Legal analytics is trending! And the data underlying it is the topic of this week's Feed.

What's new?

This week Baker McKenzie, a global litigation firm, announced a partnership with SparkBeyond (AI platform, and a good friend of Demyst) to bring AI to the forefront of legal analytics. Specifically, Baker McKenzie aims to “deliver machine learning enabled judgement.” While the firm is bringing these capabilities in-house, there are a number of data vendors that are providing the raw ingredients as well as some level of AI solutions in the market today (see our data landscape to learn more).

What are the use-cases?

There are many. Which case types are more likely to result in settlements? Which legal firms don't have good records against certain judges? Which insurers have a longer cycle time and across which business lines? Based on the use-case, insurers can allocate resources and optimize the number of cases that go to settlement vs. trial (which could be lengthy and costly). Insurers also love this data because it enables them to benchmark their performance vs. their competitors, by line-of-business, and by geography. The ultimate insight would be a change in legal strategy (e.g., a choice to settle all personal injury cases in the Central District of California, whenever an insurer faces a certain litigation firm) that produces process efficiencies.

What data is available?

Some of the common attributes are defendant, plaintiff, case type, judgment, filing date, and judge (we'll call these the "basics"). Beyond the basics is where the secret sauce of data vendors comes into play. We've listed a few nuances below:

• Docket engine: More sophisticated vendors (e.g., UniCourt, Gavelytics, and Premonition AI) have "docket engines” that mine the text of case filings to identify special tags, such as: covid-related, bad faith, etc.
• Federal vs. State/Local: Federal case filings are standardized (PACER), but state and local filings are not (every jurisdiction is slightly different, with varying disclosure requirements). For example, certain states require defendant disclosure, while others don't. This creates an imbalance when trying to execute aggregated analytics across defendants in a certain industry (e.g., insurers).
• Aggregated Case Duration: If a case is appealed, what is the true duration? Vendors invest significant time and resources to accurately quantify metrics such as these.

Demyst's Perspective?

We've partnered with several insurers across legal analytics and we've seen these organizations benefit from a deeper understanding of their own performance relative to their competitors across key metrics, such as bad faith and case duration. We're also optimistic that, over time, state and local filings will become standardized and overall data quality will improve.

Other Recommended Feeds