SAA/CoSA/NAGARA 2018 recap: Session 605

Session 605 – Taming the Web: Perspectives on the Transparent Management and Appraisal of Web Archives [RIM]

Session 605 offered different organizational perspectives on the management and appraisal of web archives. The perspectives included a municipality, a university, and state and Federal government.

Local Government Perspective – Austin, TX

First up, Katherine Cranford described the types of records found on their websites – many of them permanent. She explained how stakeholders approach web archiving from different perspectives. They manage web content by connecting their document management system, OpenText eDocs, to their websites via API. This ensures documents are protected and maintained according to records schedules. They use ArchiveSocial according to their social media policy. To ensure only necessary information is in their content management system, Drupal, they use policies. She recommends using a style guide if policies don’t work. She emphasized the ongoing importance of content audit and governance.

University Perspective – Johns Hopkins University

Next, Jordon Steele explained how they use the Archive-It service to capture websites and Facebook. The web archiving labor includes:

  1. Deciding on seeds (working with IT and student center to get a list of all officially registered groups)
  2. Performing test crawls
  3. Troubleshooting issues
  4. Saving crawls
  5. Quality assurance
  6. Metadata creation (embedded in Archives Space)
  7. Preserving archival records
  8. Performing reappraisal on a regular basis
  9. Repeat (annual or semi-annually)

Jordon discussed the ethical considerations of documenting student groups. They managing the tension between their ethical obligation to document campus life and the ethical obligation to ask permission. If they decide not to ask, can they mitigate using redaction or access restrictions? Could they apply standard restrictions to the web archiving platform? They are trying to determine what they should do based on their priorities.

Jordon mentioned the following key resources in developing their program: Collecting Policy for Duke University Archives, Middlebury College Web Archives, University of Virginia Data Documentation & Metadata, and Documenting the Now.

State Perspective – State Library of North Carolina

Next, Krista Sorenson explained how the State Library works with the State Archive to manage state publications, documents, and public records. They began using Archive-It in 2005 and ArchivesSocial in 2012. They perform bi-monthly capture of state agency websites and content, including publications only available on web.

After 13 years, they reevaluated their approach. They are focusing on user experience as they know patrons may find it difficult to find what they need. They performed an audit and are reconsidering their approach to metadata and documentation. They’ve determined they have to periodically review their approach and create clear documentation to make a well-managed, transparent web presence.

State Perspective – State Archive of North Carolina

Jaime Patrick-Burns discussed hot they capture websites, blogs, and social media of official state organizations using Archive-It and ArchiveSocial. For quality control in ArchiveSocial they monitor accounts and for Archive-It, they download crawls, look at data, and check seeds to see how they appearing. Then they add rules and do test crawls of their 700 active seeds. They take top 5% and bottom 5%, review all errors, check how they appear in the Wayback Machine, and record actions taken. With this approach, they are looking at the seeds most likely to cause problems. They are rolling out a new approach to divide seed list and check a section at a time so all seeds get checked annually. Their ongoing issues include the maturation of web archives, scalability, communicating with stakeholders, and limits on the number of accounts in Archivesocial.  

Federal Perspective – National Archives and Records Administration (NARA)

Kyle Douglas gave an overview of the NARA guidance on managing web records. While the NARA Guidance on Managing Web Records is from 2005, it is still applicable. NARA is working on new guidance and considering various options, including pursuing Capstone-like approach to manage top-level web records.

NARA asked agencies about how they are managing website records in the 2017 Records Management Self-Assessment (RMSA). In response, 55% of agencies said they are managing their websites as records and 45% said they were automatically capturing web records. 28% said they were transferring to NARA.

NARA is in the process of developing Use Cases for Website Records as part of FERMI. The use cases can be used by agencies to evaluate vendors’ ability to manage web records. Kyle also pointed to Documenting Your Public Service as a resource.


SAA/NAGARA/COSA 2018 Recap: Session 305

Making Policy Work: Creating and Implementing Information Guidance in the Age of Open Government

Christopher Magee from NARA kicked off the session by discussing the importance of policies: they can help you improve consistency across your organization. When creating policies, be sure to include as many users as possible and be transparent about your work and overall purpose. Be sure to consider immediate as well as long-term business needs.

