Blog Viewer

Practical Tips for Gathering and Analyzing Messaging Data - Alex Chatzistamatis

By Alex Chatzistamatis posted 04-08-2022 12:15


Please enjoy this blog post authored by Alex Chatzistamatis, Principal Solutions Consultant, Nuix.

In the not-so-distant past, short message format data like texts or chats were not typically a requirement for most organizations to make discoverable.  The pandemic certainly redefined many areas of the legal industry, and I’ve certainly seen a huge uptick in organizations needing solutions to address messaging data. Microsoft 365 Teams alone had nearly doubled usage between March and April 2020 and over 70% of Fortune 500 companies subscribed to it during those 12 months, due to COVID (source below).  Organizations were quickly forced to start figuring out how to deal with data from Microsoft Teams, Slack, Zoom, Yammer and other social media platforms.

If you’re the kind of organization that is dealing with this struggle in house, there a few ways to gather and analyze the data.  In terms of gathering the data, you have a few options depending on your desired workflow: 

  1. Direct source download (native) – this involves going directly to Microsoft, Slack, Zoom, etc and creating an export using native tools. Often, the data output is not what you’d expect.  For example, a Slack export in JSON format is hard to deal with.  Manually looking at these kinds of messages like the way people historically looked at emails is very tedious.

  2. Central platform (multi API connectors) – there are several commercial options which allow you to effectively do a direct source download, but through a central platform, so that way you don’t have to make too many manual hops.

  3. Direct connectors (API integration) – commercial software solutions like Nuix provide direct connectors to cloud sources like M365 and Slack which allow you to skip a traditional “data download” and load the data directly into your processing tool.  

When it comes to analyzing the message data, you have a few options depending on how you obtained the idea:

  1. If you are not going through a connector, you’ll want to make sure your data processing tool supports the file types from the different chat/messaging platforms. Not only is the file type important, but you’ll want to determine if the tool can parse the content and metadata properly.

  2. If you are going through a connector, the same things I just mentioned above still apply, but you’ll also want to make sure you fully understand and are comfortable with how the process works, available metadata, support for attachments, reactions, etc. The biggest things to consider are how these tool deals with authentication, bandwidth throttling, transient connectivity, etc.

Lastly, you will want to identify a solution which helps make the analysis easy and presents the data in a way that mimics the original chat/message platform.  Who wants to be look at 1 message at a time?  Not me. I’d rather collect, process, filter and have the context of the conversations to make better informed decisions for any downstream work.