eDiscovery Leaders Live: David Stanton of Pillsbury Law
David and I discussed how he got started in eDiscovery, the expansion of eDiscovery in his firm, and the growing risks and demands that impelled Pillsbury to build out their eDiscovery operations. We talked about how to build a strong internal eDiscovery operation. David talked about that build-out, especially the importance of the people involved, and about the imperative for them to compete on value. We also examined the goals behind eDiscovery; principal among them, as David noted, speed to narrative. David talked about the technology stack that powers the work his team does, in particular the power of advanced AI and reusable AI models, and closed with his thoughts on an ideal eDiscovery platform.
Recorded live on January 15, 2021 | Transcription below
Note: This content has been edited and condensed for clarity.
Welcome to eDiscovery Leaders Live, hosted by ACEDS, and sponsored by Reveal. I am George Socha, Senior Vice President of Brand Awareness at Reveal. Each Friday morning at 11 am Eastern, I host an episode of eDiscovery Leaders Live where I get a chance to chat with luminaries in eDiscovery and related areas.
This week our guest is David Stanton, David is a partner at Pillsbury. He’s been at Pillsbury since 2004 and has put together an extensive list of accomplishments, which I will not try to detail. I’ll just point out a few things. First, David leads the firm’s Information Law & Electronic Discovery practice, where they take a multi-disciplinary approach to the operational and legal challenges of large data sets in litigation, internal investigations, and mergers and acquisitions. He’s also proficient in cybersecurity, privacy, incident response – those areas that border on eDiscovery or on which eDiscovery borders. He collaborates with in-house lawyers and IT professionals on implementing robust information governance programs and related policies and procedures. Among other things, he created PEARL, which is Pillsbury’s E-Discovery Alliance of Resource Leaders. He’s been recognized by Chambers Global, Chambers USA, The Legal 500 U.S., and Who’s Who Legal.
Before joining Pillsbury, he was an attorney at Chadbourne & Parke and before that at another firm, Troy & Gould. David has his B.A from St. John’s College and his J.D from University of California, Berkeley, Boalt Hall School of Law. David, welcome.
Hi George. Yeah, thank you. Thanks for the introduction.
I’m glad to have you with us.
I’m glad to be here.
Getting Started in eDiscovery
I’d like to start with your journey of the last five years and the transitions that you’ve gone through such as the transition to the cloud.
As you mentioned, I’m a litigation partner. I’ve been working in this eDiscovery space, I guess it’s for a couple of decades now, and like many people kind of sided into it. At the time that I was coming up, there wasn’t a career path that really existed around eDiscovery. It just happened by happenstance and maybe I had an aptitude; I was brought up in a family that was exposed to a lot of computers early on and so fell into managing this thing called e-mail, back when discovery was early on. I just had that faculty, I was just that kind of an attorney, as I’m sure many in the audience are. We had an interest in technology, had an interest in dealing with these kinds of challenges and problems, and over time took on more and more responsibility in that area.
Out of curiosity, was this something that you actively pursued from the onset or were you like some of us, where you said “This is really eDiscovery, I don’t want anything to do with it. Please don’t pull me into this?”
“There did come an inflection point where I really did start to recognize that this was going to be a thing, that eDiscovery was going to become more and more important.”
I wouldn’t say that I actively pursued it, because it didn’t really exist but I was up and coming so it wasn’t something to pursue. There did come an inflection point where I really did start to recognize that this was going to be a thing, that eDiscovery was going to become more and more important.
The discovery already is such a big part of litigation. I was at a pretty big firm, as you mentioned, by 2004 and most of what the cases were about, for better or for worse, was the discovery process. More and more of those matters were involving data and I enjoyed getting to know all of the vendors and all of the sponsors and going to conferences and educating myself about this entire space.
An Expanding Practice
You mentioned this last four or five years, and this was a bit of a windup for that, but at that time, in the 2010 to 2014/2015 time period, we were working at Pillsbury with a whole lot of different vendors. We had, like many partnerships, a federated approach where litigators could make their own decisions about which vendors they wanted to work with. The eDiscovery operations happened in a very transactional way. They were based upon a case. A matter came in, then you put out bids and you got some offers to do the project and you selected your vendor and you ran with them. It was all pretty informal at the time.
For a while we tried to embrace it, we tried to improve upon it and put some stability into those systems. We pursued that through that operation that you talked about when you made the introduction, that we called PEARL, where we identified a select handful of vendors, got to know them, put together master services agreements, standardized our contracts and terms with them, made it much easier to on-board those relationships and bring those folks into a matter, gave a panel approach if you will. That was fairly successful, it was somewhat helpful.
