Conducting True Review Of eDiscovery: Where Strategy And Technology Meet

Wednesday, December 11, 2013 - 15:03

The following is a brief summary of Conducting True Review of eDiscovery, a webinar sponsored by Kiersted. The discussion was moderated by Josh Gilliland, e-discovery attorney in San Jose, and author of the Bow Tie Law Blog. Participating were Gail L. Gottehrer, Axinn Veltrop Harkrider, LLP, a class action attorney focusing on complex civil litigation and an e-discovery thought leader; Joanne Lane, a recognized legal and e-discovery technology expert with over 20 years of law firm and corporate legal department experience; and Paul Prewitt, Kiersted’s eDiscovery Review and New York Operations Manager with 15 years of experience with high-level complex litigation. 

Josh Gilliland began the discussion with a quote from Judge Andrew Peck in Moore v. Publicis Groupe: “The objective of review in e-discovery is to identify as many relevant documents as possible, while reviewing as few non-relevant documents as possible.” He then asked the participants their advice regarding leveraging the Rule 26 conference to achieve Judge Peck’s objective.

Cast The Net Wide

Ms. Gottehrer emphasized the importance of conducting a meaningful Rule 26 conference, rather than treating it as a “drive-by” – i.e., something you have to do in order to comply with a court order. She pointed out that in this age of e-discovery, it is very dangerous to put off major decisions. Rather, it’s wise to gather as many stakeholders with knowledge of the relevant data as early as possible in the process, not only from an e-discovery perspective, but also from a strategy perspective, as questions raised by non-lawyers can help a litigation team frame the issues. Teams may also learn that an individual hitherto considered tangential to a case is actually far more central, which means that the person’s data must be preserved. Such information may also turn out to benefit the team’s strategy.

Ms. Gottehrer went on to point out that in asymmetrical data scenarios, it is beneficial for the defense, which has considerably more data than the plaintiff, to know what it has in terms of data so that it can try to reach an agreement that data that is inaccessible or that would be unduly burdensome to collect, review and produce need not be included in discovery, or at least to set parameters around the scope of discovery. If this is not addressed early on, a team risks being served with a motion to compel, which could wreak havoc on the schedule in the case. Similarly, the format of production should be agreed upon as early as possible.

Getting The Lay Of The Land

Ms. Lane agreed with this Rule 26 “wide net” approach, as each individual brings a unique perspective to the table. Like Ms. Gottehrer, she believes that attorneys should go into a Rule 26 “empowered” with knowledge of the data, with an understanding of the costs and burdens that may be involved in production. This will enable them to effectively leverage their arguments to keep discovery reasonable and cost-effective. In her practice, Ms. Lane works with business representatives and application owners to get the “lay of the land” – gathering metrics around volume, effort and costs, and identifying methods to find a germane, responsive set of data.

Ms. Lane also observed that as an e-discovery professional, she and her colleagues often receive requests that might be for a Rule 26 conference or other submission, but the lawyers do not always provide the context for the requests – which would greatly enhance the efforts of the support team. Mr. Prewitt weighed in to agree both with Ms. Gottehrer’s point about “building substance” into the Rule 26 and with Ms. Lane’s assertion that, from the technology provider point of view, understanding the context of a request can be a game changer.

Early on in the case, Ms. Lane said, data experts can play the critical role of enabling attorneys to be strategic in their agreements around the scope and format of discovery – and ensuring that the impossible is not promised. The bottom line consists of educated decisions, “with the right people weighing in at the right time.”

Search Terms: A Continuous Dialogue

Ms. Lane reminded the audience that there is no 100 percent guarantee for search term validation – that technology cannot, as some assume, definitively locate all relevant material in one simple step. It’s important to keep in mind, first, that perfection is not the standard – “reasonable and defensible” is; and, second, that search is an iterative process best executed by a combination of machine and human in which a variety of search engines, for example, keyword searches and clustering, is utilized. Ms. Lane described having a high level of engagement with the search process, and recommended ongoing frequency analysis and re-evaluation of search terms and methods. Ask questions, she said: Does the business use unique jargon for its industry? Or, might a certain search term be more meaningful if combined with others? She emphasized the need for a “feedback loop with actual people reviewing for substance.” While attorneys sometimes seek smaller document sets in order to maximize cost-effectiveness, this might not be the best strategy for winning the case, she cautioned. While it’s important to reduce the number of irrelevant documents, teams should be careful not to inadvertently filter out relevant documents in that process.

Ensuring that participants in the process are “speaking the same language” is an important aspect that is sometimes overlooked. Different search tools use different syntax, and all team members who discuss, develop and execute searches should be as specific as possible with each other. Ask clarifying questions, and never make assumptions.

Afterward, the results should be validated to make sure all parties are in alignment. That said, while identification of search terms may be done in preparation for a Rule 26, there should be a mechanism for dialogue to continue throughout the matter, with search terms being revisited as frequently as necessary.

Getting Your Arms Around The Case

Ms. Gottehrer allowed that in her experience, this kind of agreement rarely occurs at a meet-and-confer, which usually takes place too early for her to be comfortable discussing search terms. Rather, she often runs terms herself in order to get her arms around a case. If key custodians are providing certain information, running the search terms can help to find hot documents, to confirm or raise questions about what she’s been told or what is alleged in the complaint, or to test her own hypotheses in the broadest sense.

