Single line of text or managed metadata

Question

I am importing over 200,000 documents into sharepoint. I have a column called Vendor and SubVendor. I did a quick Remove Duplicates in Excel on this columns. Vendor column has 163,000 duplicate value and 37,000 unique values.

Should I create a Managed-Metadata Type for Vendor or Single Line of text. Which will benefit me in long run (say for view, search, query, etc.)?

Same story goes for SubVendor.

Answer 1

Performance: I have a 20,000+ item library using managed metadata that runs smoothly, but can't speak to 200,000+. Looking at how they work on the back-end (hidden columns that are synchronized by the managed metadata service), I could see performance difficulties arising at that size...: http://www.n8d.at/blog/sharepoint-2010/anatomy-of-managed-metadata-fields/. You'll want to run some tests or google around to be sure.

Managed metadata really has a lot of limitations today in terms of user experience: InfoPath, Workspace, Office 2007 and below, Datasheet Mode, etc.: http://www.sharepointanalysthq.com/2011/06/managed-metadata-column-limitations/ & http://www.cleverworkarounds.com/2010/10/28/un-managed-metadata-a-couple-of-gotchas/. It's also a little more work on your part to manage: the Term Store Management Tool is not very efficient at editing lots of terms in a short amount of time (you have to create, rename, merge, etc., one at a time), you have to handle the terms a little more delicately in workflows, etc. You'll want to be comfortable with PowerShell if you want to do anything too fancy.

Where managed metadata really does well is in situations where you truly need to manage a controlled vocabulary (information integrity), you want to be able to encode relationships between terms through the use of hierarchy or term reuse, you want support for synonyms, or you want the fancy new Folder Hierarchy or Key Filters navigation features. Without these features, it could be difficult for example to ensure that if a sub-vendor belongs to a particular vendor, that an item tagged with a particular sub-vendor will always also get tagged with the relevant vendor; i.e. there's no way to tie two columns together, or identify relationships between column terms, in Single Line of Text without some kind of custom code or something.

Some questions to help determine which to choose:

1. do you need to manage the relationship between Vendor and Sub-Vendor? if yes, consider managed metadata (hierarchical terms). text will potentially mean multiple columns that you can't keep in synch except with custom solutions.
2. how will you handle information integrity on values in your vendor column? Will users be able to enter incorrect or alternate names of vendors? Managed Metadata can help keep this clean, but so could a Lookup List or Choice column.
3. how will your users need/expect to work with the data? mass editing of tags? you'll want text. complex OOTB filtering? managed metadata.

Some more good reading: http://blogs.msdn.com/b/ecm/archive/2010/06/22/introducing-enterprise-metadata-management.aspx http://spcert.wordpress.com/2011/02/05/notes-on-managed-metadata-in-sharepoint-2010-%E2%80%93-a-key-ecm-enhancement/