But what we found is that our efforts at stability were kind of fleeting, because every one of the organizations that we made a relationship with continued to evolve internally, they continued to combine with other organizations, the project managers that we got to know got reassigned or left or went somewhere else. We didn’t have any control over what was happening inside the organization around those kinds of decisions. It continued to be very difficult to manage and maintain this multiplicity of vendor relationships around eDiscovery projects. And at the same time, you started to get a lot of breach activity out there and data incidents became more and more notorious and we saw a risk for the firm.
One of the perspectives that we at Pillsbury have on all of this is that we are fiduciaries around client data. When we are litigating a case or working an investigation and we arrange to take custody of a client’s information assets, we have an obligation. Our duty of confidentiality, our duty of supervision, our duty of competency attaches to those state assets. We need to operate mindful of those risks and responsibilities. We started to see that it made sense for us to consolidate all of this data that was out there. I think that’s a thing that a lot of organizations have begun to realize. There was a time when I would do these kinds of panels and presentations and I’d say, “It’s 4 o’clock, do you know where your data is?” Do you know as an organization where you have sent emails over the past 5 years? Do you know? Do you have a sense of who has hard drives with gigabytes of your employees’ files, emails from the C suite, unfiltered repositories of information that have been gathered up for this litigation and that litigation and just sort of sent out. We saw that as risks for our clients, and we saw that as a risk for Pillsbury.
Pillsbury’s Response: Building a Strong Internal Operation
We looked to consolidate that operation internally and effectively to in-source the eDiscovery vendor operation, to build our own alternative legal service provider inside the firm. That journey over the last five years has been about situating that operation internally, identifying an appropriate technology stack, championing the program, getting the approvals, putting a staff and a team together and then sort of building and maturing that operation over time.
There are a number of challenges there. I’d like to touch on one part first and then come back to in some ways the bigger issue of how you actually do all of this. One of the challenges historically has been that this is a law firm after all. Partners feel rightly or wrongly that if there is a set of preferred providers or set of preferred tools, those are just preferences; they use other things. And then you also have clients who say, “Yes, we understand you want to do this but here’s how you’re going to do it” or “Here’s how we’d like you to do that”. How do you address those parts?
Competing on Value
It’s a good question, and we’ve been dealing with this for a while. I see them as very different types of clients. While my partners are in a big way my clients, the firm’s clients have a different relationship. When firm clients have made a decision and they’ve gone through the process of identifying and selecting appropriate vendors to handle their eDiscovery projects, it’s very rare that we will contradict those decisions. Sometimes we see that as having been a decision that was made and perhaps not thought through or the resources that the client vendor has might not be appropriate or we think that they’d be better off to work with Pillsbury directly, and we’ll lobby to get that work. But we tend to fall in line and work in those relationships and normally it works out great.
I’ve long said organisations should centralize their data, they should centralize their eDiscovery operations. They can do that through select vendors, they can do it by hiring things in house, they can do it by engaging a firm like Pillsbury who has those operations. But as they mature they naturally tend to head in that direction and so the more mature organizations seem to have that sorted out to some extent and we work in that.
With the partners inside the firm, the internal clients that we deal with, it’s a little bit of a different relationship. On the one hand, yes, you’re a sophisticated person, you’re a sophisticated partner at a big firm. Certainly, you should have some prerogative to guide your case in the direction you want to go in. Oftentimes though, those decisions historically have been made without a lot of forethought and perhaps based upon relationships and not necessarily thinking through the consequences of how that might affect the case or the relationship with the client over time. The bottom line when it comes to kind of dealing with the partners, I think anybody who is sort of in my position or a position like mine when they’re trying to build an internal operation, you have to compete on value. You have to prove to your internal facing clients that they are better off to work with the firm’s own team than to do that through third party vendors. And I think over time we’ve been winning most of the work that comes in our direction.
Putting Together the Platform and the Team
To compete on value you’ve got to put together that team, you’ve got to put together that platform, and you have accomplished this entirely or to a large degree. How have you gone about doing that?
We’ve been lucky, we’ve had good sponsorship. We’ve tackled a few challenges that I think are going to be common for anybody who’s dealing with this kind of thing that we’re thinking about. One is, who is advocating for this opportunity, for building this out? Clearly, I was on board with it. In Marty Metz, Pillsbury’s former CEO, we had a tremendously supportive sponsor at the C level, who was communicating regularly with the firm’s board and was able to articulate the value and help people to understand how the firm was going to be addressing the risks. There’s no substitute for that kind of relationship.
I think anybody who is in eDiscovery learns to value the relationship between legal and IT. When you’re doing what I do inside the firm, that becomes a really instrumental relationship. It’s key for us to have strong connections between our groups of attorneys and the firm’s own IT department, so that was huge.