Gail Snow from King County Records and Archives in Seattle, Washington spoke on promoting transparency during the policy creation and implementation process. King County is a large and complex local government where it is extremely difficult to get any policy adopted across the entire county. Some of the challenges include having policies which technically only apply to the executive branch of government as well as providing policy guidance to employees who are not stationed at a traditional desk. One solution that worked for them involved prioritizing the overall records management policy, an on-boarding policy, and an exiting/transferring employee policy, as these were common functions throughout the entire county. They also leveraged services provided by the RM program to get buy-in from other government branches and elected officials. In addition, they used the county’s Public Records Committee to sponsor policies. To increase transparency and accountability, they have pushed all RM policies to the web.

Continue reading “SAA/NAGARA/COSA 2018 Recap: Session 305”

SAA/NAGARA/COSA 2018 Recap: Session 201

Email Archiving Comes of Age

This session was composed of lightning talks about various email archiving projects, including the first NHPRC electronic records case studies focused on email archiving.

Chris Prom from the University of Illinois at Urbana-Champaign reported on the CLIR Report on Technical Approach on Email Archiving (CLIR 175). The report is available here. The purpose of the report is to document how archivists are currently preserving email as well as to frame email preservation in terms of what technology is available and how it can be used. The report includes topics such as why email matters, technical definitions, lifecycle models, tool workflows, as well as an agenda for email archiving moving forward.

Continue reading “SAA/NAGARA/COSA 2018 Recap: Session 201”

SAA/CoSA/NAGARA 2018 recap: Session 706

Guest post by Scott Kirycki, Digital Archivist, University of Notre Dame Archives

Session 706 – Opportunities in Change: Transition to Functional Records Scheduling in Washington, Wyoming, and North Carolina

Session chair Courtney Bailey (Records Analyst, State Archives of North Carolina) introduced the panel’s topic with the SAA Glossary’s definition of functional analysis: “A technique that sets priorities for appraising and processing materials of an office based on the relative importance of the functions the office performs in an organization.” The presenters then described how they applied functional analysis in creating functional records schedules for their states. They spoke in the order that their states implemented functional records scheduling (Washington, Wyoming, North Carolina), and Bailey mentioned in particular that she learned from what her colleagues had done, thereby illustrating that it is not always necessary to “reinvent the wheel” in records management.

In Washington, Russell Wood (State Records Manager, Washington State Archives) faced 28,000+ record series along with issues such as duplication of schedule numbers that were supposed to be unique. The schedules’ organization was based on which office created a record, and this led to the need for monthly changes to the schedules as office structures evolved. Switching to an approach based on the function of the records enabled Wood to develop a general schedule for records common to all state agencies and eliminate duplication. Making the schedules smaller helped get record creators on board with using them.

Mike Strom (State Archivist, Wyoming State Archives) told of a similar process of simplification by moving from office-specific to more general schedules. So far, the original 8,000 schedules have been reduced to 600. The process has involved some agitation, however, among agencies that were used to having their own sets of schedules. Strom emphasized communication and education as the keys to addressing the agitation. He gave examples of a training video that explains how to use the schedules and crosswalks that show how old and new schedules relate. He also suggested seeking input on proposed changes from the people who use the schedules.

Bailey spoke as well about the need to involve stakeholders in the making of schedules.

Over 200 individuals participated in meetings to evaluate North Carolina’s schedules as they were developed, and Bailey posted the schedules on a blog so those who were not at the meetings could provide feedback. To determine agency functions, she looked at their websites and previous schedules and asked records managers within the agencies to identify functions.

Later, when the schedules were ready, records analysts did training with the units. For additional training, Bailey wrote choose-your-own-adventure-style tutorials.

To show that the schedules are not arbitrary, Bailey included legal citations even though it was challenging to figure them out. She made an appendix listing the titles of series that come to the archives. This appendix helps record producers and managers, and also offers the public transparency about what kinds of records archives consider archival.

The session concluded with a question from the audience about whether members of the public might find general functional schedules harder to use when making records requests than agency-specific schedules. Wood recommended writing either type of schedule in a way that the public can understand. Bailey answered that uniformity in retention has cut down on confusion for the public. Strom indicated that Wyoming is not using the new schedules directly with the public in records requests, but he suggested that the new schedules may help cut down on the time it takes staff to find records to respond to requests.

