How an Enterprise Migration and an Extended Discovery Stay Created the Perfect Storm for Sanctions

A legal hold system that “works” can still wipe out the evidence you need, and you may not realize it until the sanctions motion lands. We unpack a major Rule 37(e) preservation decision from DDC in United States ex rel. Staggers v. Medtronic that every litigator, legal operations leader, and discovery counsel should study, especially if your company is migrating email or sitting under a long discovery stay. 

We walk through the timeline problem that drives the whole outcome: a case served in 2016, discovery stayed for years, and an enterprise-scale Enterprise Vault to Microsoft Exchange migration happening in the background under a two-year retention policy. You’ll hear exactly how “dropped holds” and normal retention can quietly delete older ESI while dashboards still report that mailboxes are migrated. Then we split the analysis the way the court does: system-level migration steps can be reasonable under the Sedona Principles, while custodian-by-custodian legal hold timing can still be unreasonable and costly. 

We also dig into the practical litigation lessons: why deposition testimony is not a substitute for missing ESI, why “replacement” has to be substantially complete, why intent to deprive remains a high bar, and why you must prove prejudice with specific documents rather than generic argument. Finally, we highlight the real financial exposure that too many teams underestimate: multi-bucket fee awards for spoliation investigation, motion practice, and even fees on fees. 

If you’re responsible for eDiscovery, legal holds, retention policy risk, or managing outside counsel, listen through and pressure-test your preservation plan against this checklist. Subscribe to Meet and Confer, share the episode with a colleague, and leave a review with your biggest takeaway.

Thank you for tuning in to Meet and Confer with Kelly Twigger. If you found today’s discussion helpful, don’t forget to subscribe, rate, and leave a review wherever you get your podcasts. For more insights and resources on creating cost-effective discovery strategies leveraging ESI, visit Minerva26 and explore our practical tools, case law library, and on-demand education from the Academy.  

Transcript

Why Autopilot Holds Fail

Kelly Twigger 0:31

Welcome. Our episode today is a critical one for legal operations and for the litigators that represents corporate parties or any organization. It highlights the need for the ultimate form of documentation and cooperation between in-house and outside counsel to effectively manage legal holds for active litigation during a system-wide migration of data. It's not the first case to address this issue. And that's perhaps even more important. There is precedent for this situation that we're going to discuss today. Today's episode reiterates several of our themes here on the case of the week: that the timeline is critical in terms of the duty to preserve. Documentation of your process is critical and key, as you're going to see here. And the need to have specific factual evidence to back your arguments on discovery motions is going to continue to come up. If you take one thing away from today's episode, let it be this: the duty to preserve is not satisfied by a legal hold system that works on autopilot. Somebody has to be reading the complaint, reading the discovery requests, and asking week after week, month after month, whether the right people are on the hold. Because in a world where retention policies delete data on a clock, a custodian that you forget today can be a sanctions motion tomorrow. And that's particularly important where discovery is stayed for a period of time. A discovery stay is a danger zone for preservation failures. Welcome back to the case of the week segment of the Meet and Confer podcast. Today we're talking about the latest decision in United States XREL staggers versus Medtronic, a Rule 37E decision from the district of the District of Columbia, entered on April 6th, 2026 by United States Magistrate Judge Michael Harvey. I'm your host, Kelly Twigger, the CEO of Minerva 26 and a discovery strategist and practicing litigator for almost 30 years. I started the Case of the Week series in 2020 to emphasize the role that courts are playing in telling everyone who touches electronic discovery what they need to be doing. This series is designed to teach anyone who touches the discovery process, litigators, litigation support folks, legal operations folks, and counsel generally how to identify potential issues that need to be paid attention to in your matters. It's issue spotting. It's telling you what's coming up day after day in discovery so that when you come across these issues for your clients, you are prepared for them. Today's case is one I want every litigator to pay attention to because it does something rare. It gives us a side-by-side comparison of preservation conduct that the court called reasonable and preservation conduct from the same company in the same case that the court said was not reasonable. Same litigation, same defendant, two very different outcomes. Let's dive in. This is a false claims act uh key TAM case. The realtors that brought the allegations are two former Medtronic sales representatives who marketed and sold Interstem, a surgically implanted device for incontinence. Their theory, which survived a motion to dismiss in 2019, is that Medtronic Salesforce and a unit called Support Link inserted themselves into the doctor-patient relationship, assessing patient