#CaseoftheWeekCase Law

Episode 166: Proceed with Caution When Your Search Term Hits are a Multiple of What You Actually Produce

In Episode 166, Kelly Twigger discusses the importance of identifying and resolving discovery disputes in advance of the discovery cutoff as well as the court’s novel approach to allowing the parties to test the relevance of documents not produced given the disparity in produced documents vs. search term hits in Valeo Schalter und Sensoren GmbH v. NVIDIA Corp.


Introduction

Hi and welcome to our Case of the Week segment on our podcast. My name is Kelly Twigger. I am the Principal at ESI Attorneys, a law firm for ediscovery and information law, and the CEO and founder at eDiscovery Assistant, where we take the insights from our practice and provide a knowledge platform for you to leverage the power of Electronically Stored Information (ESI). Thanks so much for joining me today.

Case of the Week Episode 166

One quick announcement about the upcoming University of Florida eDiscovery Conference, to be held virtually and in person on February 12-13 2025. This is 16+ CLE hours of practical education from some of the best in the business. It is free to attend virtually, so please mark your calendars and get registered. I’ll be moderating and participating in the case law session at the start of the conference, together with United States Magistrate Judge Elizabeth Chestney from the Western District of Texas, Maria Salacuse of the EEOC, and Bansri McCarthy from Morgan Lewis. I’ve had the pleasure of participating on the planning committee for the University of Florida eDiscovery Conference for the past 11 years and we start working together in May, so a number of months ago, to put together an amazing program about the current issues facing us in electronic discovery. You can view the full agenda and register at ufediscoveryconference.com.

On with the show.

This week’s case emphasizes the importance of timing in discovery. Now we’ve seen a number of decisions in ediscovery case law over the last four years where courts deny discovery to parties for the failure to complete it prior to the discovery cutoff. That’s a new trend. We hadn’t seen that previously with paper discovery. Today’s decision addresses that issue and then goes on to discuss some highly technical issues related to the discovery of ESI.

Now, do you need to be addressing each one of the sources of ESI that we’ll discuss? That depends on your case, but what you do need to do is identify the sources of ESI early on so that you can have a plan to deal with each one within the bounds of the discovery dates that the court gives you.

Let’s dive in.

This week’s case comes to us from Valeo Schalter und Sensoren GmbH v. NVIDIA Corp. This is a decision from United States Magistrate Judge Virginia DeMarchi from January 9, 2025, so just from a couple of weeks ago.

If you’re a listener on Case of the Week, you know Judge DeMarchi well. She has authored 182 decisions in our eDiscovery Assistant database, which I dare say ties her for one of the top spots. She sits in the Northern District of California and is very prolific in writing thoughtful ediscovery decisions. She is often presented with and handles some of the most complex discovery issues in the country, together with her fellow judges in that district.

As always, we tag each of the decisions in our eDiscovery Assistant database with our proprietary issue tagging structure, and this week’s issues include failure to produce, hyperlinked files, form of production, search terms. and 30(b)(6) corporate designee. We’ll include the link to the decision for you to be able to review in the show notes.

Keep in mind that each of the decisions in eDiscovery Assistant is a public link so anyone can view the full text of those decisions. We built that specifically to allow everyone to have access to the full text of the case law to be able to do your analysis on these ediscovery decisions. Case law is changing so quickly. We have more than 5,000 decisions for 2024 in our eDiscovery Assistant database, and we’ll be issuing our 2024 Case Law Report at the UF eDiscovery Conference. So if you’re registered there, that will be part of the available materials. You’ll also receive a link via our weekly case law newsletter to be able to download that report If you’re interested. If you’re not signed up for the blog, you can do so at ediscoveryassistant.com/blog.

Let’s dive into the facts and the analysis for today’s decision. This is a short decision but it raises some really interesting takeaways for us from a practical perspective for ediscovery.

Facts and Analysis

The parties are before the Court here on four different discovery disputes raised with the Court, by letter, on various responses that NVIDIA gave in response to plaintiff’s discovery requests. Now, we’ve talked about this before, but raising discovery disputes by letter versus standard motion practice and the briefing schedule has become a much more effective way to bring discovery disputes to the court’s attention. It both moves things forward faster and is much more efficient for the parties and the court, which is the key goal of Rule 1 of the Federal Rules of Civil Procedure. So, always find out whether your court has a letter procedure to be able to deal with discovery disputes. Some courts even just allow you to contact the court by phone and set up a brief conversation with the court to keep things moving forward. This is a vast difference between motion practice, which can take months and set back your discovery efforts.

