Message Image  

Data Can Be Records, Too

By Kevin Parker posted 02-03-2016 02:03


I often find myself convincing technologists that data can be records, too.

What is a Record?

Put very simply, records are information assets that you keep.

  • Some you want to keep (because you see business value in them).
  • Some you have to keep (because some authority wills it).
  • Some you happen to keep (because you are not actively applying disposition).

Whatever the reasons, records are information assets that you keep. Those information assets can come in many forms and media types. To really manage the records that you want to keep and have to keep, you must consider all of their forms and media types.

Media Types that Can Be Records

These are some examples of media types that may be records.


Paper and other physical records are perhaps the easiest for us to grasp—figuratively and literally. Birth certificates, driver's licenses, diplomas, contracts, purchase receipts, bills of lading, invoices, etc. are all evidences of something and can be considered records. Easy enough. The hard part is knowing which documents are not records, which is beyond the scope of this post (and there are lots of resources on this topic).


Images of paper documents, whether microfilm, aperture cards, or microfiche, can be records. For many decades, microforms have been used to store records, often replacing the original paper documents as the true records. Much of AIIM's history has been in this arena.

Digital Documents

With the ongoing digital transformation of the enterprise, documents are increasingly born digital. While office documents can easily be seen as records, a long-term preservation format like PDF-A should be considered. Digital documents can also have associated metadata that aid in records management.


Organizations are struggling to manage email as records. While many still print and file emails, more businesses are taking steps to manage email electronically. Email headers provide some metadata and should be included.

Multimedia Digital Assets

Information assets in forms like digital images, sounds recordings, and videos can be records. This might include photo evidence, CAD drawings, video footage, voice recordings, movies, and much more. Like email and digital documents, these digital assets include metadata that should be managed with the content as records.

Web Content and Social Media

Web pages, blog posts, status updates, and conversations can be considered records under certain circumstances. Some formats like blog posts are simple enough to be considered records, but start adding comments, ratings, and shares and then it gets more complicated.


What about structured data inside relational database management systems, NoSQL databases, XML files, and other structures? Can these be considered records? The real question is: if information inside databases is valuable or regulated, how can you not manage the data as records?

6 Considerations for Archiving Data as Records

Know Your Preservation Needs

What information do you need to keep and why? Does your data include:

  • Financial transaction records?
  • Auditable evidence of actions taken when adjudicating cases?
  • Sales data for customer relationship management?
  • Scientific research?
  • Other business data?

Is there a subset of this data and a human-readable format that will best support the business need for preservation for future reference?

Don't Keep All Data Forever (Maybe)

Some say "just keep data in databases and keep it forever because you may need it some day." Yet many enterprises are creating and capturing data faster than their technology systems can process it for insights. Depending on what the data is, what it is used for, and several other factors, you very likely do not want to keep it all forever.

Don't Wait for Technology Obsolescence

If you keep data in its database for long-term preservation, you will eventually run into the same problems as other digital information assets: software and systems will become obsolete. Database upgrades can cause changes in relationships and metadata, and record integrity can be compromised. So be mindful to properly preserve what is valuable before it becomes too expensive or too difficult to rescue data stored in obsolete or broken technology systems.

Consider the Best Long-Term Preservation Format

It is tough to point to raw data and see information in context in a form that supports the intended purpose of preserving the records in the first place. Is there a preferable format for long-term preservation? This could be some type of report or set of reports that joins related data together in meaningful ways. This is something a manager or attorney or other human person can point to and see meaning.

Manage Cases with Data and Documents

Case management often includes both data and digital documents. When treating a set of information as a case, preserving the documents as evidence without the supporting data leaves the organization with an incomplete picture of what happened. Once the best long-term preservation format has been determined, data can be exported and formatted, and then managed as a document together with the other documents in a case. This gives a complete picture in a human-readable format.

Govern Data Repositories

Once data is archived for records preservation and retention, should it remain in the database? There are compelling reasons to keep data in database management systems for business intelligence. However, all data that remains in such a system after it is archived elsewhere as records should be treated as convenience or utility copies. This would apply to data that has been extracted, transformed, and loaded into a data mart or data warehouse for BI purposes.

Also, if records resulting from data are dispositioned (destroyed or transferred), keeping it in a database (including backups of the database) means the organization has not really disposed of the records. Be very careful here—and seek competent legal counsel to help your organization.

Long-Term Preservation of Data Records Needs to Mature

This post raises more questions than it answers. What have you done that works well for managing data records? What resources can you share that deal with this topic? Let's figure this out together.


J. Kevin Parker, CIP, ECMm, ERMm, BPMm, SharePointm, Capturep | |




23 days ago

Check out this article in KMWorld Magazine about Data and Records Management. I have a couple of quotes in there:

02-12-2016 11:22

Stefan, you add some good questions and thoughts. I will try to expand on our conversation here in future posts. This is really on my mind these days.

02-12-2016 11:19

Monica, your question is one of my main concerns as well, and the answer is that I don't know yet. Some "data" in database systems and content management systems (which use databases!) are metadata and some are master data, and a subset of this might be considered "reference" data. For the system to continue working and for current and future data and records to make sense, this data has to remain in the system and accessible. There is a lot to think about.

02-09-2016 01:42

Thanks Kevin for sharing your thoughts on this important aspect! I especially like your thoughts on governance on data repositories. I have seen records management projects for many years now and while the data have constantly evolved (trending towards big data and multiplying structured data "swimming" uncontrolled in data lakes), compliance driven records management practices have not evolved with the same speed. Sure, many of the companies I am working for have learned that it might make sense not just to archive a contract or a complaint file, but to enrich it with structured data from their CRM or case management system. But in the end they are still archiving unstructured (PDF) content.
How to enforce structured data governance? I see two ways: Either make sure you NEVER loose context to all the data you store to make sure you can delete it from all systems (archive, business application, DWH etc.) once you have to do so. While this might still work for the business app, it might prove difficult for the DWH (or any other BI systems): Do you also have to delete any reports or insights you gained using the affected data? One approach is to anonymise the data instead of deleting it (process to be confirmed by your legal dpt.). If this works, you could even consider to do so during the data load and you don't have to worry about data retention and destruction in your BI system.
But isn't there a holy grail? Couldn't we just throw any information into a big bucket and use this to store it throughout the whole lifecycle? Linking it to all business apps, BI systems and just apply RM policies? This is exactly what I see appearing on the market and I'm looking forward to see some really convincing success stories.

02-08-2016 10:42

Thank you for this. I was looking for more information around this subject.
Donda's presentation helped a lot too.
I have one question:
How do we ensure that deletion of data doesn't impact the integrity of the remaining data. And if that is the case, how can we mitigate that risk?.

02-07-2016 15:27

Updated LinkedIn with presentation and job aid. Hope it is helpful!

02-07-2016 07:45

Kevin, I love this article and the way it is formatted for easy read and understanding. Dondra, I want your presentation so I am checking you LI site. How blessed am I to know the smart people? Very blessed!

02-03-2016 11:31

Thanks, Donda! I was hoping you'd chime in. I was excited when I heard about your presentation on this topic.

02-03-2016 10:54

Great article Kevin! I worked on a project at my old employers around structured data and was able to find a path to archive some of this 'record' data for identified structured systems. I presented at last year on this topic and would like to share the presentation with you and others to help 'get them started'. There's also an inventory job aid that was very helpful in the information gathering process I can share if folks want to email me. I'll try to add both presentation and job aid to LinkedIn. Feel free to connect directly to further the conversation or pick my brain. =)
Have a great day!