Travis County Records Manager
Not an official web site of Travis County, Texas government
Email Retention
Page 2
Page 3
Page 4
Consulting
AIIM Training
ERM glossary
The Dublin Core
BPM
Video
Contact Shawn Malone
Other Shawn Malones
Site Map
Blood Simple
Bicycle leashes comparison
The Traditional Records Management Approach
 
Classifying, categorizing, and indexing records has been the most successful way to store, find and use them for most of human history. The best records management practices were developed over decades to comply with federal and state laws and regulations. During most of that time, almost all government records were paper documents that could be filed in hierarchical structures. Records managers adjusted to microfilm technology completely. But legal and regulatory management of electronic records is maturing too slowly to keep pace with the growth of unstructured electronic records like email.

The traditional approach arranges records so that they can be found and so that bulky physical records that are taking up valuable space can be destroyed. But if finding electronic records is cheaper and faster using keyword searches than by categorizing and filing, and if storing them for long periods doesn't take up space, then those records don't need to be arranged.

 



In addition to federal and state case law, legal opinions by various attorneys general, and other authorities, there are the 12 records retention schedules promulgated by the Texas State Library and Archives Commission (TSLAC). These are designed to classify, categorize and retain the evidence of the business of local governments in Texas. About half of these schedules have some application to Travis County government, and the most important of these is Schedule GR. It includes 135 records series, some of which are divided into sub-series, and has special notes as well. Some of these typical notes call for a type of record to be retained for a given number of years after a trigger event, like the completion of a project or an audit. Since some trigger events cannot be known when an email is sent or received, some electronic mail records cannot be categorized at the beginning of their life cycles. This is the After-Audit Axiom!
     
If there are 4,500 County employees using email, and each one has an average of 25,000 emails stored somewhere, then there are 112 million records for an analyst to examine, classify, and categorize so that the retention assigned by the TSLAC schedule can be applied to each email. If each email takes three minutes to analyze, that's 2,704 FTE years to apply the traditional records management retention analysis to the existing email inventory. If this calculation is off by a factor of ten -- if it's ten times easier than we think it is -- that's still 270 FTE years.

The majority of emails could probably be deleted or would fall into one or two series on the GR schedule issued by TSLAC. A small percentage of emails (and attachments) pose a classifications challenge to traditional records management methods that could require a thorough and thoughtful analysis to properly assign retention periods.

If 270 FTE years is too great a resource allocation to apply to email records analysis, what if we tried to train all County email users to analyze each email and apply retention? How much training would be required? Austin Community College included one class in a semester as part of its two-year Records Management associate degree program. The Texas State Library and Archives Commission offers a half-day class on the fundamentals of retention and classification. The Association of Records Managers and Administrators considers Records Information Management practices one of its six "core competencies." Training programs like these traditionally have focused on building the expertise of an individual to apply a retention regime to structured series of records that manifest repetitive elements.

Some see the exponential growth in the number of unstructured electronic records and opine that individual employees need to be given more duties to classify and categorize the records they manage themselves. Any expectation that a significant training requirement along these lines could be mandated at Travis County is unrealistic. If 4,000 employees are trained for half a day on records management fundamentals and how they apply to electronic mail records, they may not have a chance to apply that training to destruction of a given email for years.

Any technology that makes it harder for GroupWise users to do their work will push them to use web-based email alternatives that are easier to use, increasing the likelihood that some County business communications will move outside the appropriate channels.

If a third party vendor were hired to build a product that sat on top of GroupWise and helped users categorize their emails as they were sent and received, experience suggests that the resulting technology would be less stable and harder to use than the current email environment.

A policy that called on users to evaluate each email, assign an appropriate retention period, and save it for as long as they're supposed to would be difficult to understand, teach, and enforce. The relatively small customer base for Novell GroupWise means that some of the most powerful and popular products that could apply technology to helping users retain email are not available to Travis County. Any technology tool that was developed to assist users in this task would have to be complex enough to automate retention and destruction analysis but simple enough for all users to adopt easily; this balance would be difficult to strike. A policy that centralized application of a retention regime without user involvement would be, for practical purposes, technologically impossible.