I want you to keep it in mind as we talk about the timing issue in today’s decision, because if you have a full-on motion practice for a discovery issue, you’re going to run up against your discovery cutoff and you’re going to have things you want to do past that discovery cutoff, and, absent really good cause, courts are not extending those discovery cutoff deadlines these days. Instead, they’re telling you have to get your discovery done in the time that they give you at the outset.

And that is the first thing that the Court considers in these discovery disputes before her — is whether or not District Judge Lee’s standing order, who was the District Court Judge here in effect for the case, precluded her even hearing these discovery orders. So Magistrate Judge DeMarchi looked first to District Judge Lee’s standing order, which provided that the close of fact discovery was the date by which “all discovery must be completed, including motions to compel, hearings on discovery motions, and any additional discovery resulting from orders on discovery motions.”

Because the disputes at issue were raised with the Court days after the close of discovery, Magistrate Judge DeMarchi found that the disputes did not comply with District Judge Lee’s standing order. However — and it’s a big however for the plaintiffs — the case was subsequently assigned to District Judge Pitts, whose case management order allowed for discovery motions to be filed no more than seven days after the discovery cutoff.

Now, by my count, the plaintiff actually filed its letters here eight days after the discovery cutoff, not seven, which is not addressed by the Court. So it could be that my count is off, maybe the dates are off, not really sure. In spite of those date cutoffs, the Court here heard from the parties on January 7th and found that most of the matters could be resolved expeditiously on the merits. To me, that really raises the question of whether — if any of the issues was more involved —Magistrate Judge DeMarchi would have relied on the date cutoffs to deny the motions.

That’s something to keep in mind all the time. You need to be aware of the court deadlines. Too often, we don’t calendar all of the deadlines in the case management report and they sneak up on us or pass us by before we realize it. Someone on your team needs to be responsible for calendaring those dates, setting up reminders with sufficient time in advance of them, and modifying them when the case management order is modified. That happens a lot in litigation and those dates need to constantly be updated. Calendar everything, and set reminders of all deadlines at some point in advance that make sense.

Here, I probably would have set a reminder at least a week in advance of the date to ensure that my client had raised all open discovery disputes as required. Don’t take the chance that your judge will allow the disputes after the fact like Magistrate Judge DeMarchi did here. However, you can also use this decision, if you’re in the Northern District of California or in front of Magistrate Judge DeMarchi, to have her order that, because the complexity of the dispute is minute, as it really seemed to be here, that she could continue to rule on them and move the case forward.

Let’s look at the individual issues that the court addressed here. The first issue is one that both highlighted the difficulty of using search terms to identify responsive documents and also really shocked me that the Court ordered it. Plaintiff complained that there was a large discrepancy between the number of documents from NVIDIA that hit on the agreed upon search terms and the number of documents actually produced as responsive. So, big gap. The Court had little information to determine the basis of the discrepancy, and it did something I’ve never seen a court do before. It ordered the parties to conduct an experiment to determine whether documents that were not produced were actually responsive. That experiment was to be as follows: Plaintiff would choose a custodian from NVIDIA; NVIDIA was then to provide a list of all of the control numbers of documents that hit on the search terms but that were not produced; then Plaintiff was then to select a hundred of those control numbers and NVIDIA was to review those documents for privilege. Any documents that were not privileged from that 100 were to be provided to the plaintiff for them to review for responsiveness.

If you’re shaking your head in disbelief, I am right there with you.

The absolute cornerstone of party-based discovery is built on the notion that we have to trust the other side in producing the responsive information required by the Federal Rules of Civil Procedure. This approach from Magistrate Judge DeMarchi really seems like a slippery slope to go down.

Now, back to that process. If the plaintiff finds anything in those documents that it believes to be responsive, the parties are to meet and confer. If an issue still exists, the parties are to bring it to the Court or advise the Court that the issue is resolved. The Court also required the parties to set out a schedule for the stages of the experiment and file it with the Court. It’ll be interesting to see how that one plays out.

The second issue at play here is one of our favorites on Case of the Week — the production of documents from hyperlinks. Because there are just 10 documents that the plaintiffs are requesting, the Court ordered NVIDIA to produce them without a lot of discussion and, frankly, without any argument from NVIDIA.

The third issue here simply required NVIDIA to provide the Bates numbers to plaintiffs of the financial projections it had previously produced. Now that’s a small issue, and there’s no info here as to how much data NVIDIA had produced, but it is often easy to bury documents in a collection and that’s because Rule 34 allows documents to be produced as they appear in the ordinary course of business.  As far as ESI goes, that means that parties basically lump all the information together and provide it to opposing counsel without any form of organization.

