have different goals and objectives
• All records are documents, not all documents are records
• The ideal records management solution should be an extension to the collaborative environment, not the other way round
• Collaborate to create the document, apply locking controls when the final document is declared a record
• If the required controls are complicated, the records may need a separate database designed specifically to provide those controls
• If you are in charge of designing the required controls, please give some thought to the economic implications of your decisions

And never forget, people can always circumvent the system. Put too many controls in place and they simply will not create the document - no document, no record...