symptoms, documenting them, encouraging implantation of the device in a way that caused physicians to submit false Medicare claims. Now, the realtors filed in March 2015 and served Medtronic in 2016, August of that year. And then, and this timing matters more than almost anything else in the case, two months after Medtronic was served, the court stayed discovery. That stay remained effectively in place for more than five years until December 2021, by which point Medtronic had completed an enterprise level migration of its email system. As always, the timeline is critical in discovery issues, and here it is almost as positive. By the time Medtronic ran wide-ranging searches for relevant ESI through its email system, the legacy system holding most of the ESI was gone. And remember the timing here. This was 2016 and prior. So email is the primary form of communication that's relevant for discovery. This is well before we started having text messages, chat, Slack, Teams, all of that data that is really complicated discovery today. Here's what makes this case unusual. While litigation was paused, Medtronic was running a massive enterprise project that had nothing to do with the case. It was just a system-wide migration. The company had used Enterprise Vault since 2007 to archive emails. That was a very common system that a lot of corporations put in place in that around that time in the early 2000s. Enterprise Vault retained emails indefinitely. And by 2013, it was holding more than a billion email messages at Medtronic. So in 2013, Medtronic adopted a two-year email retention policy. And in May 2015, it started migrating its archive from Enterprise Vault to Microsoft Exchange Server 2013. The migration finished four years later in 2017. Enterprise Vault was completely decommissioned in February of 2021. The migration mechanics matter here. For a custodian who was not on legal hold, only the most recent two years of email got copied from Vault to Exchange and then dropped on a two-year rolling retention cycle. Anything older stayed in Enterprise Vault until it was decommissioned and then it was gone. For a custodian who was on legal hold, the rule was different. An active directory flag was placed on the account on Microsoft Exchange, and the account was supposed to ensure that all data, not just two years, got copied to Exchange and the retention policy was suspended while the hold was active. Once the hold was released, the two-year policy kicked back in. Here's the trap that Medtronic kept tripping over during this process. Picture a health economics manager, H E M as it's referred to in this decision, and let's call her Sarah. She left the company in 2014. Suppose that Sarah was on a hold for an unrelated 2010 matter. So her full mailbox was migrated to Exchange. The 2010 hold was released in 2017. The retention policy starts running then. By 2019, anything older than 2017 is deleted from exchange. In 2022, this case that was filed in 2016 adds her to the hold, but Medtronic system reports her mailbox was migrated, so no look back to Enterprise Vault is triggered. And Enterprise Vault had been decommissioned for over a year. So her data from 2016 and prior to that is gone with no alarm having fired anywhere along the way. A two-year retention policy plus a dropped hold equals data destruction every time, invisibly, while your system reports that everything is fine. Now, let's plug that fact pattern into the actual case. Medtronic placed its initial hold in September 2016 on 303 custodians, which were mostly Salesforce people. That's a lot of custodians. Over six years, it added 369 additional custodians for a total number of custodians of 672. That is an enormous number of custodians to manage. When Medtronic finally started running wide-ranging searches in August 2021, it discovered that 20 of those 672 had partial or total data loss due to the migration. The realtors moved for sanctions and identified 14 custodians who had said they should have been on the hold earlier. Those people consisted of health economics managers, support link personnel, Cindy Kent, who was a marketing director, and later the VPGM of the unit responsible for Interstem, and Annette Middlemark, who was a medical education manager training physicians on Interstem. One production note here before the analysis. This is a very long, factually detailed opinion, over 20 pages with custodian by custodian analysis. If you advise clients on email migrations or consult on them, you should read this opinion in full and appreciate the scale of evidence that the parties had to compile. Multiple sworn declarations, a 116-page statement of material facts, over 300 exhibits and depositions just on the spoliation issues. This is not a 30-page motion that you draft over a weekend. It's a sanctions fight that cost the defending side tens or potentially hundreds of thousands of dollars on its own. And as we'll see in the end, the losing side ended up paying the other side spoliation investigation costs. All right, those are the facts. Let's talk about the court's analysis. Rule 37E, which is the motion for sanctions that the realtors filed, has five threshold requirements. The loss material has to be ESI. There has to be a duty to preserve. The relevant ESI should have been preserved. The loss has to be because the parties failed to take reasonable steps, and the ESI cannot be restored or replaced through additional discovery. Only if all five are met does the court reach curative measures under Section E1 or sanctions under E2. And E2 requires, as we know, a finding of