Friday, February 15, 2013

Considerations In Folder Structures

I'm working on developing default folder "core" structures for matter-centric collaboration in iManage (otherwise known as "MCC").  The basic idea with such workspaces is to make it easier to file documents into the right matter, increase compliance with document tagging (for instance, identification of a document as a pleading instead of a letter in the DMS) and, for some firms, file email. In an MCC environment, each client-matter number has a corresponding workspace, and all documents are stored and browsed to within workspaces.  In contrast to the situation in non-MCC environments, document libraries as such as invisible to users, and it is not difficult to find a folder for a specific client-matter.

I see the goals of MCC as increasing the amount of information that is in a matter-number-tagged and controlled environment (as compared to an email silo or shared drive), and also rectifying the tendency of document type tags to be dominated in practice by the "Miscellaneous" category. Document type is one of the key pieces of metadata for finding documents in a document management system, and also potentially is a guide to records management treatment.

This post provides links to some of the key MCC resources, and also outlines key considerations for development of sets of folders, without revealing specific folder recommendations.

Hildebrandt Baker Robbins over at e2.0 (their confusingly-named site that addresses the electronic file 2.0, not enterprise 2.0 business collaboration software) has a number of pages with useful resources.

Probably the most useful is the MCC Design Awards section, which provides actual pictures (for the most part) of award winners from 2007 to 2010.  Another suggests "How to Start With MCC", which notes that MCC is tied to matter opening, how attorneys and staff work with documents, records management, and (of course), knowledge management.  Another extensive MCC resource is the International Legal Technology Association's somewhat dated "matter centricity" page.

MCC Principles

Simplicity

As with other taxonomy projects, I believe that MCC adoption and success benefits greatly from simplicity (see "KM and the Simple Stick").  In practice, I think this means that no more than 5-10 substantive folders, and starting out with as little as two, one for email and one for documents.  The MCC Design Awards and other industry trends indicate that approaches calling for dozens of folders for various matter types have failed to be fully implemented.


Simplicity primarily relates to the number and granularity of folder types. While lawyers may like to have their electronic documents organized in dozens of different categories and sub-categories (befitting their categorical brains), they and their staff are rarely able in practice to leverage such an extensive list.  Unless a top-down, consistent approach is consistently enforced, a small number of folders is better.


Ease of Use

One key mistake that I believe was made in early MCC tries was to have extensive use of multiple levels of folders.  This is problematic in that the primary mode of filing in MCC is through drag-and-drop, which is much harder to do with multiple folder levels.

Another important aspect of ease of use is quick access to recently accessed workspaces. The vast majority of the time, the workspace / folder you need to save a given document into is one you've touched recently.  Workspaces and folders within them therefore need to be readily distinguishable, and made available in ways other than through browsing and searching all of the workspaces.

A third aspect of ease of use is allowing users to customize.  You can't let them change the most visible aspects of workspace and folder structure, but one should allow for appropriate degrees of customization.  For instance, users should probably be able to add as many folders at the "top level" of their personal workspaces as they like, but not be able to so easily modify the "top level" of a client-matter workspace.

Utility

A good MCC design will acknowledge the primary scenarios in which people use folders.  This means you have to account for people looking for documents and information as well as people's work in filing documents and email.

People looking for information have different needs than people filing documents.  For one thing, they will want quick access to search.  Many award-winning designs have provided folders that generate matter-specific searches, of documents, email, and documents + email.  Also, they may not be familiar with the substance of the matter, suggesting that some consistency in the folders used can lead to greater utility for information searchers.

It is essentially impossible for a central team to accurately predict the precise information management needs of widely distributed matter teams working on different types of cases.  Accordingly, the design should allow matter teams to create and control folders, at least, below the "root" level.

A Place For Everything

A user should be able to quickly (within a few seconds) identify the correct folder for 80-90% of the common documents generated in the course of legal work.

Miscellaneous

Most of the winning designs do not have a miscellaneous category.

Types of Workspaces

A good MCC design accounts for, first, client-matter workspaces; second, personal workspaces; third, legal administrative groups like practice areas, departments, and firm committees; and fourth, non-legal administrative workspaces (for groups like marketing and HR).

You can develop lists of needed workspaces through examining and analyzing two primary sources, the current set of documents and the client-matter or nonbillable client-matter number equivalents used to save them, and the current organizational structure of your firm.

Closing Thoughts

As with tax and constitutional law, the principles of MCC have the potential to conflict. Balancing the predominant principles with your firm's culture and way of working will be the keys to a successful MCC rollout.

No comments: