Without Contextual Messages, You May Be Missing the Context of the Conversation: eDiscovery Best Practices

This week’s post for IPRO’s blog discusses another topic from last week’s ILTACON 2021 conference – the session that I moderated – the importance of contextual messages in eDiscovery today.

That session was titled Better Together? E-discovery with Teams and Other Collaboration Platforms and we discussed discovery considerations associated with collaboration apps like Slack, Teams and Zoom.  We also discussed a few cases tied to the topic, including Sandoz, Inc. v. United Therapeutics Corp., which discussed production of contextual text messages.

Contextual text messages are messages that are part of a conversation that don’t hit on search terms for relevance but are still needed to fully understand the conversation between the parties.  Because each message is stored individually, you can’t get the full understanding of the conversation that took place without including those contextual messages.

So, why did we discuss a case involving contextual text messages in an educational session about discovery of collaboration platforms?  And why don’t contextual messages apply to emails?  You can find out on IPRO’s blog here. 😉  It’s just one more click!

Speaking of the session, I want to thank the panelists I worked with for a great session with lots of great insights – Damon Goduto with Lineal Services, Rose Jones with King & Spalding LLP, Jack Thompson with Sanofi and Martin Tully with Redgrave LLC – as well as Gordon Moffat of Pillsbury Law for doing a great job coordinating the session for us!

So, what do you think?  How does your organization handle contextual messages in discovery?  Please share any comments you might have or if you’d like to know more about a particular topic.

Disclosure: IPRO is an Educational Partner and sponsor of eDiscovery Today


Disclaimer: The views represented herein are exclusively the views of the author, and do not necessarily represent the views held by my employer, my partners or my clients. eDiscovery Today is made available solely for educational purposes to provide general information about general eDiscovery principles and not to provide specific legal advice applicable to any particular circumstance. eDiscovery Today should not be used as a substitute for competent legal advice from a lawyer you have retained and who has agreed to represent you.


  1. “Contextual text messages are messages that are part of a conversation that don’t hit on search terms for relevance but are still needed to fully understand the conversation between the parties. Because each message is stored individually, you can’t get the full understanding of the conversation that took place without including those contextual messages.”

    This issue is fascinating, because it inadvertently lays bare the conceit at the heart of keyword searching and culling: There are documents that can help one contextualize a pattern of activity — i.e. that are responsive because they are about the issue at hand — that themselves don’t hit on search terms that a lawyer thinks are most indicative of responsiveness.

    There is a difference between how we do something (indexing messages individually, only reviewing messages that hit keywords) from what we’re suppose to do (produce responsive information). And that same difference applies to documents in general (not just emails, but also Word and Excel files, too): Atomic units of indexing that don’t hit any lawyer keywords may still be responsive.

  2. “For emails, we never have to worry about this issue because each email is a snapshot of the conversation up to that point – search terms that hit on any part of the conversation will retrieve the entire conversation (whether all of it is contextual or not). The issue with email is deduplication, not context.”

    Oh, and not necessarily true for email. For a number of reasons:

    (1) One’s email client has to be configured to automatically quote/include the text of the previous message if one hits “reply”. Today most clients are configured that way as a default (which was not the case 20 years ago), but that doesn’t mean every person at a company has left that default in place. Some may have switched it off, in which case any search term hit will _not_ retrieve the entire conversation.

    (2) Even if the “reply” button automatically includes previous message text, that does not mean that the user has left that text there. Speaking for my own habits, when email threads get too long, I regularly cut out all but the few sentences that I’m replying to. I will often reply inline, cutting up chunks of the previous email and deleting what I am not addressing in my current response. Even though I am not addressing some deleted chunk, though, doesn’t mean it’s not responsive to something. And if the keywords don’t hit on that chunk that I’ve deleted, then “end of thread” email will not work.

    (3) I also occasionally don’t hit “reply” to an email, and instead start a new message/thread. This is true if I’m taking the conversation from a large group down to a selected few within that group. It’s also true if a thread has gotten too long, and I just want to start fresh.

    (4) Finally, I’ll sometimes move conversations from short messages/Teams to email, and back again. My conversations with people don’t stick to one medium. An email to me might be met with a short message response, or vice versa. As such, email is part of that context.

    (5) Hell, I should be writing a blog, myself 😉

  3. Addressing your points, Dr. J:

    1) True, and many people don’t realize that the default can be changed to not include the original text of emails when responding to them. I think it was more like 25-30 years ago that the default was NOT to do that, but it is to include them now and certainly most people use the default setting (as evidenced by all of the eDiscovery solutions employing email thread identification). While I neglected to mention that in the article (because I was primarily focused on contextual messages for text and collaboration apps), we did talk about that in the ILTACON session.

    4) That’s common these days and I addressed that in a post here (https://ediscoverytoday.com/2021/06/23/addressing-bifurcated-conversations-in-discovery-ediscovery-trends/) where I talked about how conversations today are often bifurcated, trifurcated or even quadfurcated (yes, that’s a real word, I looked it up) across email, text, collaboration app and more. We also talked about it in the ILTACON session. That’s why you have to collect from all of these sources today to get a complete picture of a true “conversation”.

    2) and 3) Perhaps you do, Dr. J., but most people I know don’t do that — they let the email threads continue to grow and grow. I guess that’s part of your charm! 😉 Regardless, I think, as a general rule, most conversations within email can be captured within one to a handful of emails at most.

    5) Hell, I’d read it! 😉

    • (1) I suspect you’re correct about the timeline.. probably was longer ago than I remember. But the option is still there, today. Tongue-in-cheek note to companies with a lot of litigation: Turn off auto-quoting in your email clients, amirite?

      (4) Ah yes, of course. Now that you link to it, I actually remember reading it. Apologies for pointing out what I already knew that you already knew.

      (2) & (3) I sense a market opportunity here: An email client in which you compose your message as usual, but then the client breaks that message up into thousands of individual emails, each comprised of a single word. On the receiving end, the receiver’s client then reassembles the individual messages in order. But when doing discovery, any “4 corners” email that has a keyword hit will only have that individual keyword in it. Is that not a logical response to current discovery practices? (Note: Clearly I’m again making a tongue-in-cheek Jonathan Swift-ian “modest proposal”, and am not actually serious. But as with all satire, it illuminates underlying weaknesses of the system.)

      (5) 🙂

  4. Or, for security reasons, the receiver’s client could reassemble the individual messages randomly. Of course, the email won’t have any evidentiary value anymore, but at least it will be secure! 😉

Leave a Reply