intent to deprive, which is a very high bar. And here's the structure that I want each of you to internalize as you're listening to this. The court does not evaluate preservation as one yes or no question. It evaluates preservation here as two separate things. First, was the migration process, the system level decisions about how to move a billion email messages reasonable. And second, were the whole timing decisions for each individual custodian reasonable? Different questions, different answers. And the court is explicit quote, the reasonableness of Medtronics migration processes and decommissioning of enterprise vault does not end the inquiry. Rather, the undersigned must still assess whether the company reasonably instituted legal holds over key custodians. Close quote. The court found that the migration handling was reasonable, relying on the Sedona principles, third edition, in which the district of the District of Columbia and many federal courts treat as authoritative guidance. The relevant principle, quote, perfection in preserving all relevant electronically stored information is often impossible. Close quote. This goes back to uh Judge Shinland's argument in the Zubalay cases back in the early 2000s. Perfection is not the goal. Reasonable steps are required, not heroic ones. So, what did Medtronic do that the court called reasonable? It copied data for tens of thousands of custodians on legal holds from enterprise vault to exchange. It built a look back process. When it realized that former employee account deletion was creating gaps, it built a reconstruction process. It was all imperfect, but it was a plan executed in good faith by people with the right skills. Now the realtors argued that decommissioning Enterprise Vault in 2021 was unreasonable given the pending litigation. The court flat out rejected that argument by distinguishing two cases that the realtors relied on. In Ray Peters, in which a bank decommissioned Lotus notes while subject to a preliminary injunction, but made no real effort to preserve data before destroying backup tapes. And in Estate of Moreno versus correctional health care, where the defendant sent one email to outside counsel, put a hold on one person, deleted everything else, and the CIO testified that the retention policy was designed so that, quote, bad emails would be destroyed. The court called those efforts lackluster. Medtronic's approach, which was multi-pronged with lookbacks and account reconstruction, was imperfect, but not lackluster. So, and so it distinguished from those cases argued by the plaintiffs. Here's the line to file away for when you need it. And this can apply to any system, not just email. This is a quote from the court. The mere fact that relevant data was lost does not mean that the party holding such data failed to take reasonable steps to preserve it. Close quote. Keep in mind that with the sources of ESI that we are dealing with today, text messages, chat, instant messages, we are increasingly likely to have spoliation issues where this reasoning can be applied. So file this one away for something to be thinking about. This is the issue spotting functionality of the case of the week that I want you to be storing. Now, thinking about this case and about the migration that happened here, I want to just tell you a quick story from my own experience. I managed the exact same transition at issue here from enterprise vault to exchange for a client in 2009 through 2012. And I can tell you personally, if you can't tell from the facts of this case with a billion uh email messages and tens of thousands of custodians, that it is a complete and total nightmare. The fundamental problem was that data in Enterprise Vault could only come out through a front-end search. So there was no clean way to export data from the back end. So during migration, you had to manually pull the relevant date range for each custodian on each matter, even when the date ranges overlapped a lot. And the decision that we made with the client was to manage the preserved data outside of the new exchange environment entirely, separate storage indexed by custodian and by matter. That way, when a hold was released, we could delete the data tied to the matter without destroying data still needed for an unrelated ongoing matter on the same custodian. Now that process, which sounds to me like just based on the volumes of data and number of custodians, that was probably about a third of the size of what Medtronic was dealing with, took more than six years to finally release all of the old data. Six years. If we had trusted the new system to remember the old relationships, we would likely have ended up where Medtronic ended up in this case. Today, that kind of migration could be handled much more efficiently using AI. But in the 2015 to 2021 window that Medtronic is dealing with here, that was not an option. Remember that we had spreadsheets. We didn't have a lot of the technology that exists today to allow us to filter and sort data effectively to be able to provide what was responsive when we had sufficient, very structured preservation obligations on multiple matters. So the lesson to learn here is if you're managing preservation during a system migration, you cannot trust the new system to remember what the old system was holding. Build the documentation yourself outside the system, tie it to specific matters, have documentation that tells you exactly what is in the new system. All right, let's talk about now switching from the actual system migration to the timing of legal holds that were