Since 2014, a Fortune 500 medical equipment management and services company has seen a dramatic uptick in the volume of data it hosts. In order to control – and hopefully reduce the associated costs of having more data – this company and Legility have become much more intentional with how data is managed, how outside counsel accesses data and in deploying better workflows, automation and advanced review technologies. The results: When the increased volume is paired with total Legility billings since 2014, cost per GB has been reduced by 52% entering 2018.
When we looked beyond just the Catalyst hosting, we noticed a steady decline in per GB charges. We used the total billed per year divided by the total number of GB hosted in Catalyst. In 2017, we reduced the cost per GB by $20, a 31% decrease – far outpacing the periodic hosting charge reductions we implemented. In fact, if the numbers hold true for 2018, we will have reduced the client’s total cost per GB by well over 50% in four years. To explain where this savings is coming from, we separated the Catalyst hosting costs from all other billed items. As was expected, the hosting costs ran nearly parallel with the GB hosted, with divergence occurring only when the aforementioned hosting price reductions were applied. The other costs were subject to much more variance. When analyzing the trend line, however, what we’ve billed for all other services has declined despite the increase in data volume. When factoring all services and in total billings other than Catalyst hosting charges, in 2014 we averaged $33.89 per GB. In the last half of 2017, Legility averaged $16.19 per GB despite having 28% MORE data than in 2014. That’s a 52% reduction in billings while hosting 1 TB more data, and results in $600,000 in annualized savings.
How We Did It
The key to successfully and consistently reducing total billings has been the overall management of the data. With a better understanding of the data corpus, how it resides in Legility’s hosted environment, and how we can access and pull from the data when needed, we have created a much more efficient data management process. This includes culling data that is not needed, an initiative that further helps to keep overall data volume under control. The downstream impact of better initial culling is that, when needed in ECA and during a matter, we can pull more relevant data – not junk – from the corpus when it is needed, reducing the size of stored data in the hosted environment. The additional benefit is that this makes attorney review more efficient, as well.
Culling efforts and Legility initiatives to identify the most relevant documents sooner during a matter have also paid significant dividends. With the consistent deployment of the Catalyst Insight Predict continuous active learning platform, attorneys are reviewing fewer documents, and what documents they are reviewing are much more likely to be relevant to the case. As seen in the chart below detailing four matters on which Predict was used, we averaged eliminating 71% of data before attorney review. This is independent of the richness of the data corpus which, as you are aware, can dramatically influence the number of documents that must be reviewed. Additionally, of the documents the Legility Managed Review attorney team saw, just one in four documents (26%) was not relevant, on average. This means that 75% of the documents they reviewed were relevant to the case, an outstanding result that highlighted the benefit of using a robust platform like Predict.
When totaled, these initiatives will continue to help this client reduce its per GB and attorney review costs, and Legility will continue to find ways to be more efficient and knowledgeable about the client's data footprint.