In actuality, each custodian ended up assigned to review nearly 30 GB of data. This sudden and surprising development almost quadrupled the scope of the task.

Hoping to rapidly and inexpensively shrink its data set, the purchaser relied on e-discovery experts FTI Consulting to determine which of the following three options would best complete the task at hand:

  1. Apply FTI Ringtail’s predictive coding tool to its documents. Under this scenario, Ringtail’s analytical engine would process a random sample of documents that an informed reviewer would then code as relevant or irrelevant. Then, using this sample as a baseline, Ringtail would determine if the remaining processed documents were similarly relevant.
  2. Apply the more traditional technique of keyword filtering, using search terms validated by the client and FTI’s experts.
  3. Apply the more traditional technique of keyword filtering, using search terms provided by the Bureau.

FTI expected to generate the best, most accurate results from predictive coding. However, unlike its U.S. Department of Justice counterpart, Canada’s Competition Bureau had never before permitted a company under antitrust investigation to disclose materials processed through predictive coding, and it was unclear if it would set a new precedent by approving this technology.


FTI tested a sample set of data and determined not only that predictive coding was the least expensive option but also that it promised the best precision rate, with an estimated 80 percent of relevant documents likely to be identified. FTI also determined that the least effective scenario would be relying on the Bureau’s suggested highly generic search terms, which yielded many irrelevant documents during the test run. At the very least, the purchaser and its legal team now knew they had to persuade the Competition Bureau to use the company’s recommended keyword search terms. But in the best-case scenario, they would convince the Bureau to avoid keywords altogether and instead embrace predictive coding.

Fortunately, FTI has a longstanding trusted relationship with the Competition Bureau. After FTI presented its compelling test sample results and explained its e-discovery process in detail, the Bureau approved the predictive coding, confident that it was a viable technique. In addition, the purchasing company’s legal team determined it could reduce the number of data custodians from what was originally an unwieldy 42 to a more manageable 14 and still meet the requirements of the SIR. Even so, those 14 data custodians were still responsible for nearly 30 GB of files each, meaning FTI and its client had a large number of documents to process. FTI also used Ringtail’s Mines and Cubes data mining tools as well as its Document Mapper data visualization application to further streamline the review process.

Impact: The company reduced the data set by 95% and produced the relevant documents to Canada’s Competition Bureau on budget and on deadline.

The purchaser started with a data pool of 4 million documents, which was narrowed to 1.3 million after the number of data custodians was reduced to only 14. After processing the documents through Ringtail, FTI and its clients produced approximately 218,000 total relevant documents to the Bureau. FTI stayed on budget throughout the review process and the documents were submitted on time. To conduct its e-discovery process, FTI relies on a dedicated, experienced team of expert reviewers with extensive knowledge of the Ringtail solution.

This results in higher precision rates, which provide FTI the credibility to effectively negotiate with government agencies like the Competition Bureau.