Document Capture is the on-ramp to ECM. Without documents and the ability to import them into a system, ECM is meaningless. Document Capture can be done in a variety of ways; scanning, faxing, emailing, API integration, etc. However you stuff your documents into your ECM system, their is one major obstacle to making those documents useful, metadata. All ECM system’s use metadata, even file systems. A good metadata model is perhaps the most important component to ensure the success of an ECM implementation.
Some metadata is obvious, like author/creator, date created, and the primary identifier (think primary key for those of you RDBMS folks). All of the systems that I have worked with have these metadata fields, whether they expose them to the user or not. This leaves the ECM architect to determine what other metadata fields to associate with the documents the ECM system manages. I have come up with four guidelines, when it comes to creating metadata fields.
- What fields would be useful to help users find content. This guideline seems obvious. But how you go about answering this question may not be. The first place to start, is your end users. Ask them how they currently find documents. When they look at documents, what information in the document is most important to them. It is also helpful to look at the documents themselves to find helpful metadata fields. If you have an AP document, chances are it has a date specifying when a bill was paid. This may also be good candidate for a metadata field.
- What information needs to be stored with the document, but is not a part of the document itself. Often times a user may want to make a comment about a document. Or a document has a revision number that needs to be captured. This information is important, but will most likely not be in the document itself.
- What are your metadata dependencies. While you are architecting your metadata model, look for metadata dependencies. For instance, say you have the metadata fields Business Unit and Document Type. The Accounting Business Unit will have documents of Document Type Invoice. You don’t want users in the Business Unit Professional Services having access to search Invoices. If your ECM system supports this functionality, you will want to group specific Documents Types with Business Units.
- Keep the number of required, user entered metadata fields to 5 or less. If your users are required to enter values for more than 5 metadata fields, you run the risk of data corruption. Entring metadata into a checkin form is not fun, and users can get bored of it fast. Use automated methods to catch as many fields as you can.
If you are considering creating a metadata schema for an ECM system, this list should give you a good head-start. You should also consider how the particular ECM system you are using handles metadata.