placed on specific custodians by Medtronic. This is the part that earned Medtronic uh curative measures and, or that earned the realtors, I should say, curative measures and a fee award. Even within a reasonable migration system, the court still asked whether each specific custodian who lost data was put on hold at the right time. And the court here went through a custodian-by-custodian inquiry, which does make this opinion more lengthy. Of the 14 custodians, the realtors challenged the court rejected five and accepted nine. The five that it rejected were health economics managers that were added in mid-2019 within a reasonable time frame after the realtors served revised RFPs asking about Medicare reimbursement. So, in essence, Medtronic added new custodians when they received discovery requests that implicated them. But two other HEMs, Mike Neal and William Paul Harris, should have been added in May 2019 at the same time, but they weren't. Neal got added in August of 2022, but William Paul Harris was never added. And wait till you hear this. Medtronic's explanation was that there was confusion between William Paul Harris, who was the named custodian, and a different employee, Paul T. Harris, who was a sales manager who was on the hold. The court didn't buy it and found that, quote, a sophisticated enterprise like Medtronic, with tens of thousands of employees and tens of thousands on legal holds, should have processes in place to prevent confusion about employees with similar names, perhaps especially when the names are common like Harris, close quote. Now that's pretty brutal, and that's a pretty large system-wide process to put in place. But that's the bar. Here's what that process actually has to look like in practice following this court's rationale. First, your legal hold database needs to capture capture more than name. At a minimum, it needs to capture employee ID, job title, department, and email address. Second, every time a new name is added, the administrator will have to confirm the unique identifier against the source that prompted the hold, the complaint, the RFP, the witness list, not against memory. Third, build a same surname flag. When you add a Harris or a Smith or other common names in your enterprise, the system surfaces every other employee with that surname and forces a deliberate choice. If you can't produce those three steps in writing, you cannot tell a court that you had a process that reasonably addresses the problem that surfaced here for Medtronic. The other group that was at issue were five support link custodians. And this is where the court's reasoning becomes key to pay attention to. Medtronic argued that it didn't know the support link personnel were key custodians until the realtors served a second set of RFPs in May 2022. In response, the court pulled the realtor's October 2016 RFPs, six years earlier, off the shelf and pointed out that the plaintiffs had specifically asked for documents relating to, quote, instructions, advice, andor information provided to support link representatives and all documents sufficient to identify Medtronic support link representatives. The May 29 RFPs included the same request, and so did the May 2022. And the court's question was this one, and it's a pretty good one. Why was the same request sufficient to put Medtronic on notice in 2022, but not in 2016? Medtronic offered no answer to that in its response. And the court held that support link custodian should have been on hold around the time of the initial September 2016 holds. The same problem occurred with custodians Kent and Middlemark. The 2016 complaint and RFPs should have put Medtronic on notice. Now that's an argument that you can take and use the next time opposing counsel says, we didn't realize until later. Pull the original RFPs that you think should have put counsel on hold. If the request was the same, the notice was the same, and you can leverage this decision for that. So nine custodians were unreasonably late. Their data was destroyed during the gap because of the interaction between dropped holds, the two-year retention policy, and the eventual decommissioning of enterprise vault. Let's turn now to whether or not the ESI uh could be restored or replaced, which is one of the prongs under the Rule 37E analysis. Now, the ESI, the lost ESI cannot be restored or replaced through additional discovery, is this last threshold element that we have to hit. And this element gets glossed over a lot, and Medtronic really tried. Give it a serious run to try and defeat it. The court looked at the advisory committee notes to Rule 37E, which state quote, because electronically stored information often exists in multiple locations, loss from one source may often be harmless when substitute information can be found elsewhere. Close quote. But quoting Hollis versus Civa Logistics with emphasis in the original, the court also quoted: the question is whether the electronically stored information can be restored or replaced, not whether the moving party could get similar information by obtaining statements from witnesses. Deposition testimony does not substitute for missing ESI. The burden of showing irretrievability is on the moving party. And the numbers here of what was lost are very striking. The real leaders put forward without any contradiction from Medtronic the volumes of data that were destroyed. Over 189,000 records from Kent, 160,000 from Langer, 78,000 from Medlow, 112,000 from Milmark, 16,000 from Morzak, and 42,000 from Neil, plus undetermined volumes from Klaus and Haben and William Paul