This post explains how a “rolling productions” approach to both small and large eDiscovery projects can help you prepare for common challenges.
Legal project managers, litigation support professionals, and other eDiscovery practitioners are well aware that their projects are often unpredictable and that outside forces drive consistent requirement changes. While originally implemented in the software development industry, Agile development methodology can help address this challenge in eDiscovery.
What is Agile Development?
Basically, Agile development is a set of principles that were created to allow software and application development teams to deliver their products more frequently and with a significantly higher quality level.
Development teams do this by delivering projects in phases, essentially prototypes, or incomplete products. The results of these phases are reviewed by project stakeholders and feedback is used to restate and define—or redefine, as needed—the project requirements after/prior to each phase. The prototype delivery concept and continuous improvement model are core principles of Agile Project Management.
Using Agile Project Management Concepts for eDiscovery
You can use this prototype model in eDiscovery to refine the scope, sharpen the timing estimate, and save costs by limiting superfluous data processing and review. Successful workflows accomplish this by leveraging “rolling deliveries,” or “rolling productions.”
A rolling production/delivery is a subset of the required files to be produced to the opposing party/the court in the eDiscovery agreement. They have been utilized as the practical response to expanding scope from the initial meet and confer agreements, as well as, the sheer volume of ESI that is being generated by today’s users.
The key is to treat each production/delivery as a prototype, or phase deliverable, prioritize your data sources, and analyze the results to determine the best course of action.
An Agile Approach for Smaller Datasets in Discovery
An approach for smaller datasets (between 10,000 and 25,000 records) is to prioritize your initial searches to a key custodian or sources. Then, use what is found in that set of ESI to determine the next steps. The results of the initial /deliverable can be analyzed to shape the criteria for the next phase. For example, email to/from fields can be used to determine additional custodians; key phrases can be used to refine the search criteria; modified/sent times can be used to limit the date range, etc.
An Agile Approach for Larger Datasets in Discovery
Rolling productions/deliveries for larger datasets (over 25,000 records) can be approached with a Predictive Analytics (a.k.a. Technology Assisted Review (TAR) or Predictive Coding) workflow. A subject matter expert (SME) would review a subset of ESI (key custodians) to identify the relevant and non-relevant documents. The Predictive Analytics engine would find documents with similar characteristics and assign scores to the rest of the documents based on the SME’s relevant/non-relevant decisions. Review teams then prioritize the review of documents based on these scores, looking at the documents with the highest scores (most relevant documents) first. Analysis should be performed on the most responsive documents in order to identify similar ESI and to focus the next round of review on the sources with the highest relevancy potential.
Active team involvement with the analyses of rolling productions/deliveries, and changes to requirements and scope is an important piece of the puzzle. Case teams need to actively communicate with their opposition and the court about the results and decision points of their initial ESI assessments and Predictive Analytics workflow. All adjustments to the agreed upon protocol must be documented.
By employing the workflows and best practices described above and utilizing the Agile Project Management concept of prototyping to comply with eDiscovery agreements, you can minimize the burden on all stakeholders and achieve favorable outcomes for your clients.