SAA/CoSA/NAGARA 2018 recap: Session 204

Scheduling the Ephemeral: Creating and Implementing Records Management Policy for Social Media


Bethany Cron, “Creating Records Management Policy and Guidance for Federal Government Agencies”

Kristen Albrittain, “Implementing Agency-wide Social Media Records Management Across 100+ Enterprise Accounts”

Laura Larrimore, “Communications and Content Creator Perspective on Social Media Records Management”

Link to the PowerPoint presentation with presenters’ notes here.

Key Takeaways

  • Do not delete content from live sites.
  • If you do have to delete content, have a documented process in place.
  • Have (and share publicly) a Comments Policy.


Beth, Kristen, and Laura presented a program in which they each brought their unique perspective to the management of social media as records within the Federal Government. By doing so, they effectively demonstrated how vital cross-team collaboration really is, and how clear policy informs good practice.

I’d like to begin the session recap by focusing on Laura’s presentation, as hers is a perspective many archivists and records managers may not have heard before.

Communications and Content Creator Perspective on Social Media Records Management

Laura is a social media person who creates content. She shared insight into her daily workflow, which is running agency social media accounts such as @USPTO and @CommerceGov, as well as the accounts of individual leaders, such as the Secretary of Commerce. Laura works closely with leadership in communicating their priorities as heads of the agency and the agency’s priorities. She also works with leaders to assess how the public received the content sent out, and what the leaders might want to say next.

Troubleshoot #1

A week before the 2016 election, Laura was brought in on a 6-month detail to the Secretary of Commerce’s office to help with the presidential transition, focusing on website and social media. Like many other cabinet leaders, the secretary had a named account in addition to the agency accounts. One of her main tasks was to retire the outgoing Secretary of Commerce’s presences, and stand up the presence for the incoming Secretary of Commerce.

There was no precedent for this. This was the first presidential transition in the social media age. So, she collaborated with other staff to do what they could with what they had.

The transition team looked at how the Department of Commerce (DOC) had handled changes to the website, and adapted that model to fit their needs. DOC had a history of taking a screenshot of the site right before it changes in a major way and then making it available as an online searchable archive. They do this with various redesigns and when a new Secretary is confirmed. This allowed DOC to make changes to the website, but direct the public to old content in as close to its original state as possible.

Troubleshoot #2

The Secretary of Commerce’s official government account used her real name as her Twitter handle when it was set up.

pennyYou should always ID the leader’s position in the handle, so that it is clear the account is a function of their position. By doing this, the leader can keep their name for personal use, and so that it is intuitive that when the person’s time in that role ends, so does the use of their official government account. Laura likened handles to a company car – it helps you do company business, but once you leave the company, you do not get to keep using the car.

Since the team had this naming issue, they had to work with Twitter to move the followers and content to a new account. In doing so, they would be able to release the @PennyPritzker handle back to former Secretary Penny Pritzker.

The team indicated on the @SecPritzker that is was an archived account and inactive.

Troubleshoot #3

Snapshots are good, but they won’t capture deleted tweets. Another day-to-day aspect of records management is having a protocol for if, when and how you might delete tweets. It makes sense to determine if your agency would ever delete anything, for what reasons, and develop a protocol for this prior to such time as you actually need it. For example:

  1. Take a screenshot of the error tweet
  2. Save it in the ‘deleted tweets’ folder
  3. Write up why it was deleted and how to improve/avoid the issue in future
  4. Email screenshot and description to leadership

Lessons Learned

  1. The social media people and records people need to communicate and have a plan BEFORE a big change occurs.
  2. The idea of dealing with records can be more intimidating than actually doing it.

Creating Records Management Policy and Guidance for Federal Government Agencies

During her presentation, Beth talked about NARA’s high-level requirements and best practices for capturing records created when Federal agencies use social media. This guidance utilizes principles that can be adopted by a variety of institutions, not just Federal agencies.

NARA’s policies are social media include:


Beth identified not a few substantial challenges in managing social media as records.

Identifying records

While one can consult the Federal definition of records as defined in the Federal Records Act (44 U.S.C. 3301), some questions to ask yourself include:

  • Does it contain evidence of an agency’s policies, business, or mission?
  • Is the information only available on the social media site?
  • Does the agency use the tool to convey official agency information?
  • Is there a business need for the information?

If you have answered yes to any of these questions, then there is a chance that these records meet the definition of Federal records. However, are comments part of the official record? Should they be?

Appraising records

Social media posts are ephemeral in nature may not hold ephemeral value. Due to how quickly the social media environment shifts, the process of appraising, capturing, and bringing social media under intellectual control is incredibly challenging.