Harris, the custodian, who was not included on the hold because of the name mix-up. Medtronic did what a sophisticated defendant should do. And this was a really great call on their part, even though it wasn't ultimately successful in defeating the element. It ran reverse searches across every other preserved custodian for emails where the impacted custodians appeared in the two from CC, B, C C or body text fields. It recovered 124,000 of Kent's, 67,000 of Langers, 42,000 of Middlemarks, tens of thousands that were genuinely replaceable. However, that did not defeat the motion because, and this is another issue to keep track of, replacement does not have to be perfect to fail this element, but it does have to be substantially complete. Under Borum v. Brentwood Village, cited by the court from this same district, where, quote, at least some responsive emails were irremediably lost, and the moving party cannot fully recover them. The irretrievability threshold is cleared. Many thousands of emails remain unaccounted for here, and that gap is what mattered to the court. Run the reverse searches anyway. They reduce your prejudice exposure and they demonstrate good faith. So even if you find yourself in this situation, don't just assume that you're gonna lose. Run the searches, try to essentially offset any losses that you have and see what you can come back with in terms of whether things are irretrievably lost. Reconstruction is harm reduction. It's probably not going to change the outcome on this element, but it does show a lot of good faith and it goes to the prejudice and the intent to deprive that the court will consider later. Now, let's talk about intent to deprive, because that's the court's next step in the analysis. The court found that the realtors had not demonstrated that Medtronic had an intent to deprive to get two E2 sanctions. Even with misstatements in declarations and evolving explanations, the courts apply a very stringent standard when it comes to intent. Again, citing Hollis, the court noted that where conduct, quote, can be adequately attributed to incompetence, courts should not infer malice. And the operative line here: the failure to institute a legal hold is generally insufficient to support a finding of intent to deprive. Close quote. Now, put that side by side with many of the other cases that we've covered on Case of the Week about this intent to deprive analysis. And we've talked about how the courts have varying bars depending on who the judges you're in front of on this E2 analysis. In let's look at a couple of cases where the courts have found intent. In Nagy versus Outback Steakhouse, the District of New Jersey found intent through selective preservation of video footage. The court outright rejected the incompetence defense. In Jones versus Riot Hospitality, which we covered on episode 135, the Ninth Circuit upheld intent based entirely on circumstantial evidence, missing texts that the plaintiff couldn't explain. And in Skanska versus Bagelheads, which was episode 107, the 11th Circuit defied intent as the willful destruction of evidence with the purpose of avoiding its discovery by an adverse party. So, why no intent here then in Medtronic? Two reasons that I can see. First, Medtronic statements were attributed to incompetence and miscommunication, not to a scheme. Second, and this is really the important point that it's worth taking in, the realtors did not put forward evidence that Medtronic forgot to put certain custodians on hold because those custodians' data was bad for the company. Now that begs the question: what happens in the next case when a plaintiff can show that a party strategically left key custodians off hold because their data was unfavorable? Nothing in this decision forecloses that argument. And arguably, the other cases that I cited to you would suggest a finding of intent in that situation, as long as the plaintiff could demonstrate it. If you suspect that pattern, look at who was added, who was left off, and whether the gaps correlate with seniority, proximity to the alleged conduct, or damaging facts that surfaced elsewhere. For defense counsel, the same point cuts the other way. Assuming opposing counsel is studying your custodian list for patterns. Your whole decisions need to be defensible on their face. Now the court did not find intent, but moving to the E1 analysis, the court did find that the realtors were prejudiced by Medtronic's loss of information. And this is the part that drives home a principle that we talk about constantly on the case of the week. You have to prove prejudice custodian by custodian with documents, not arguments. Under the sum proof standard, the realtors have to show with at least some evidence that the lost ESI would likely have been favorable to them. The realtors cleared that bar for six of the nine impacted custodians by producing actual artifacts. A 2012 email from a senior AGM about voiding diaries and prior authorization. A 2012 internal email about sales reps coercing patients with marginal results, and the 2014 support link script called warm intake. But the realtors did not clear the bar for three custodians because they didn't have similar artifacts tying those three to the alleged conduct. Same case, same theory, same legal hold problem, different result because the evidence presented to prove the prejudice was different. This is what creating a factual record means in practice. Prejudice is custodian by custodian, issue by issue. Turning to remedy, now, after finding prejudice, the realtors wanted Medtronic precluded from