We won approval by the board and we had folks who gave us considerable autonomy to go about doing what we needed to do. And then we had to hire. We were transitioning folks in-house. We also had an opportunity to hire an outside director. Again, we got really lucky, we ended up with somebody who is just fantastic in Gordon Moffitt, who is at the firm here.
We run a technology shop and we also run a review shop. On the review shop, we started small and then worked with a lot of flex staff and managing that sort of thing and over time we’ve brought on more and more full time people. Obviously, having good people, having their morale be good, having them see a vision for an operation internally that we can build together, to have that entrepreneurial spirit be alive and well inside of a law firm – it’s a gratifying thing and it’s a really important thing about putting a team together that’s going to collaborate and be successful.
And then finding good technology. We’ve looked to distinguish the firm and to make ourselves competitive around the technology stack that we have put together and I think we’ve done a good job with that. We have leaned far into analytics, we’ve leaned far into predictive coding and the use of TAR. We made those, early on, a commonplace feature of how we approached the eDiscovery process. We leaned into statistics, we leaned into random sampling and validation, so that we could achieve a good quality result and learn to measure that in a lot of different ways.
And keeping an eye on the metrics by which we measure organization across a lot of different criteria. There’s “How fast are you doing things?”, “How cheaply are you doing things?”, “How much revenue are you generating?”, some sort of soft metric of how the morale is across the team, how many people are you bringing on.
Speed to Narrative
“We call it speed narrative. How quickly are we turning a big set of documents into a story that our attorneys can tell, that supports their strategic aims, their settlement aims, their litigation aims across the board?”
The thing that’s the cohesive identity that we have inside of the organization here is getting to the story. We call it speed narrative. How quickly are we turning a big set of documents into a story that our attorneys can tell, that supports their strategic aims, their settlement aims, their litigation aims across the board? That becomes the driving motif that everybody is about. Everybody on the team is looking to improve our speed to narrative.
It seems to me that if you’re not trying to figure out what story to tell and how to best tell that story, why are you doing any of what you’re doing?
Right. It’s such an interesting thing though, because what we do in eDiscovery is oftentimes make a lot of binary decisions. You look at a document, you say is it this or is it that? Is it that or is it that? You might have a heuristic and that might be multiple tags in the way that they get set up, but you’re making a yes or no decision. And there can be tens of thousands of yes or no decisions that get made over the course of the day. Getting those decisions right is very important. It’s very important to learn how to measure the quality of those decisions and how people are doing and, “Are they getting them right?, “Are they getting them right quickly enough?”, and all that sort of thing. And that’s the day to day sort of blocking and tackling.
A Core Key Attorney Review Team Inside the Firm
Being able to step back, to pull the lens back and look at what everybody has done over the course of a day or what everybody has done over the course of a week and to be able to find a mechanism to filter the intelligence up to the partners – it becomes an interesting thing. And it’s a thing that’s particularly advantageous to be working with these groups that are inside the firm, that are already and always part of that conversation – to already have embedded in the way we do business a model where all the document review attorneys are valued as attorneys. We take them very seriously as key members of the team and welcome and appreciate their input and collaboration and conversation around what they’re seeing, what they’re learning, how they’re coming up with things. So there’s time in the course of a day or it’s certainly in the course of the week to step away from their documents and convey that information in one form or another And different conversations along the way that allow that to percolate up. And then get recycled back into the analysis as we go. That’s a big part of how the firm operates.
The Technology Stack You Chose and Its Implications
What you’re highlighting there is some of the value that comes from bringing capabilities in house and not relying solely on external capabilities. It seems to me that at some point you hit a juncture where you don’t have the capacity or the capability that you need immediately. How do you handle that?
We’ve had two places where that happens. The technology that we took on, we took on in the cloud from day one. We had the private cloud environment at the onset. Since then that has been migrated, as many things have, into an appropriately profiled pod or space in AWS.
Being in the cloud gives you the advantage of being able to add more workers and add more machines and if you have properly constructed technology, add more processing, add more memory and take advantage of all the multithreading value and wonder that is the cloud today. That ability to have turnkey resources is really important.
When working in a technology stack, then you have a different relationship. You still have eDiscovery vendors, but we have a different relationship, more stable relationships. When you’re working with a provider like Reveal who has really great technology and their personnel – who also work in an on demand environment, who do still do transactional work on this case or that case – become flexible resources that you can bring on board as needed. If your internal staff is handling day to day administration of your cases and processing and loading and running productions and off-boarding data – which also you need to build into your administrative time, if something comes and it’s just too much to handle, you can arrange it so that you have hourly resources available through the provider to come on board and help with that. Similarly on the Reveal side, we’ve always had a flex combination of full-time and flex staff.
