M-Files Best Practice - Metadata

Here at TEAM IM, we've worked with many clients over the years and have helped design hundreds of document management systems with M-Files. Over the years, we've come to find there are key practices when choosing your vault design and metadata which will optimize the structure and use of your system.

Keep it descriptive. The metadata tags you add should correctly and accurately describe the contents of the document or object, who should see it, or how it should be used.

Keep it unique. Remember that if all content items in a set have the same metadata value, then that metadata value does not help distinguish content items within that set. Unique metadata tags will enable optimized search results, workflow automation, and routing. It will also help to prevent a cluttered metadata card by removing non-unique and useless values. 

Keep it specific. While tagging a document with the following sentence "Form 12A, Published 2022" will still work with some search capabilities in M-Files, this is far from optimal in terms of data tagging, organization, or routing. Instead, break this down into its specific components, and if you really need a combination of these values, add a concatenated property to maintain utility.

Example:

• Name or Title: Form 12A 2022 (concatenated automatically)
• Document Type: Form
• Form ID: 12A
• Year Published: 2022

Avoid redundancy. If a value appears in one field, it is not needed in another field. This will reduce the amount of data entry in your system. If you do need a value in more than one place, use intelligence tools, with data automation or concatenation to reduce the manual data entry work. 

Utilize Metadata Card Configurations. It may be tempting to create many new metadata properties in order to gain clarity on the metadata card. However, before you do so, make sure you look into what metadata card configurations are available to you which might help clean your data up and provide clarity without the additional data tags. Configurations like property groups, colors, filters, and labels can greatly improve the user experience and understanding of the data.

Example: Presentation 1 may have a field labeled “comments” while presentation 2 may use the same field labeled “summary” and presentation 3 may use the field but label it “description.”  In all cases, the Comments property is used, but the label and the purpose changes for each presentation.

Modeling Considerations

When designing your vault structure,  creating data models, and choosing your metadata tags, there are questions that you can ask which will help point you in the right direction.

1. Are there any limiting criteria on the options available?  If yes, what are they?
2. Are there any limiting criteria on the fields available?  If yes, what are they?
3. Are there as few value options as possible for this field while maintaining all requisite categorization or labeling schemas?
4. Can the options list be broken down into smaller components or broken out into other metadata properties?
5. Have alternate models that achieve the requisite categorization and labeling schemas been considered?

As always, if you're interested in learning more, you can always read further in our other Insights, or reach out to us at www.teamim.com.