What is the best discovery technique available? Is it the old reliable search terms or cutting-edge analytics that surface the most relevant data?
Let’s look at the strengths and limitations of both methods in the current eDiscovery landscape to help decide.
Search Terms
Lawyers and their eDiscovery partners have been using search terms as an integral part of litigation and investigations for years. Even with advancements in technology, reviewers still regularly use search terms. The method is familiar, and results garnered foster transparency while maintaining a high degree of certainty. As a bonus, the courts accept search terms as an acceptable eDiscovery tool.
Although the pro column is robust, there are just as many issues that users would place on the con side of this debate. First and foremost, it is difficult to know at the beginning of a review which words will pull out the most relevant and high-value documents. This often results in having to redo the process multiple times or requires more manual review to wade through a plethora of irrelevant documents. Vague or context-specific terms will often clutter the results. Another hurdle is that search term tools cannot account for typos, even when it is just one character missed or varied spelling of the same term. This raises risk that the output will have missing documents.
Analytics
Whereas search terms are absolute, analytics do a much better job at parsing through substance, syntax, and spelling. While it may not be time to completely ditch search terms, utilizing analytics on its own or in conjunction with less sophisticated technology is proving advantageous. Many legal teams are already using active learning tools, AI portable models, concept searching, and more to enhance their review workflows. The benefits include ease of use, the ability to jumpstart review, speed, less risk that relevant documents will be missed, scoring consistency, integrated quality checks in the tool, progressive cost savings, and defensibility.
Even with all these benefits, there is still hesitation to put trust in machines to do work that humans have historically completed. Fear of the black box is a hurdle that persists as technology evolves. Analytics relies on algorithms, which some may view as a scary word. However, algorithm simply means a process – and those used in trusted and vetted tools generally provide solid output in a fraction of the time. With more data to review than ever before, this is a compelling feature of analytics.
Making the Right Choice
So, should legal teams be using search terms or analytics for document review? The truth is there is no clear answer. It depends. While AI is gaining more adoption, it may not be available for some teams or be the best choice for smaller cases. Legal teams should strategize with internal stakeholders and their outside partners about what works best and ways to reimagine document review processes.
Remember, when using search terms – whether alone or in conjunction with another tool – look at the capabilities search term solutions offer. There are varying degrees of sophistication on the market. A provider with deep document review expertise can advise on models that incorporate dictionary features, concept exceptions, and hit count analyzers. Best practices include putting thought into the keywords used, considering gaps, testing out a small sample before applying the keywords to an entire dataset, and looping in opposing counsel to improve transparency efforts.
The potential of analytics tools can be maximized by taking a three-step approach. First, thorough preparation is crucial. This includes thinking about the end goal in mind, assessing the data available, and focusing on fact development. For example, if settlement is on the table, it may be highly beneficial and cost-saving to use analytics for early case assessment. Next, training is key to enhancing effectiveness and garnering consistent results. Lastly, it is important to understand and anticipate weaknesses, as no tool is perfect. For example, short messaging data can lead to collection and review obstacles if the right technology or expertise is not available.
A recent study discussed in articles published by ALM and EDRM found that the eDiscovery market is valued at over $14 billion and human document review makes up at least 60 percent of those costs. This should be factored into eDiscovery strategies and legal technology budgets. Being armed with actionable insights empowers legal teams to make informed decisions about which tools are optimal for their organization. Legal teams should know about technology options and have resources in place to guide
[View source.]
