By Sam Anderson

The classic movie Glengarry Glen Ross coined the phrase “Always Be Closing” for the salespeople in a real estate office where they are given a motivational (and graphic) speech about how to succeed in sales. Blake, the motivational(?) speaker stresses the importance of focusing on activities that lead to closing accounts by saying:

“A… Always, B… Be, C… Closing. ALWAYS BE CLOSING!”

In eDiscovery projects, ABC is just as important to success, but with a different C — A… Always, B… Be, C… Culling. ALWAYS BE CULLING! There’s also a D, but we’ll get to that.

The Debate on When to Cull

Out of an over-abundance of caution, the historical eDiscovery mindset has been to “collect everything, then figure it out later.” The process is this:

  1.  collect any potential custodian’s entire data corpus;
  2.  load it all into a review platform;
  3.  use a combination of culling and searching approaches to reduce the population to identify potentially responsive documents;
  4.  review those documents; and
  5.  produce documents deemed responsive but not privileged. This so-called “safe approach” means all culling occurs downstream in the most expensive phases of discovery.

However, because of the exponential growth of Big Data, the cost of the “collect everything” mindset has become overwhelming. As a result, legal professionals are using alternatives, such as choosing to employ index- and search-in-place technologies upstream to cull out clearly non-responsive ESI. This early screening leaves only potentially responsive ESI to be collected and moved downstream to discovery activities such as processing, hosting, and review itself. Reducing volumes where the data resides — before collection — is considered the most cost-effective approach for culling ESI.

So, which is correct – culling upstream or culling downstream? The answer is – both! And much more.

The Prism Periscope

With so many discovery challenges today, including Big Data and an increasing variety and complexity of data sources (including mobile devices, collaboration/chat apps, and other cloud solutions) that are routinely discoverable, you can’t limit culling to just one step in the discovery process. You should… ABC — ALWAYS BE CULLING!

Of course, culling your data collection incorrectly can lead to issues with opposing parties or the court. Therefore, understanding how to perform data reduction defensibly is imperative. The Prism Periscope approach, developed by Prism Litigation Technology’s eDiscovery experts, leverages a combination of technology and best practices to perform culling throughout the eDiscovery lifecycle, thereby defensibly reducing the amount of data moving downstream, while also providing powerful analytics and search capabilities that quickly locate relevant information. The more you can take advantage of opportunities to cull non-responsive information, the easier it is to identify the key evidence in the case.

The Prism Periscope approach identifies six key areas where culling activities can occur:

  1.  Legal Hold / Preservation: Can culling actually begin at preservation? Of course! By conducting custodian relevancy assessment interviews and data source classification, sources of data or even entire custodians can be identified that need no preservation, much less collection. Preservation is not only expensive, but over-preservation can paralyze an organization’s ability to effectively execute a data retention and destruction program.
  2.  Proportional Discovery Assessment®: The patented workflow in Evidence Optix® can be utilized to develop a defensible proportionality discovery plan that limits scope through accurate and fact-based relevancy, burden, and cost analysis. This adds a defensibility layer to support decisions regarding which custodians and data sources to preserve and collect.
  3.  Language-based Analytics for Search Term Optimization: Crafting well-designed search terms and logical expressions is an underappreciated science. Poorly crafted terms often lead to overbroad or underinclusive results that are costly. By learning the “language” of a case quickly, discovery teams can more accurately target relevant information while isolating the noise. This process also leads to an earlier understanding of pertinent topics, connections, timelines, and patterns to help develop the case strategy. Time spent here results in a more surgical and targeted approach, dramatically reducing the volume of data collected and expediting the eDiscovery timeline.
  4.  Targeted Collection + Analysis: Index in-place technology, such as X1 Enterprise Collect™, enables rapid identification, searching, and collection of data in place behind the corporate firewall. Laptops, desktops, file shares, email, and cloud data sources, including both metadata and text, can be searched in real-time to quickly understand the story within the data itself, without disruption to employees. After analysis is complete, use targeted collection to collect only what you need to create a defensible ESI deliverable that can be uploaded directly to Relativity or exported to your review platform of choice. This targeted collection results in significantly reduced costs.
  5.  ESI Processing + Data Minimization: Culling also occurs in conjunction with processing. The ability to leverage data minimization methods (e.g., deduplication, de-NISTing, email thread analysis, metadata filters, entity recognition, pattern matching, ROT and non-data files) can eliminate a significant amount of redundant or irrelevant data that doesn’t need to be reviewed.
  6.  Accelerated Review: The institutional knowledge gained throughout the Prism Periscope process can be leveraged to optimize the review phase for speed and accuracy. The case’s linguistic narratives and search methodology, in tandem with predictive coding techniques such as TAR/CAL, accelerate review workflows through the advanced classification, clustering, and review of a much smaller document set, thus reducing the number of documents requiring eyes-on review. Finally, integrated quality control and statistical sampling procedures maximize efficiency and precision, thereby lowering associated costs.

The Prism Periscope approach can be illustrated as follows:



The key to a successful approach for culling throughout the eDiscovery lifecycle is defensibility. The Prism Periscope is a consultative approach which is customized to the needs of individual organizations and cases, leveraging best practices and technology to maximize the benefits of the culling activities, without introducing unnecessary risks. Prism’s clients save as much as 70% over traditional eDiscovery costs using the Prism Periscope approach!

In fact, the key to a successful eDiscovery project is actually ABC… D! A… Always, B… Be, C… Culling, D… Defensibly. ALWAYS BE CULLING – DEFENSIBLY! You’ll never regret bringing in experts who know their ABC…Ds!

Sam Anderson is the Vice President of Operations at Prism Litigation Technology