Meta-e Discovery was able to achieve a 29% cost savings and a 48% reduction in total time spent on the first-level review of ninety-one thousand documents for a plaintiff’s litigation production.
Our client, plaintiff’s counsel working with an outside review vendor, needed to review 91,000 documents for an upcoming production. They had already wasted precious time working with a separate vendor using a linear review workflow and were now feeling not only the burden of sunk costs incurred but significant pressure to complete their review and production in a timely manner.
In a typical linear document review workflow, the review team would need to look at each of the 91,000 documents one-by-one in order to complete the review in defensible fashion. With a team of eight reviewers plus a review manager it would take approximately six weeks to complete first level review. Linear review also requires some means of apportioning the documents to the review team which can lead to a less than efficient process.
Active Learning instead immediately provides the reviewers with documents to start coding. As each document is reviewed, the analytics engine begins putting those documents most likely to be relevant in front of the reviewers in real-time. Each coding decision made by the review team serves to progressively train the system. Eventually, the vast majority of relevant documents have been reviewed while any remaining unreviewed documents can be skipped with high statistical confidence that they are not relevant. This has the practical effect of greatly reducing both the time spent on the review and the associated costs.
In this particular matter, only 47,000, or 52%, of the original corpus of 91,000 documents ended up requiring eyes-on review. This cut the effective time needed to complete the first level review down from six weeks to three. In addition to the time saved, the reduction in document volume yielded a savings of $22,000.