Document Review in 2025: ediscovery to infinity and beyond
Guest article by Kara Kirkeby & Jodi Vickerman of Kroll Ontrack
E-discovery continues to evolve at the speed of light. With developing state and federal rules, new opinions from the bench, and a constant stream of technological innovations, law firms and corporate clients must evolve their practices to stay competitive. Of the many key e-discovery practices that are evolving, document review is currently undergoing rapid transformation. Fifteen years ago, practitioners were reviewing boxes of paper documents in dusty warehouses. Law clerks were hired to sift through the firm’s files stored in the basement and to then discard past clients’ paper discovery files that were no longer needed. Now in 2012, that outmoded relic that has been vehemently replaced with innovative technology that is fundamentally altering the manner in which document review is conducted. In fact, the roles played by technology and key personnel involved across the Electronic Discovery Reference Model (EDRM) are being redefined as we speak.
Today’s metamorphosis, however, is only the tip of the iceberg. Let’s venture into the future to envision what document review might look like 15 years from now. If history repeats itself, we expect a world unknown to any legal practitioner alive today…
Staffing and Administration: A Few Good Managers
Today, legal document review clings to the brute force, lawyer “eyes on every document” approach where a multitude of licensed attorneys swoop in, receive training on the case and painstakingly review every document for privilege, relevance, and key issues. These attorneys – who may work directly for the law firm or be under a contingent contract arrangement with a document review service provider – make document coding decisions for hours on end, often under tight timeframes. The exact number of reviewers needed on a project is determined by the volume of documents to review and the time allotted for a final production deadline. It is not uncommon for document review projects to require 100 or more reviewers to work weeks on end to meet aggressive deadlines.
In addition to a large staff, in 2012 document review teams rely on quality document review managers and administrators. These key personnel, working closely with legal counsel representing the client on the matter, aim to get a project done as quickly and accurately as possible by investing heavily in the people, process, and quality control measures implemented to conduct document review. These managers hire reviewers, organize training sessions with counsel, assign document batches, answer questions, validate quality control, and prepare final production sets – non-trivial work to say the least. As such, the responsibility lies heavily on this individual – or individuals – to assure all bases are covered.
Document review can account for 50 to 90 percent of total e-discovery costs, but advancing technology continues to maximize the efficiency of review while reigning in the often ballooning cost of brute force review. As such, continuing advances are likely to outdate the existing review staffing model to the point where the current practice will be unrecognizable in 15 years. By 2025, we predict technology advancements capable of reviewing thousands of documents per hour, as compared to human attorneys reviewing 50-80 documents per hour today. Consequently, the reviewing attorneys previously involved in the traditional linear document review will instead focus their efforts on managing documents by translating key aspects of the litigation into principles that the advanced “smart” algorithms can interpret. Then, these attorneys will analyze statistical metrics and reports associated with the computer’s assessment of the document set.
As staffing resources reflect this shift, the ability to implement technology to quickly and efficiently filter out non-relevant documents will become the primary factor in determining the amount of time a project will require in 2025. Accordingly, futuristic technology reviewing 30 megabytes will require the same amount of time and input from the review manager as 30 terabytes of data would, and data volumes will no longer prove a deciding factor in the length of a document review project.
While greater investment in review technology will reduce the need for a large team of reviewers, the role of the review manager will prove exceedingly important. Even in 2025, it is unlikely that machines will fully replace the need for legal advocacy and analysis, even in document review; however, this role will change significantly, requiring more expertise in the review technology used and less emphasis on managing a large group of reviewers. By minimizing teams of human reviewers, the need for extensive quality control and assessment will also fall by the wayside. Thus, the entire composition of a “review” team will evolve from a large staff of brute force reviewers to a small team of legally skilled, technologically savvy and statistically inclined attorneys.
Technology: Smarter Machines, Lower Completion Time
The driving force behind the evolution of document review will be the technology employed. Current document review technology involves software platforms designed specifically for linear review. While the linear document review model is the most common practice today, numerous e-discovery authorities are supporting the use of “intelligent” review technologies driven by powerful computer algorithms. Despite the existence of these advanced processes, many practitioners remain on the sidelines, waiting for the bench to issue an official opinion condoning its use. However, a recent conference between the parties and Magistrate Judge Andrew Peck in Da Silva Moore v. Publicis Groupe  discussed the use of this technology. The legal community may be on the cusp of formal case law discussing and supporting its efficacy.
