Categories
eDiscovery

Chat Data Review – Adding context to modern eDiscovery

Discovery is a complex world where the number of innovative new data types continues to grow every year. The use of collaboration platforms in business is one of the most recent trends to impact the eDiscovery community. A survey by Wakefield Research confirms that hybrid, flexible work models are here to stay.

Similarly, in a survey conducted by Slack’s Future Forum, they found that email is losing its status as the default business communication tool. Almost half of IT Decision Makers believe that email will be replaced as the primary communication tool by 2024. And, when forced to choose, 36% of users would rather go without work email than without Slack, Teams, or other collaboration solutions.

These business trend findings indicate that eDiscovery professionals are currently experiencing a major paradigm shift in the types of data they deal with daily. In 2021 alone, over 600 billion chat messages were sent globally among businesses and Microsoft™ Teams™ reported a usage increase of more than 330 percent.

Do these numbers startle anyone else?

The eDiscovery industry has responded with many unique ways of reviewing these data types. However, some of the most common and accessible solutions have major limitations.

Challenges with Conventional Chat Data Review Solutions

Before native chat review solutions became available, eDiscovery professionals relied on the existing export capabilities of these enterprise communication tools. Most communication tools come with discovery-friendly capabilities, allowing their data to be exported into semi-reviewable formats.

Excel export chat data

chat data PST format

Two common review approaches were:

  1. A conversation or channel could be reviewed as an excel file where each message, participant, and sent time stamp was represented as new rows. Each excels contained the entire conversation in one sheet.
  2. Second, each message could be displayed as its own individual record in the review tool OR in an email client such as Outlook. In this technique, reviewers need to click on a new record to read the next message in the conversation. Inline images and file transfers are not supported as standard attachments would be.

There are four primary reasons why these solutions are not ideal:

  • Early analysis of a single chat channel within excel

Consider a business context where the group chat has 10-15 people and has been an active channel for 2+ years.  When did the participants join?  The excel export will not notate this.  It also will not provide any data visualizations or interactive displays for reviewers to easily engage with the chat.  The filtering and searching features in excel can absolutely provide a lot of this information if the reviewer 1) takes the time to set up their excel view with filters or possibly prepares a graph of the timestamps and 2) understands excel well enough to dig this information up themselves.  Some service providers will assist with this redesign for a fee.

  • There is rarely any chat metadata from the source system (Teams/Slack)

The problem here is that every eDiscovery professional uses metadata to strategize their review.  Key metadata like the start and end date of the conversation isn’t easily retrievable or searchable when using either of these solutions for chat data review.

  • Emoticons and Images are not supported (non-text content)

Excel, for example, doesn’t display emoticons, screenshots, or file transfers within collaboration systems. Review teams must attempt to correlate chat “attachments” to where they were supposed to appear within the conversation.

  • Courts and regulatory bodies have rejected productions in these non-native formats

In Charter Communications v. Optymyze, when the plaintiff requested Microsoft Teams’ data, the defendant produced the data in a format where each chat message was an individual document, separate and apart from any other messages the users exchanged in that conversation. Later, the court granted the plaintiff’s motion and ordered the defendant to produce the Teams messages in native format.

Understanding the Concept of Native Chat Review

Chat data is fundamentally different than traditional review data types requiring additional attention during Discovery. People tend to communicate differently when in chat. The language is more casual, and expressions such as emoticons and gifs are often used.  However, each of these new types of communication can still indicate intent or action.  Often in e-discovery, you may be looking for affirmations or instructions in emails.  A smoking gun email from the boss saying “Yes, go for it” or “project approved” or “buy it now.”  In a chat, this may simply be a reaction to a message or a THUMBS UP emoticon!  Small nuances like whether or review strategy support emoticons start to become key case strategy.  If that support is missing, you may miss key evidence.

In the same context, a native review system that shows attorneys when file transfers occurred and can one-click display those attachments play a significant role in finding the context of the lengthy channel.  Context plays a crucial role when dealing with chat data.

Native Chat Review solution aims to address these challenges by delivering a review experience similar to the data types native platform.

How does Native Chat Review Solution Solve Chat Data Review Challenges?

Based on the challenges we’ve heard about so far here is what to expect when using any Native Chat Review solution for eDiscovery.

The solution should:

Provide advanced filtering, searching, and early analysis to reviewers

A native chat review ensures that reviewers can perform quick searches and filter with simple clicks, but also it should display key details by default.  Participant names, number of messages per participant, chat timelines, etc. should all be visualized on the page by default. Solutions following this concept help reduce cost and time by helping reviewers target critical content faster.

Offer quick insights

Compatibility with predictive coding TAR models and unsupervised machine learning models like PII, Key phrases, and Clustering helps eDiscovery professionals plan their case strategy. Native Chat data review captures metadata appropriately to provide more context to the reviewer. Timeline charts give access to data visualization with chat specifics such as metadata.

Support Emoticons and Multimedia Messages

Chat data often comes with multimedia messages (photos, videos, etc.) and emoticons.  Understanding the importance of emoticons to intent and action this is a show stopper.  The system must support emoticons to be considered effective.  Furthermore, native chat review solutions should also allow you to view images, screenshots, and shared files within the viewer or in no more than one click.

Offer the experience of the custodians’ native chat view

With existing chat review options, it was challenging to find the context of the entire conversation, especially when the chat is a group chat or too prolonged. Native chat data review features provide the exact feel of the custodians’ native chat platform. It helps reviewers to quickly go through the chat enabling the visual context of the entire conversation flow.

Native Chat Review in Knovos Discovery

Concluding thoughts: Native Chat Review solution untangles the conversation’s context

More and more companies are joining the hybrid work model giving rise to collaboration tools usage for professional communications. Therefore, eDiscovery professionals should expect a corresponding hike in relevant chat data during their Discovery activities. Native chat data review solutions can help mitigate the impact this emergent data type has on both eDiscovery and in-house data compliance teams.

Knovos dedicates 75% of our resources to research and development to empower eDiscovery professionals to deal with the Discovery changes and deliver the best results. Talk to our experts to learn about our Native Chat Review solution and find out how you can reduce your eDiscovery costs and increase the review accuracy.