If you want to have more organization to the productions that you receive, you need to negotiate that at the outset. I refer to that as manner of production — how the documents are organized. Ask for data to be organized in a certain way when it’s produced or to have metadata fields assigned to it that allow the receiving party to be able to sort and find key information. That’s an easy remedy, but it is only available to you if you think about it before you send your requests for production. It’s something that can go in your ESI protocol. It’s something that can go in your instructions. The Federal Rules of Civil Procedure do not distinguish between form and manner of production. Produced in the ordinary course of business is what I refer to as manner of production, and so, because Rule 34 also allows a party to provide for the form of production that it wants in its request for production, you can include that language in your request for production or in your ESI protocol to ensure that you receive a more organized production.

The final issue before the Court dealt with whether NVIDIA’s 30(b)(6) designees were prepared to testify on specific topics. If you’re not familiar with a 30(b)(6) deposition, a requesting party can use the subsection of Federal Rules of Civil Procedure, which is 30(b)(6), to request a list of topics that it wants to hear from a corporate designee on. It wants to better understand those topics and how they work at the corporation. That can be any number of things. It can be how information is created. It can be relationships with third parties. It can be any number of topics that can occur. What happens is the requesting party for the deposition prepares a list of topics that it wants to hear from someone on, in this case at NVIDIA. NVIDIA then provides witnesses who are designated on the specific topics outlined in the subpoena. Typically, a 30(b)(6) subpoena is a standard subpoena with topics listed on Exhibit A, and then each topic is organized by number and each witness is assigned specific topics.

So, essentially, the plaintiffs were arguing that some of the witnesses who were provided by NVIDIA were not prepared to testify on the specific topics that it felt they were designated for. But the Court found that the parties were really in disagreement on who was designated for which topics and whether or not they were prepared to testify on them during the depositions, and the Court ordered the parties to meet and confer and provide the Court with more information.

Obviously, here on our Case of the Week I am Monday morning quarterbacking and we have no insight into the transcripts from those depositions. But the best trick to ensure that you know which witness is covering which topics is to ask up front at the depositions, using the topics in the notice. If there are 10 topics and you are taking the deposition of the 30(b)(6) deponent, ask each witness whether they are testifying on each specific topic in the list and then confirm it at the end. Go through topics one through 10, confirm they’re testifying on one, two, five and seven, and then, at the very end, say, “just to confirm and rehash, you’re here and prepared to testify on topics one, two, five and seven on behalf of NVIDIA.” Then you have the language that you need to be able to go back to the court if, in fact, that witness is not prepared to testify on that topic.

In addition, I would say this. Lawyers prepping witnesses for these depositions will, or I guess I should say should, prep their witness to know exactly what topics they are responsible for. Having that in the transcript can eliminate any question about whether a witness is prepared on a topic for which they are designated. I’d also encourage you to explore the witness’ background and why they are the person who is designated to be on a specific topic. A lot of times in these instances we’ll have corporations that will “prepare” a witness who they know is a good witness and could maybe cover three of the five topics, and they’ll “prepare” them on the other two topics, even though they may not be the best witness for those topics within the organization. That is your job as the deposing party to flush out and ensure that you have made a record for the court as to what their experience is on that topic and whether they are the best person to present that topic for the corporation.

Those are the four topics that the Court covered. Having issued rulings on each of the disputes, the Court then required the parties to ultimately seek an extension of fact discovery from District Judge Lee for the limited purpose of completing the issues outlined in this decision. That is really in keeping with the language of District Judge Lee’s order, because Magistrate Judge DeMarchi has essentially ordered that the parties can continue some of these discovery disputes outside of the District Judge’s language, which said that everything had to be completed as of the close of fact discovery.

Takeaways

What are our takeaways from today’s decision? And we’ve discussed some of them in the context already, but let’s go back for a minute to Magistrate Judge DeMarchi’s experiment and the Pandora’s box that that can open.

I have to believe, because Magistrate Judge DeMarchi is an excellent judge, that there was some additional information in the parties’ papers that led her to this experiment of allowing the plaintiff to test NVIDIA’s production for responsiveness. I guess I could say this — that there’s nothing in the actual facts that talk about the vastness of the disparity. So I don’t know whether there were 200,000 hits and they only produced 10,000 documents, which would be quite a disparity and might be the basis for this experiment. So, there has to be more information here, factually, rather than just to be able to assert that if you believe that there’s a large enough disparity between the number of search term hits and what’s actually produced that’s responsive, that you should be able to argue for this experiment that Magistrate Judge DeMarchi allows the parties to engage in here.

