Standardizing Coding Practices and Optimizing Performance

Whether a particular firm engages in document coding in-house or through a vendor, standard coding practices should be developed within the firm to ensure accuracy and consistency on every document coding project.
Quality coding can only be achieved when coders are given sufficiently detailed instructions to capture information that will meet the expectations of the attorney or paralegal reviewing the data. For a combination of reasons, the overwhelming majority of coding manuals miss the mark on even the most basic details of coding.
The nature of document coding requires constant analysis of the document population and clarification of specific issues in order to improve the coding protocol and the consistency of the deliverable to the end user.
This article outlines the benefits of utilizing a detailed standard manual and demonstrates the importance for firms to have self-prescribed instructions to meet the demanding expectations of reviewers for accurate and consistent data capture.
Factors used to determine best coding practices.
When creating coding guidelines and practices, we looked at two sources to develop a coding manual that would produce the most accurate and consistent coded data.
First, as a coding vendor we have reviewed hundreds of coding manuals from both scanning vendors and law firms. These coding manuals were developed for both general coding projects as well as in-depth customized projects. Initially, we reviewed each of these manuals in detail and catalogued the most commonly used practices in the industry. From these common practices, we developed specifications with both the industry standards and the search needs of reviewers in mind.
Second, we analyzed the problems that we found with coded data. Through quality assurance review and communicating with coders, we were able to uncover difficulties that coders experienced while interpreting the specifications for coding when they are not sufficiently clear and detailed.
Basically, what we found is that if a coder is not provided with a specification on how to treat a specific occurrence, then they are forced to make a decision on their own. Failing to provide this level of data capture guidance to coders will inevitably lead to like documents being coded in several different ways. Using this information, we created specifications that developed standards.

Acknowledge the nature of the litigation.

The overwhelming majority of coding projects for law firms come up very quickly, as is the nature of litigation and the discovery process. As a result, most coding manuals are not complete in their instruction. Coding manuals are put out quickly, and therefore, address only what the reviewer expects to find in the document population. No instructions are created for the unanticipated documents. By creating a detailed coding manual that is used as a standard practice throughout the firm, an effective solution lies ready and waiting for any coding project no matter how quickly the need arises.
With the basic coding details out of the way, when a project develops, more time can be spent tailoring a project specific coding manual to fit the needs of the document review. With a good standard manual, the groundwork has been laid for a design team to concentrate on altering the manual to more specifically define what the review team expects to obtain from the coded data.
Using a hypothetical example, a standard document type list is created to encompass the basic categories of documents so the coding manual can be used firm-wide. When the coding manual was created, insurance documents were not
given their own document type because they did not often appear in the firm's
typical litigations. Should, however, a particular litigation arises that has a strong focus on a client's insurance documents, the additional burden can be
minimized. Since the standard practices are in place, the design team can spend their time adding details to the coding manual that specify how these documents should be treated, starting with a document type for insurance documents.

Specify your needs and expectations.

It is important to acknowledge that when something isn't specified by the law
firm directly, the coding vendor will either fill in the blanks with their own practices, or the coders will be left to make their own interpretations for the best practices for capturing data. The bottom line: if you have any opinion on how something should be treated, state it in your coding manual.
Defining document categories meaningfully is one example of how to maintain uniformity and accuracy. Document types are the most subjective of all of the standard bibliographic coding fields. A particular document can sometimes logically fit into more than one document type category. If you expect to find a particular document in a defined document type, state that in the definition.
An organization's Annual Report illustrates this point. By title alone it is very likely a majority of coders would apply the overused document type "Report". Annual Reports are often formally printed, which creates some misinterpretation that they fit into a document type such as "Publication" or "Marketing". However, they also encompass a great deal of financial information; therefore, Annual Reports may also be referenced as "Financial" documents. The corporate information encompassing the document leaves open another alternative for applying "Corporate" as the most significant document type indicator. The document coding company will attempt to standardize all of the interpretations made by coders by putting all Annual Reports into one category, but it is still unknown what document type the reviewer believes is most appropriate for an Annual Report. If the details are specified, the guesswork is taken out, and the uniformity of the coded data is maintained. The amount of time it takes to define particular document type categories will ultimately save reviewers from having to search through coded data for all of the possible interpretations made.

A further benefit to standardizing the firm's coding practices is having a firm-wide continuity between how the coded data will appear from one database to another. As one document review project comes to an end, another may be in process and reviewers are confronted with yet another database full of coded data. Just as they have grown accustomed to how one database of coded data was formatted, they are put in front of another database captured in an entirely different format. At the very least, if a firm standardized its format for data capture, consistency will be maintained from one project to the next.
Don't reinvent the wheel.
When attempting to standardize a coding manual, a law firm should seek out the advice of one or more companies that specialize in document coding. Every coding vendor has a standard coding manual, and reviewing their practices provides both a stepping stone to developing your own preferences and a tool to determine how detail oriented a particular vendor will be with your coded data if you should choose to have coding outsourced. If a vendor has clearly put a lot of time and effort into their standard specifications, it illustrates how important accuracy and uniformity is to that company.
Using one or more coding vendors as a resource can also be helpful in determining what can and can't be done in coding. Some law firms write a standard manual for an in-house audience. Practices are included that are of a subjective nature, rather than objective. If this manual is passed onto a coding vendor, expectations may not be met because the details of the manual are focused on data capture that is outside the vendor's means.