The last few years have brought more technology than ever into our lives. Video meetings have replaced in-person meetings, everyone’s favorite 90’s-era chatrooms have been rebooted as the new digital collaboration space, and even office post-it note reminders have gone digital. With all that digital static buzzing around our heads, there’s more noise than ever to cull when it comes to eDiscovery.
In business settings, one of the most popular new technologies is Slack.
Used by over 70% of the Fortune 500 and boasting over 12 million daily active users, this platform is renowned for simplifying and streamlining virtual communications. However, it can be a nightmare for litigation and internal investigations—unless you’re armed with the right tools and processes to navigate chat discovery.
That’s exactly what Kyle Kelly, Renee Bea, and David Slarskey discussed in a recent interview with Logikcull: the best ways to mitigate discovery risks posed by chat and how to handle thorny evidentiary issues around Slack data.
Here’s a brief overview of the interviewee’s expertise with Slack discovery:
As the Senior Technical Program Manager at Coinbase, Kelly is a veteran in the Slack eDiscovery arena. He handles enterprise-level data collections across the entire EDRM with a commitment to efficiency, security, and defensibility. In this interview, he shared how he single-handedly tackled a massive Slack discovery project.
Renee Bea, partner at Slarskey LLC, is a seasoned trial lawyer with more than 15 years of experience representing clients in litigation, trial arbitration, and appellate proceedings. Her practice focuses on complex business disputes, fraud, financial services, and intellectual property, with the occasional deep dive into the dark depths of Slack discovery.
A litigator, trial attorney, and also partner at Slarskey LLC, David Slarskey has broad experience in all aspects of dispute resolutions. From pre-complaint negotiations, all the way through trials and appeals, he’s done it all. He has also had some outstanding wins when leveraging technology to deal with Slack data in previous matters. Check out his recent case study with Logikcull.
Read on for a deep dive into their practical guidance on how to navigate Slack discovery end to end, or get it straight from the sources by watching the full interview below:
This blog covers the first half of the conversation with these Slack discovery experts. We’ll share the second half of the interview on Logikcull’s blog very soon. Stay tuned!
Logikcull: Kyle, can you tell us how you're using Slack at Coinbase?
Kyle Kelly: Slack is a ubiquitous tool that is utilized throughout the company for all different types of things, similar to social media. A lot happens there, so we have to be mindful of what we're actually sharing in Slack. We need to have constant reminders of good practices, as you want good corporate hygiene and culture on these types of messaging platforms.
Everything happens on Slack, so we want to make sure that we are not only being mindful about it from a discovery perspective but also making sure that it's being used effectively.
From a corporate productive tool standpoint, we want to make sure that we're being good about safeguards and certain channels, because there's a lot of sensitive information being shared.
Logikcull: Coinbase has been a Slack user for a really long time now. Can you tell us a little bit about how your policies towards Slack and particularly your retention policies have evolved over the period that you've been using it?
Kyle Kelly: A lot has happened in just a little over two years. There was a time when the Slack retention policies were very minuscule and not really well-thought-out. Over time, we thought intentionally about it.
Currently, we have a 90-day retention policy across all of Slack. As you think about your retention policies internally, you need to ensure that you're mindful about how it's going to impact the company.
Additionally, ensure that you're going to have a vetted third party outside counsel to approve that retention policy, as it may be challenged.
Slack also allows you to be really granular with your retention settings. Let's say that you have files that have an impact for longer than you want the messages to exist. You have the flexibility to decide whether or not to change your retention policy based upon your messages or your files. This has helped us navigate through our regulatory requirements and the maturity of the company as a whole.
Logikcull: Renee, when you come into the picture as the counsel advising someone like Kyle, it might be at the planning stage, where you're putting together those retention and information governance policies, or it might be when things have already happened. When you approach an engagement where Slack data is at issue, what advice do you give your clients, wherever they might be in that process?
Renee Bea: I think the number one thing is that the client needs to preserve any Slack data available right away. As soon as that threat of litigation is there, or frankly, in an investigatory situation, you just want to have that data available. I would not rely on the tool to do that preservation work for you.
Slack is an innovative tool that is constantly changing. Its capabilities are constantly changing, as Kyle alluded to in their own evolutionary use case. But until you can nail down what your particular settings are and what might be applied to the specific data that is at issue for preservation, just grabbing that data and making sure it's available to you later is the key first step.
Logikcull: Kyle, you have handled some mind-boggling Slack discovery challenges and you've done it all in-house. Can you tell us about one of the most daunting ones and how you got through it?
Kyle Kelly: We got a regulatory inquiry for more or less every Slack message that was available. At this time, we hadn't thought through our retention policies too well.
So we had a lot of Slack to go through, and there was not a good way to get a handle on that. We had some complex search terms that we could apply through email and Drive and other data types, but to apply that to Slack was going to be difficult.
I had to find a way to ingest all that data and make sense of it. The alternative was to produce all of it and, honestly, if I had done that, I probably wouldn't have a job right now.
Going back to the JSON files which the Slack corporate export provides, I was able to ingest that in a meaningful way within Logikcull.
Fortunately, I was able to utilize the same Boolean logic-type of searching that is able to utilize in both email and Drive and significantly reduce the number of records. A ballpark number might be six to eight million records of Slack. Applying these search terms that were provided and agreed upon with outside counsel and the regulators reduced that down to roughly 32,000 messages, which is a dramatic difference.