making certain arguments. And the court declined those preclusions. Instead, the court ordered what the advisory committee called a common curative measure. The parties may present evidence and argument to the fact finder about the loss of ESI from those six custodians in which there was prejudice. The court then looked at fees and the argument from realtors that Medtronic should pay their fees on the motion. And this is where the real financial exposure lives for Medtronic, and it's broader than most practitioners might assume. The court did not just award the cost of the sanctions motion itself against Medtronic. The fee shifting order instead covered three separate buckets. First, fees and costs in conducting discovery into the spoliation, including preparing for and taking the depositions aimed at establishing the spoliation. Second, the cost of litigating the sanctions motion itself. And third, fees on fees if the parties cannot resolve the fee amount on their own. The amount, according to the court, will be reduced to reflect the realtor's limited success, but the scope of those costs is very real. And a sanctions motion at this scale costs the defending side tens of thousands of dollars on its own. Now add paying the plaintiff's side investigation and motion costs on top of that. That's the spread that you are wagering when you decide to litigate a whole timing dispute instead of just negotiating it. So Medtronic loses on roughly half of what the realtors asked for, faces a multi-bucket fee award, and at trial, if the case ever gets there, will have to defend itself in front of a jury that will be told relevant emails from senior personnel are missing. Not nothing, but it's not the existential sanction that the realtors wanted either. All right, let's talk about our takeaways, and they are plentiful from this particular case. First, build your custodian list from the complaint and the request for production and know that it is an iterative process that continues throughout the litigation. If a discovery request asks for documents for a category of personnel, you are on notice that that category is important from the date of that request. The court used Medtronic's own later decision to add sub support link personnel as evidence it should have added them in 2016. Don't let that argument get used against you. The custodian list has to be re-evaluated continuously. Know who's left the company, what IT initiatives are happening that might impact the availability of data, and what new categories of information are being added via discovery. That level of knowledge requires a close relationship with those at the client that are managing this process, or that counsel know and understand what's happening internally at the client. My role as Discovery Counsel is often to be a regular liaison between IT and legal to understand what those issues are that the company might be facing, to be able to advise outside counsel. You need someone like that on every matter. I've even had situations where a legacy database is discussed in emails produced five years later that isn't relevant, but because it no longer existed, the other side tried to leverage it. Creating sanctions motions in cases that won't be successful on the merits is a real litigation strategy these days. Don't fall into the trap. The only way to stay out of the trap is to make sure you've got the process in place and you're documenting it. Next takeaway. Sophisticated parties don't get to plead name confusion. Now, I think legitimately here with these similarity of Paul Harris and William Paul Harris, these are legitimate confusion. That's a tough one. But you're going to have to, in light of this court's decision, capture employee IT, job title, employee ID, excuse me, job title, department, email, not just names. Confirm every new hold edition against the source document and against the department and the information they should be covering, not just memory. When a common surname goes into the hold, your system surfaces every other employee with that surname and forces a deliberate choice. One other piece there that we've come up against many times. Women sometimes change their name when they get married. And oftentimes they either get married or get divorced and change their names back during a hold period. And the email address or the name that you're searching for might change. That's something you have to take into account on your legal holds. Next takeaway: a two-year retention policy plus a dropped hold equals data destruction. You need to plan for it. If you have an enterprise retention policy, you need a process that reevaluates every released hold against every other active or reasonably anticipated matter before the retention clock starts running again. When I was telling you earlier about the process that I managed for a client over a six-year period, we revisited that hold structure quarterly. And if there was a case that was released during that quarterly time period, like in between when we would have done the review, we took immediate steps to get rid of information. We wanted to make sure there was not additional information available that would have to be kept on a subsequent matter when it should have been released. So we took immediate steps to be able to release those things. That's a process. Someone needs to be managing that process continuously. We had to document it step by step when the attorney signed off on the hold, when a release notice was sent, how long after that release was sent, we actually released the information or deleted the