She agreed that the process should be a dialogue and gave some tips as to how this should proceed. When defense counsel asks what search terms opposing counsel would like to request, it’s likely, for example, they will ask for very broad search terms, at which point the defense may counter with a request that this search be refined, otherwise the defense may request a cost-shifting order. Such a back-and-forth demonstrates good faith and may help the parties avoid a motion to compel or motion for protective order. Discovery motions are time-consuming and their outcomes frequently leave all parties unsatisfied: it’s better to provide the documents the plaintiffs want, whether or not they will prove helpful. Over time, Ms. Gottehrer re-emphasized, perspectives – and therefore search terms and names – will change, so one should expect the process to be revisited over the course of the matter.

Initial Disclosures And Effective Early Case Assessment Technology

Ms. Gottehrer noted that early case assessment is the litigation team’s opportunity to “cut through to what’s important and focus more on what you want to know” and to develop an “intelligent perspective” about the data such that the team can be aware of concerns as well as the preferred scope of discovery before heading into the Rule 26 and subsequent court conferences.  

When asked about technology that supports attorneys as they identify their initial disclosures, Mr. Prewitt noted that many products can help a litigation team to understand its data by categorizing documents and other data, as well as by enabling them to see the origins of data and who’s speaking with whom – which may help the team uncover important relationships. Such tools may prove critical to developing the kind of preparedness Ms. Gottehrer described.

Mr. Prewitt then demonstrated some of the tools Kiersted Direct has built precisely to facilitate effective early case assessment. Mr. Prewitt’s first slide was a screenshot of the Direct Domain feature, which shows the user where the bulk of the data is coming from, including the domain and frequency of email communication between parties. (Click to enlarge images.)

The next screenshot showed Kiersted’s Direct Location Culling, which enables users to see the data as it was housed and organized in the native environment from which it was collected – giving users a feel for how day-to-day business was conducted.

Mr. Prewitt’s last slide, a screenshot of Kiersted’s Snippets tool, spoke to Ms. Lane’s point that informing the technology provider of the intent behind a search is crucial, as it will minimize false positives and maximize relevant data: hyperlinks embedded in search terms open text for a user to instantly see the term in context.

Form Of Production

Despite the amendment of the FRCP Rules in 2006, form of production remains a contested issue. When asked about lawyers’ “fixation” on static images and Bates numbers, Ms. Lane responded that this was more a case of being accustomed to these standards, which worked well for paper files. She remarked that people do still seem to insist upon tiff and other image files – essentially electronic equivalents of paper – rather than native production. The perception is that native production presents many challenges, especially with regard to tracking; verification, especially in the context of presenting evidence; redaction; placement of endorsements and confidentiality legends; and “hidden” data, such as tracked changes.

Ms. Lane noted that it may also be that litigators prefer to avoid native files for strategic reasons but the primary reason may well be, quite simply, that native files are not requested by opposing counsel. She concluded that the resistance to native file production may come down to educating lawyers that many of the above challenges may be mitigated. Ms. Lane observed she had noticed an uptick in “hybrid” productions, with a mix of tiff files along with certain native files, especially Excels and PowerPoints. Finally, she said, litigation support service providers ought to build better accommodation of native files into their solutions so that time and money are not wasted on converting files between various formats.

Privileged Review

“Privilege is tricky,” Ms. Gottehrer commented, so she prefers to use a tool that will allow her to batch out privileged (or potentially privileged) documents so that specific reviewers can be assigned to these high-risk files to ensure that nothing slips through – such as in an email thread in which all explicit identifiers of an attorney have disappeared, but his paraphrased advice remains.

Mr. Prewitt responded that once the potentially privileged documents have been batched out and reviewed for privileged/not privileged, advanced analytics (such as those in Kiersted’s review tool) may be deployed to ensure quality control: indeed, he had found the Kiersted tool to be effective in identifying email threads much like the one Ms. Gottehrer described. The tool may of course also be used to identify HIPAA and other personal identifiable information.

Ms. Lane commented that an issue coding tool is key when working with document requests from agencies such as the DOJ and SEC, and that, while in her experience it’s best to keep issue codes simple, using the tool is also highly useful in enabling the review team to produce documents organized and categorized per the requesting party’s preference.


Mr. Prewitt said that while most of the review tools provide some level of productivity tracking, they are rarely adequate for the needs of the review team, as they require review managers to pull data out of the tool and load it into other systems. Kiersted’s TrueReview dashboard shows the key metrics – individual review rate, overturn analysis, review by day – at the user’s fingertips. TrueReview allows the user to quality control the work the reviewers are doing early on, so the team can identify who may need directions to better understand the merits of the matter and improve the quality of the coding.

Ms. Lane said she strongly encourages users to leverage their vendor tools as much as possible, and suggests that they engage with their vendors directly and make requests for reporting capabilities to be added as needed. And, she was glad to see tools that enable customers to gain a view across their entire litigation portfolio, so that trends may be identified, matters categorized, etc., as this allows them to take the intelligence gained from one matter forward to new cases.

Please click here to view this webinar, where listeners can gain deeper insights on all of the matters summarized above from the participants. Kiersted’s whitepaper, Conducting True Review of Ediscovery, which contains information about how Kiersted Direct addresses the issues discussed above, may be downloaded here.


For more information, please email Linda Gordon at