Locating records

Where is the social media post of record located? Who owns that post? While a copy can be found on the platform it was distributed through, but perhaps there is a screenshot of the post saved on an office shared drive somewhere. Perhaps an automated tool is capturing the posts. Since there are so many copies….

Scheduling records

…scheduling social media through either a general records schedule or a programmatic schedule is needed.

Negotiating public expectations

Since social media is considered a public space, there is an expectation that all posts will remain public and available, preferably in its native format, including any content that has been deleted or altered in some way.


The Federal Electronic Records Modernization Initiative complements NARA Bulletin 14-2.

Implementing Agency-wide Social Media Records Management Across 100+ Enterprise Accounts

Currently, social media is unscheduled, so NARA has been treating everything as if it’s permanent. Staff spend a great deal of time evaluating social media platforms to determine whether posts are original, substantive content or are mainly being used to point followers to more substantial content. Staff also evaluate how content users are using the platform – appraisal is not only about what is being communicated, but also a question of who and why.

NARA’s Corporate Records Management team ultimately decided to take a Capstone approach, designating records as either permanent or temporary based on the content owner. They focus their energy on original content created by senior executives (or their representatives) in the course of their work.

nota bene Content created by all other offices is temporary. Kristen emphatically stressed the following: disposition dictates that COPIES of the content should be deleted/destroyed after three (3) years. Copies, not the original social media post, which in this case is considered the non-recordkeeping copy. DO NOT DELETE YOUR LIVE SOCIAL MEDIA POSTS.

Market Research Lessons Learned

In September 2017, a one-year subscription to the social media records management tool, PageFreezer, was purchased. PageFreezer maintains a record of all NARA-created content on Facebook, Twitter, Instagram, YouTube, Tumblr, and WordPress blog channels, as well as user-generated content posted to NARA-owned pages. At the moment it’s capturing 122 separate channels across those six platforms. The system works by connecting with the platform APIs and capturing data as often as each tool allows, which means it’s as close to real-time as possible.

Authenticity, Integrity, and Completeness

One of the key benefits that an automatic capture tool like PageFreezer has over previous manual approaches is that it can ensure the authenticity, integrity, and completeness of the records.


PageFreezer preserves original content, including responses, with its original look and feel.

Benefits for FOIA and E-Discovery

PageFreezer’s digital signatures, history audits, and complete metadata satisfy Open Records requirements such as FOIA, ensuring that records meet legal requirements for e-discovery.

SAA/CoSA/NAGARA 2018 recap: Session 103

Guest post by Cathrine Giles, State Records Branch Manager, Kentucky Department for Libraries and Archives

Wicked problems are defined as societal problems that are complex, vitally important, ill-defined, and rely on contested political judgements for resolution. Even just defining the scope and nature of a problem can be difficult. We can use “wicked problems” to frame a number of challenges the archival profession faces in order to identify where our professional knowledge can be applied to wicked problems challenging society. This session, Archival Engagements and Recordkeeping Intersections with Wicked Problems, led by Eliot Wilczek (MITRE Corporation), examined three examples of wicked problems.

Snowden Becker (University of California, Los Angeles) presented first with “To Protect and (Pre)Serve: Making, Keeping, and Considering Police Records.” The wicked problem here is how to apply records management practices to body worn cameras and the footage they create, particularly since body worn cameras and their footage are uniquely invasive, voluminous, and easily manipulated.

Becker looked at property rooms as a form of archives and evidence management as a variety of archival practices. Identifying the common ground between evidence and archival management may allow us to work on solutions that are mutually beneficial. Archives and property managers have difficulty with integration of audio/visual into recordkeeping, and the divide between evidence that sits on a shelf and evidence that sits on a server is growing. Archives and property managers need better training, more resources, and better tools for handling collections related to trauma and the continued trauma the collections may bring to a community.

Becker ended by saying it’s important to avoid simplistic readings and speculations about consequences, and to ask important questions and engage with institutions we may find problematic.

Anne Gilliland (University of California, Los Angeles) presented next with “Refugees, Records, and ICT at the Borders of the State.” She reviewed several challenges that face the recordkeeping fields, such as incompatibilities and inadequacies in databases and problems with digitally integrating and accessing archival recordkeeping systems and holdings, and the issues that come out of these: data access vs data protection, tensions between the need for digital fixity and digital redundancy, to name a few. With these issues already facing the recordkeeping fields, the issues refugees face are even more compounded, especially since future lives and generations depend on what happens in the present.