Regardless of whether Da Silva signals this shift in 2012, once this technology rises to prominence – certainly before 2025 – linear document review will become obsolete. Discoverable documents will be identified from within corporate information management systems through extremely powerful, intelligent algorithms driven by attorneys with high technical aptitude. These highly advanced technologies will likely process terabytes of data and analyze the complete set of an organization’s business documents with minimal input (such as the type of matter and filed court documents) while asking a managing attorney questions about specific data characteristics for analysis. The developed algorithms will be so highly complex and data-specific that it will generate entire production sets after viewing only a handful of documents. In the grand scope, by 2025, the entire review process will take mere days or hours – rather than months or years – to complete.
Review Methodology: Shifting Paradigms
With significant changes to technology, staffing numbers, and staff composition, review methodology will see a massive change. Under the current linear method, attorneys review each document in the order they were processed. This model often requires numerous attorneys to review and search for the proverbial “needle in the haystack” from massive amounts of data. This approach typically requires exorbitant amounts of money to ensure that each document is reviewed at least once – and more commonly two or three times.
The linear review model will eventually prove unnecessary and go the way of the dinosaur. By utilizing algorithm-based searches, documents will be found and selected by the quality of their content. As mentioned above, once a quality-based review process becomes the norm, document review will take drastically less time to complete. In contrast to the linear review method, a finely-tuned algorithm will quickly find relevant materials, determine whether they are privileged, and produce them with minimal turnaround time.
Changing review methodologies will also significantly affect the considerations regarding the location and costs of document review. Today, the location of the review is a critical operational decision. Many law firms prefer review projects take place in close proximity to their offices or, at least, in a single location. However, as review teams continue to decrease in numbers, the need to conduct reviews close to the legal team will become unnecessary. Instead, review consultants will meet with clients via the web, accessing data sets and providing metrics about the nature of the data, without necessarily meeting in person.
All of these changes will bring about greatly diminished costs. Currently, employing numerous contract attorneys for 40-plus hours a week and an average rate of $50 per hour (or more if law firm associates are utilized) significantly drives up the cost of litigation. Even a dozen contract attorneys at this rate for a multi-week project produces a sizeable bill for document review alone. Conversely, a project driven by intelligent review technology would only require a handful of associates or senior attorneys training the algorithms and guiding the powerful technology. While these document consultants in 2025 will cost more per hour than any single contract attorney’s billable hourly rate, the number of hours needed to complete the review and production will be reduced significantly and minimize the overall cost of the “review” in 2025.
It is difficult, if not impossible, to know precisely what the next 15 years will bring in any area of legal practice. However, just as today’s document review methodology looks nothing like the “boxes of paper documents in a dusty warehouse” approach of 15 years ago, we can be certain that document review in the future will look nothing like it does today. Hesitancy to embrace such change often proves futile, as those who fail to innovate go extinct. Prospectively – and in order to survive – organizations and practitioners should return to the present day 2012 and determine what steps they must take to prepare for the future.
 See, e.g., Andrew Peck, Search, Forward, available at http://www.law.com/jsp/lawtechnologynews/PubArticleLTN.jsp?id=1202516530534&slreturn=1; The Sedona Conference Commentary on Achieving Quality in the E-Discovery Process, available at http://www.thesedonaconference.org/dltForm?did=Achieving_Quality.pdf
 Judge Peck Addresses Predictive Coding in Federal Court Order (Feb. 2012), available at http://www.law.com/jsp/lawtechnologynews/PubArticleLTN.jsp?id=1202542221714&Judge_Peck_Addresses_Predictive_Coding_in_Federal_Court_Order=&et=editorial&bu=LTN&cn=LTN_20120215&src=EMC-Email&pt=Law%20Technology%20News&kw=Judge%20Peck%20Addresses%20Predictive%20Coding%20in%20Federal%20Court%20Order&slreturn=1
For the latest on Judge Peck and the Da Silva Moore case visit http://aceds.org/news/da-silva-moore-plaintiffs-move-to-recuse-judge-peck-in-predictive-coding-case