I have to believe there was more information that led to that decision, but also that other courts would look at that and question whether that was an appropriate experiment. Otherwise, allowing this experiment is one that will lead to other parties to request the same accommodations and that may wreak havoc. You know, the 100 documents that were ordered to be reviewed here is not that many and it’s not particularly time-consuming, but it’s more the principle of the thing that you’re having to provide non-responsive documents to the other side. Whether they contain other kinds of business information, or anything that you might not want the other side to see, that is not privileged and would not be protected in this experiment.

Typically at this stage the parties are drafting summary judgment motions and constantly moving forward on other aspects of the case, and constantly revisiting discovery issues is hard. It’s time consuming. There’s so much detail to it that someone has to be on top of for your team.

Next, we talked about the timing issue. You need to be aware of the case management order when your discovery cutoff is and that you’re planning your discovery well in advance of it.

We’ve talked repeatedly about the need to engage in third-party discovery early on because third parties typically do not respond in a timely manner and we’ve seen decisions where courts have disallowed responses from third-party discovery because the requesting party did not put the request out there early enough in the process. Oftentimes, if you’ve done it less than 30 days before the close of discovery — which is how long someone would have to respond to a third-party subpoena — the courts will disallow the discovery. So be aware of that timing issue.

We also talked about manner of production and the organization of productions. Think about the information that is most crucial to you, and there’s a few ways you can do that. You can agree early on that you have subsets of categories of data that you want to be produced first, because that’s going to really inform your analysis of the case. There are other things that you can do in organizing productions. By adding a metadata field, you can ask for specific requests to be denoted. Are you going to get it? You’re not necessarily entitled to it under the Federal Rules of Civil Procedure, but some states are starting to allow or require the documents be produced according to request. California is one of them.

Finally, let’s talk about the 30(b)(6) notice and depositions. In highly technical matters like this one, the 30(b)(6) depositions to understand technology can be critical to a case. I already talked a little bit about the fact that sometimes corporations will play fast and loose with preparing someone on these issues who maybe isn’t the foremost expert at the company, but they want to be able to control the witness. The rules don’t require a foremost expert at the company, so I’m not suggesting that those parties aren’t meeting their obligations, but what that means is that the attorney taking the deposition needs to identify out of the gate which topics that witness is responsible for, what the basis of their knowledge or experience is for being presented as the corporate representative on that topic, and how much they really know. The best way to do that is to get it all on the record at the deposition so you can point to that testimony in any motion that you have to make to the court if you feel like the witness that you have on a topic is not well prepared to be able to answer the questions.

We talk over and over again here on how facts are critical to making a case for a discovery dispute. Judges are not willing to accept a lawyer’s arguments without a factual basis. We saw it here with plaintiffs arguing that NVIDIA’s production was vastly smaller than the number of hits on the search terms. No facts really offered by the parties, or at least that were not included in the decision. We can only use what’s in the four corners of the decision. So the Judge came up with an experiment to help the parties find out the facts. Don’t make the judge find the facts for you, because often they will not. They will punt your motion for lack of basis, in fact. Do the work to put the facts needed before the court on each dispute you have with the best way to do it. In this case, deposition testimony, numbers of documents, any kind of basis for reasoning as to why you believe the search terms should have hit on more responsive documents. What was part of the negotiation process for those search terms? Again, a lot of that could have happened here between these parties. We simply don’t know, because we’re tied to the four corners of the decision that we’re reviewing for Case of the Week.

Conclusion

That’s our Case of the Week for this week. Be sure to tune in for our next episode, whether you’re watching us via our blog, YouTube, or downloading it as a podcast on your favorite podcast platform. You can also find back issues of Case of the Week on your favorite podcast platform and be sure to subscribe, as we’ll be adding new content apart from the Case of the Week segments. Have a great week!

As always, if you have suggestions for a case to be covered on the Case of the Week, drop me a line. If you’d like to receive the Case of the Week delivered directly to your inbox via our weekly newsletter, you can sign up on our blog. If you’re interested in doing a free trial of our case law and resource database, you can sign up to get started.



Categories
Archives
Privacy Settings
We use cookies to enhance your experience while using our website. If you are using our Services via a browser you can restrict, block or remove cookies through your web browser settings. We also use content and scripts from third parties that may use tracking technologies. You can selectively provide your consent below to allow such third party embeds. For complete information about the cookies we use, data we collect and how we process them, please check our Privacy Policy
Youtube
Consent to display content from - Youtube
Vimeo
Consent to display content from - Vimeo
Google Maps
Consent to display content from - Google
Spotify
Consent to display content from - Spotify
Sound Cloud
Consent to display content from - Sound