Your single source for new lessons on legal technology, e-discovery, and the people innovating behind the scenes.

Now in Relativity Analytics: 3 Customer-Driven Enhancements

Ann Marie Lane

It’s undeniable that analytics tools, especially email threading, should be used on every e-discovery project. They make reviewing and QC’ing documents faster, easier, and more consistent, and we’ve seen time and again that the most successful e-discovery teams use analytics.

That’s why we invest so much into Relativity Analytics, and why we’re constantly looking to our customers for ideas to make it better. Even the smallest changes can have a huge impact, and who better to hold us accountable for sweating the small stuff than the people who rely on Relativity every day?

Here’s a roundup of some improvements we made in the past month—all of which started as feedback from the Relativity community.

1. A new cluster visualization option.

analytics-dial-thumbnailCluster visualization is one of the most unique features in Analytics, giving you an interactive map of conceptually similar documents. Customers use cluster visualization for everything from organizing and prioritizing review to performing quality control.

However, when drilling into a particularly complex cluster, it can sometimes be hard to see what’s there. So, we’ve addressed that with a new visualization option called the dial, now in beta and available in analytics-dial-2-thumbnailRelativityOne and coming next month to Relativity 9.5.

The dial displays your clusters in an easy-to-read circular map, sectioned off and labeled by concept. Drill into each label to explore your data, or hover over it to get more details on what’s inside. Like the circle pack, you can also apply heat-mapping if you want to see, for example, where your responsive documents lie. 

2. A new email thread visualization tooltip for faster, more accurate QCs.

Since unveiling email thread visualization at Relativity Fest last year, we’ve gotten a ton of feedback. Getting a picture of when and where a conversation branches off not only saves time, it makes the coding and QC process easier and more consistent.   

Though the response has been positive, our customers have been vocal about small improvements that would make a big difference, like having a way to more easily detect coding inconsistencies across all documents and attachments in a conversation—not only parent emails.

We’ve got you covered in two ways.


First, if a message’s coding doesn’t match the coding on its duplicates or attachments, a big red exclamation point will sound the alarm that something might not be right.

Next, simply hover over a message to see the coding information for all its attachments and duplicates to quickly identify where the discrepancy lies.


3. Store multiple structured analytics sets.

Consider this scenario, brought to our attention by a Relativity user:

We often load multiple productions into a case and would like to run near-duplicate identification to help cull the entire universe, then rerun it at the end of the review with a lower similarity percentage to help with QC.

So … what’s the problem? Well, previously in Analytics, you could only store one set of structured analytics results—meaning if you wanted to run tools like near-duplicate identification or email threading multiple times on different data sets or with different settings, there really wasn’t a good way to do it (at least without investing some manual work).

Not anymore.

Now, you can store the results of multiple structured analytics sets and even set up views that capture those results to easily see differences among the different sets. No manual work required.

Take the near-duplicate identification example. Now, that customer’s team can run near-duplicate identification at a high similarity threshold (e.g., documents must have a 99 percent similarity to be flagged as near-duplicate) to help with immediate culling. Then, at the end of the review, they can rerun it for QC, this time changing the similarity threshold to a lower score.

The same technique applies to email threading. Say you want to run email threading on different productions—all combined or each individually. Now, when visualizing a thread, you can select a different email threading structured analytics set—such as opposing counsel’s production—to easily jump between the sets.


These three enhancements are just a snapshot of what’s going on in Relativity Analytics. In the past month, we’ve also introduced new security permissions to give you precise control over who can see email thread visualizations, streamlined the process of setting up analytics indexes, and more. Check out our What’s New page for more details.

Ann Marie is a senior manager of product marketing for text analytics and review in Relativity, working closely with the product teams to understand customer needs and strengthen the platform. She has nine years of experience marketing for the technology and software industries.