Gilliland discussed how the Refugees Rights and Records Project (R3) is working to address the recordkeeping challenges faced by refugees. R3 wants to identify how the use of records play crucial roles in the lives of refugees; how the professionals involved in recordkeeping in affected countries might work in ICT implementation to identify, protect, and secure such records; and potential policy recommendations that support refugee rights into records. They strategize with parallel projects that have similar needs, concerns, and conclusions.

Eira Tansey (University of Cincinnati) was the final speaker and presented on climate change-related records with “Appraising the Archival Anthropocene.” Tansey suggested that assessing the lengths between recordkeeping and climate change helps us appreciate why climate change became a wicked problem. In appraising the records of climate change like we do an institution, you start with examining who has created what kinds of records in order to carry out which specific functions.

Tansey proposed three primary functions that climate change records perform: 1. Records establishing scientific consensus for climate change. 2. Records created to manipulate public policy by fostering doubt about climate change. 3. Records that provide accountability and suggest a path forward for justice. In the case study focusing on ExxonMobil, Tansey showed how these three functions applied. Recordkeeping has informed every stage of climate change as a wicked problem. However, because ExxonMobil is a private entity, the public does not have full access to their records; as a result, the appraisal of their records is limited to what has been released by them or obtained by a subpoena.

Tansey concluded by reminding the audience that records are never neutral. Records reflect the values of those that create them.

SAA/CoSA/NAGARA 2018 recap: Session 601

“When everything you do in your repository is a high priority, how do you prioritize the core archival practice of processing?”

This was the premise of a panel titled “Prioritizing Processing When Everything is a Priority” anchored by several archivists. The first speaker was Pam Hackbart-Dean (Southern Illinois University) who analyzed the results of a survey concerning how processing decisions are made, who makes them, what factors influence processing decisions.

Significant barriers to processing include the usual suspects: staffing, unexpected acquisitions, changing department plans, changing administrative priorities, and lack of expertise in unusual formats. Processing can often get backburnered because it is less urgent than the other fires that come up in an archivists’ daily working life. Of particular interest to records manager was the survey finding that retention schedules rarely factor into how processing decisions are made.

Jamie Martin (IBM Corporate Archives) spoke next from the corporate archivist perspective. She noted that business archivists are unique from other archivists in that they provide answers, not records. The questions they receive from in-house are often quite complex, and the answers are needed within an hour. In this kind of setting, a more comprehensively processed collection is easier for the researcher (i.e., the corporate archivist) to navigate and deliver a very quick answer. Jamie talked about how she develops 1-3 year processing plans for the records at her repository.

Jessica Geiser (University of California Riverside) talked about how there was no centralized processing prioritization plan when she arrived. Since students were the main processing personnel, she assessed the backlog by applying a scoring metric to each collection, on a numeric scale by level of complexity/difficulty. This resulted in a large spreadsheet that designated which collections could be reasonably processed by student workers based on their score and which would be delayed because of size or special needs. It also set expectations for when collections would be processed.

Joanne Archer (University of Maryland) discussed how processing priorities at her institution were determined by a number of data sources documenting patron use and interest. The user data her institution used to develop processing priorities were subject specialists’ and curators’ assessment of research value, reference statistics (RefAnalytics), circulation data (Aeon), web analytics (Google Analytics), and speaking with users and public services staff.

After the panelists spoke, the session transitioned into breakout sessions for the remainder of the panel time slot. We broke into four different groups:

* Understanding our users
* The value of tools
* Project management and administration
* Managing unprocessed materials and backlogs

The large number of attendees at this session was testament to how many people find this to be a continuing issue that vexes large numbers of archivists. However, I think this session would have been improved by two things:

The workshop was dominated by academic archivists, and as the corporate archivist pointed out, different settings have radically different demands on archivists and what archivists are expected to produce (records? answers? both?). As a result, I wish there had been more institutional diversity. For example, I imagine that being a local records repository would be exceptionally challenging for processing priortization.

Finally, although this was a session identified as part of the annual conference “RIM track” it was hard to detect any emphasis on records and information management in this session. With the exception of the survey result showing that retention schedules rarely impact processing prioritization, there was no discussion of how retention scheduling or appraisal intersects with processing.