information. And then it was all documented thoroughly so that we had dates and backup if it was necessary to defend in a court situation. Next takeaway reasonable steps does not mean perfection, but it has to be a plan executed in good faith with people who know what they're doing. If your client cannot produce a comparable record, you cannot cite to Actilion or Sedona protection. If you're decommissioning a legacy system, run wide-ranging searches first, document them, and confirm preservation for every active matter. Next, monitor your holds. And when discovery goes on a stay, monitor them even harder. The duty to preserve is ongoing. Build a monitoring cadence that involves quarterly checks, custodian additions, departing employee reviews. And if a case may sit on a stay on a motion to dismiss a jurisdictional dispute in the MDL, your monitoring obligation goes up, not down. Stays when retention policies stays are when retention policies do quiet damage that you don't realize until years later. The risk of data loss through normal operation goes up exponentially during a stay. Overinvest in that process. Don't coast. Think of it as insurance against a very expensive sanctions motion. Next takeaway intent to deprive is still a very high bar, but this decision does not foreclose intent in the next case. Courts won't infer intent from incompetence, but in Outback, riot hospitality in Skanska, courts found intent through selective preservation and circumstantial evidence. The distinguishing factor in Medtronic is that the realtors did not show custodians were left off the hold because their data was bad for the company. If you can show that pattern, or if you can show what that data might have shown, this decision is not your obstacle. Next takeaway. Generic prejudice arguments are going to fail. Specific paper-backed prejudice arguments are going to win. Next takeaway: reverse searches reduce harm, but don't close the door on sanctions. Under Borum and Hollis, replacement has to be substantially complete to defeat their irretrievability element. Tens of thousands recovered document, tens of thousands documents that are recovered isn't enough if thousands of them still remain unaccounted for. Run the reverse searches anyway, just don't assume they defeat the motion. The fee award is where the real financial consequences land, and the scope is broader than you think. Less than 1% of cases reach a jury. So the E1 evidence and argument to the fact finder remedy is structurally limited. It's not going to get the plaintiffs very far. But the fee award is real and immediate, and it covered three buckets here: investigating the spoliation, litigating the sanctions motion, and fees on fees if they happen. If you're arguing prejudice, document every hour of spoliation investigation work in real time. If you're defending, the time to settle a whole timing dispute is at the meet and confer, not at the sanctions stage. The way that we document specific things for spoliation in time charges is literally to put the word spoliation colon and then the time charge. That allows us to filter and sort. And then we have documentation that every person working on that matter was working on the specific spoliation related information. All right, that was a lot of takeaways. And that's our case of the week segment for today. If you take nothing else with you today, take this. A legal hold list is not a preservation plan. A preservation plan is a process, a process that reads the complaint and the RFPs, that identifies categories of personnel, that flags similar names, monitors compliance, handles departures of employees, and survives a five-year stay with data quietly disappearing. If your whole program isn't doing those things, you're not running a hold. You're building risk into your case. If you have clients that you are managing multiple aspects of litigation for, your firm should have someone in the mix at that client who understands what's happening to their data and what decisions are being made by IT that might impact preservation. If you've been listening today and thinking about your own legal hold program or your clients, the matters that you're tracking, or the custodians you're worried about, the retention policies that are quietly running in the background, these are the problems that we're solving with Minerva 26. Minerva 26 is a discovery strategy platform that connects the case law, the rules, and real-world workflows. We curate and tag every discovery decision in our database, which is about 50,000 cases now, by issue so that when you're preparing for a meet and confer, drafting a hold notice, or building a sanctions response, the rulings you need are already organized for you. You can book a 30-minute demo with our team if you're interested in seeing how Minerva 26 can help with your process at Minerva26.com backslash demo. Come prepared with one discovery problem that you're facing right now, a hold question, an ESI protocol that you're negotiating, or a sanctions issue you're trying to get ahead of. And we'll walk through how Minerva 26 shows you what's possible to manage on that matter. If this episode was helpful, please share it with a colleague and post it on LinkedIn with your biggest takeaway. The fastest way to level up discovery strategy is to normalize all of these conversations in public with everyone who touches them. Subscribe to the Meet and Confer podcast so that you don't miss the next ruling that changes what reasonable looks like. Thanks for listening. We'll be back next time with another case of the week.