A Better Document Review Begins With Better Data

Posted on

Leveraging Pre-Review Analytics to Pinpoint Documents is a Good Start

by Michael Wade, CTO – Planet Data

In today’s environment the demand to get more done, in less time and for less money while striving to achieve higher quality is one that we all frequently hear – and very loudly! 

One effective tool to do just that is email threading analytics.  This often overlooked tool does not only provide significant improvements in efficiency, but it also provides a significant increase in quality.   While this sounds too good to be true, there is a growing body of evidence that this along with some simple changes in your review process will allow you to achieve these goals.   At a recent conference a presenter made a really astute observation: “when you are trying to read 10 books, you would not read a few pages out of each book until you read them all.  You would just read each book in order“.  While in this context this sounds patently obvious, it is exactly what you do when you batch out documents to be coded by custodian. 

Can you imagine trying to fully understand the individual stories of those 10 books if you attempted to do it that way? Why would we expect our reviewers to understand complex chains of communications when they only see parts of the entire chain?  Organizing your data by conversation/topic allows us to present the entire story to your reviewer.  Furthermore, by having the reviewer read only the “inclusive” or the most information rich emails, we can reduce the total number of records that are required to be reviewed.

This process improves your results in several ways:

  • Supporting these technologies in Pre-Review allows you to understand the scope your case more quickly by allowing you to focus in on entire conversations rather than just snippets.  This can be used to improve search term selection, prioritize custodians and concept cluster choices earlier in the process.
  • Improved consistency in coding decisions.  Since reviewers are seeing the entire context, small snippets of a conversation may have a clearer meaning and result in a more consistent and accurate coding calls.
  • Provides a great QA/QC tool to look for inconsistent coding calls by reviewing all the documents in the thread.  In particular reviewing coding decisions on privilege are easier to review and validate.
  • Reviewing by inclusive emails can reduce the number of individual records that have to be reviewed.  Beyond the obvious costs savings, the presentation of the entire conversation to a single reviewer gives them the greatest chance of understanding the nuances of that particular chain of communication.

In its most simple form, email threading is an improved method for organizing the data to make your review process more efficient and effective.  We can further improve on this method by utilizing clustering to group the email threads by topic!   As we alluded to earlier, the main goal is to get the complete story, or conversation to individual reviewers so that they have the best possible understanding of the material that they are reviewing. 

Imagine if you had an automated way to identify all the email threads that are discussing a particular topic like “Financial Auditing Standards”.  You could then focus a few reviewers on that topic.   This gives those individual reviewers the ability to accumulate related knowledge which will result in better and more consistent coding calls. 

With the proper application of email threading and clustering you can zero in on the important documents earlier in your case and your review team will achieve higher coding quality, improved decision making and better efficiency!

1Angela Green, Lockheed Martin at Relativity Fest 2015


Leave a Reply