Applying the search terms that were provided and agreed upon with outside counsel and the regulators, we were able to go from almost eight million to roughly 32,000 Slack messages.
Then, I had to pass it on to a trusted third party. I can't imagine the bill I would have received from that third-party hosting provider if I had sent them eight million Slack records.
By being able to reduce that number significantly and only producing the relevant Slacks for the scope that was agreed upon, we kept the focus on the points at hand. We’re able to stay on course for the retention requirements we had to do in that case, and we'll have to do in cases in the future.
That was a heavy lift, and I was very impressed by taking all that data and making sense of it in Logikcull as opposed to a custom script that I had someone write many moons ago in Google Docs.
Logikcull: That’s definitely a testament to the ingenuity of your team and the fact that if you have the right platform, you can turn a mess into something that's much more manageable. Going from eight million messages down to 32,000 is over a 99.9 cull rate, which is really amazing.
David, the technical challenges of Slack are unique, but the type of conversations that people are having in Slack can also be really probative. In your experience, what’s so revealing about Slack conversations?
David Slarskey: As litigators or trial attorneys our job is to manufacture narratives or to bring out the narratives. And we're looking for whatever evidence or snippets available to create that narrative.
When I came into the practice 20 years ago, email was overpowering everyone and we were looking at emails.
From there it moved to—at least in New York with financial services litigation—Bloomberg messaging. You'll never find better evidence than you find on the Bloomberg terminals of bond traders. All of a sudden, you had a whole other source of messaging.
Now, what's interesting about Slack is that it's multidimensional, and it's different in the sense that hard paper notes are very linear.
Slack creates a much more multifaceted and complex picture of what's going on. It gives an entire community's perspective, to the extent that with Slack, you have an open channel, you're collecting a marketplace of information in a way that we really haven't been able to in the past.
You were never able to get a record of the conference where that took place, where people met and conferred. Now we have that in writing, and it's not just the inter-communication; it fragments out into other source documents.
Slack creates a much more multifaceted and complex picture of what's going on. It gives an entire community's perspective.
There's more meaning being created in Slack channels by the jokes people tell. When people make jokes about somebody being incompetent or something going completely wrong, that's the kind of evidence that 20 years ago, ten years ago, five years ago, was impossible to capture because there was just no record of it.
Logikcull: Renee, when you're looking through Slack data, how do you craft your approach? Are there any tips or tricks that you can share?
Renee Bea: I think as attorneys, we need to understand how your client uses the tool, and how end-users actually use the tool. Because Kyle can do his very best and recommend policies and rollout company guidelines, but as David mentioned, Slack is this multidimensional space where there's a lot of business and personal interaction woven into the platform.
So, step one is getting a real understanding of how your organization or your client's organization uses Slack and what capabilities they do have available to them.
And then you need to understand the integrations. Slack is advancing almost daily on its integrations with other platforms, and you need to know which of those integrations your client uses because inevitably, Slack is going to point to other sources of data.
In terms of the volume of data and files that Slack can generate, I think one of the most useful things in crafting your approach is getting your hands on the data and getting to know it. And that is something that we've been able to achieve using Logikcull to really narrow down and focus our review, investigation, or analysis to understand what we have.
The ability to ingest that data and look at it and begin to get a feel for what is happening in certain channels and for what particular custodians or witnesses are doing with their Slack experience is critical to crafting any approach in litigation.
That is something you can do in Logikcull based on their integration with Slack works and how you pull the JSON files in, and how those files are broken down. It's very user-friendly. Any time you can get your hands on that data, it's amazing how much more focused you can really be in terms of your view of the evidence, your investigation. And then you are talking about not three million files, but, like in one case that I had, 1,500, which is an incredible difference.
One of the most useful things in crafting your approach is getting your hands on Slack data and getting to know it. And that is something that we've been able to achieve using Logikcull to narrow down and focus our review, investigation, or analysis to understand what we have.
Logikcull: Renee, we've talked a lot about Slack including everything, but you run into instances where actually there's a smaller amount of information in Slack than you might expect. What goes off in your mind when you see something like that?
Renee: It usually prompts more questions because, in my experience working with Slack internally and also seeing how my clients use it, I know it is a very candid space. It goes beyond the water cooler because it blends that personal interaction with the professional one.
So if I'm really not finding information, that would always prompt me to ask questions about how I'm thinking about my search terms, if I'm looking in the right channels and if I'm looking at persons of interest or witnesses that are important to the case and how they use Slack.
You can name a channel anything you want. Some people are going to use the name project and make it really easy on us lawyers, and others are going to have their acronyms or their internal side jokes. You should get that question prompted early in the case through understanding the Slack experience or you might find out later how witnesses were actually talking about something with a code name instead of what you may have assumed.
If Kelly’s mention of “eight million Slack files” shocked you, you’re not alone. At Logikcull, we’ve seen unique docs uploaded per Slack project increase by more than 380% since 2020. This can be an enormous headache for legal teams without the right tools, which is why Logikcull is releasing a Slack Integration and Chat Filters.
As Renee pointed out, the most important part of conducting discovery on Slack data is ensuring that you preserve the key conversations and channels as quickly as you can. From there, you need to make sense of your chat data and get to the most important information quickly. But with the traditional data exporting, ingesting, and filtering methods, this process can be challenging.
However, Logikcull’s new Slack integration and chat filters put your Slack data just one click away, and allow you to find the signal hidden in noisy chat conversations faster than ever before.