Blogs

Far too long ago I wrote a post introducing the topic of inventorying the 

records and information of the organization. I'd like to return to the topic now with a look at different approaches for gathering that information. I will follow up much more quickly with posts on the questions to ask and some related items. 

 

There are a couple of primary approaches to inventorying information. We can consider these as location-centric, information-centric, and process-centric. 

 

Location-centric. This is the traditional approach for dealing with paper records and has been the staple of records management for decades. In this approach the staff conducting the inventory conduct a walkthrough of the various different areas of the organization. In each area they ascertain what file cabinets, boxes, and other storage locations exist and then conduct a physical inventory of the contents of each location. 

 

The benefit of this approach is that it is very straightforward. It is easy to train staff to conduct such an inventory and many organizations, particularly in the public sector, have worksheets with appropriate fields to fill in as part of the process. 

 

The most significant drawback to this approach is the time required to complete it. It is a very lengthy and tedious process because every box, drawer, cabinet, etc. must be examined in detail, at least the first time around. It also tends to be a rather dirty job - lots of dust, lots of ink, lots of interesting finds in boxes (a thread last year on RECMGMT-L described the unusual things found as part of the inventory, including drugs, weapons, cash, animals in varying states of life/decomposition, baby pictures, music CDs - you get the picture).

 

And this approach is quite challenging for electronic records because they are found in so many different formats and locations. What kind of document is 00000001.tif or work.doc? It's impossible to know without opening them - and that approach doesn't scale in the petabyte era. 

 

A subset of the location-based approach specific to electronic records management is system mapping. In this approach IT develops a list of all of the systems in the organization - process-level, departmental, and enterprise-wide, including applications, file shares, backup and archival media, and so forth. Next, it develops a list of the types of information created or managed by each system. 

 

This can be automated to some extent; moreover, IT should have this information already and if it doesn't it needs to do so for two reasons. First, it is all but impossible to have an effective disaster recovery and business continuity plan without a listing of the systems in the organization. Second, when, not if, the organization is involved in litigation, audits, or public records requests if applicable, it will need just such a listing of its systems and information. This approach therefore serves multiple purposes and should be a priority for organizations that have not yet completed it. 

 

It is important to note that this approach may not be 100% complete because so many individual users, processes, and departments have applications that IT doesn't know about. These could be Access databases, Lotus Notes applications, or custom-developed applications or integrations. So this needs to be followed up with a survey of some sort to the departments asking about their applications. 

 

Information-centric. In this approach users are asked what types of information they create, use, and interact with on a regular basis. The data gathering often takes the form of a survey sent to users, supervisors, and managers.

 

This approach works well with either paper or electronic records because it focuses on the information regardless of location. It can also be more informative than the simple walkthrough because the users answering the survey know the most about the documents they interact with. And it is generally the cheapest approach because the time required of most staff is minimal. 

 

However, this approach also has some issues. The most obvious is that the survey has to be designed to elicit the appropriate information. Survey design is as much art as science and organizations that follow this approach should pilot the survey first to ensure that the questions are clear. 

 

It also tends to focus on information that users interact with regularly at the expense of those items that are exceptions or not as routine. 

 

Process-centric. In this approach department subject matter experts are asked to list their processes - daily, weekly, monthly, yearly, etc. plus ad hoc and irregular/emergency processes. For each process, experienced users of the process are asked to diagram the flow of the process including exceptions and then to identify the documents associated with each process and task. There are several ways to do this, including process mapping, context diagramming, and others. The diagramming process is facilitated by someone who can probe and elicit exceptions and inputs and outputs to the process. 

 

This approach works very well for both physical and electronic records. It improves upon the survey approach through the use of facilitation. It also tends to uncover information that may have slipped through the cracks using the other two approaches through the focus not only on routine processes but exceptions. And it can be less impactful to the organization than the location-centric approach. 

 

However, it does require a skilled facilitator and the time and effort of subject matter experts for each process. This results in a more expensive and disruptive process than the survey. And care still must be taken to ensure that all of the processes and tasks are captured correctly. 

 

It also doesn't work as well for enterprise-wide electronic information systems like file shares, the email system, ERP, and so forth. These should be caught as part of the system mapping as described above. 

 

So which is the right approach? I have had the best success by using a process-centric approach and system mapping exercise followed by a quick walkthrough to confirm that there are no other information stores that were missed. This ensures that I have "seen" all the information - records and non-records, physical and electronic, and departmental- and enterprise-wide stores.

 

In the next post I will describe some of the information to be gathered, particularly with regards to electronic records.