One of the things we’ve looked to do in building that is to make a pathway where we really are trying to value these attorneys and create career opportunities. And eDiscovery does represent a maturation of the general legal sector and specialization happens in businesses that are mature. Think of a time when there were doctors and nurses, and now there are all sorts of different tiers of doctors and nurse practitioners and physician’s assistants, and it’s all sort of broken up. eDiscovery is a place where that is happening in the legal practice and you have different careers that are now being created out of people who start doing document review. We’ve been mindful of trying to create those pathways and bring flex staff into full-time positions and give them opportunities to continue to grow and develop professionally. That flex staff pool creates a body of individuals that we can call upon to handle the cases that we don’t have the resources to do ourselves.
Coming Next: Advanced AI Capabilities, Reusable Models
You have managed to build both a platform in terms of the people you’ve got internally to call upon and undertake all these various tasks, and then a platform in terms of the technology that you’ve settled on to enable all of this. You’ve moved from ad hoc and transactional to something more unified, coherent and the like, where do you go next?
Well, it’s getting fun. We’ve enjoyed a bit of success. I think we’ve proven the model to the firm. We’ve demonstrated to plenty of partners that we are good at this and we know what we’re doing. We see them as repeat business, if you will, and are very much supporting their practices. There is continuing to grow in our own ability. Where do you take a bunch of eDiscovery lawyers who are really good at what they do? My sense is, you get them better and better at telling that story. There’s professional development around, being better at searching, being better at analyzing, being better at converting what you’ve learned in a document set into consumable products that you can deliver over to the case teams that you’re supporting.
We’re having a lot of fun looking at different artificial intelligence capabilities. We have incubated a setup with a data scientist, Alexis Mitchell, who is working with Pillsbury through her organization, Implicature, and building out customized AI capabilities that are assets of the firm that will allow us to automate and script things that we do all the time. In particular, we can use AI to look back and do quality control measures and metrics and identify outlier reviewers early on so that we can take corrective action to bring them back in line. We can use some combination of AI and regular expression filtering to create very interesting chronologies of cases where we can select certain sets of documents and pull up material and start to display it in a reported format earlier in the case than otherwise you could.
We can use it to detect things – and I know you guys are at the fore of this – around detecting incidents of privilege or material likely to be confidential or things that look like different types of contract clauses that you might be particularly interested in.
And there’s a whole world of classifiers and ways to take what we’re doing day in and day out and coding documents and turn that, not just into a model, a TAR model that we’re using on one case, but turn that into models that then we can apply to every case. With a lot of client data running through the firm, being exposed to these kinds of analytics, it gives us a really unique opportunity to build models that are useful to our clients. So that stuff has been really fun. And then continue to try to just keep growing and do this better and better.
The Ideal eDiscovery Platform
In the time that’s remaining, I’ll pose one final question, which is that if you could have no limitations whatsoever, no restrictions whatsoever, what would you want out of your ideal eDiscovery platform?
It’s a great question and a fun one. I would want to have a group of people who are gratified in what they’re doing and are continuing to be challenged by interesting and important legal challenges, legal problems. That’s what we’re doing. We are a group of people and we are looking to grow and get better and better at what we’re doing. It’s no good if you have a really great platform but it doesn’t enthuse your people to want to be working on it and make them develop a kind of expertise.
On the flip side, you get into the technology and that’s where it’s getting really in my view, very interesting, because I think what’s happening around eDiscovery, we’ve gone from a variety of different platforms out there that can do the basic functions of processing, homogenizing data, presenting it, and providing sort of a templatized environment for people to go about classifying those documents and put them into different buckets. And then that’s kind of where it stops.
What’s happening now, is we’re finding more and more capability to go in and mine that data early on. So my ideal systems have different levers that allow me to go in and apply different filters to mine out certain components of data; to display with a lot of data visualization, different identities and different components and characteristics of the data that’s in front of me; to be able to automate a lot of that initial assessment, the initial shaking out and sorting of the data set.
“On the AI front, I need a system that’s going to learn from my work, that’s going to learn from what I’ve done in one case and allow me to apply that understanding in the next case”
Being able to bring in every kind of data imaginable – I need a system where I don’t have to worry if somebody is in Slack or Teams, whether I’m going to be able to display it or not. I need a system that’s going to be able to handle all sorts of different languages. It’s going to be able to handle all sorts of different data types.
And on the AI front, I need a system that’s going to learn from my work, that’s going to learn from what I’ve done in one case and allow me to apply that understanding in the next case just as I would if I were training up an associate, where the technology really becomes an integral part of how you practice law.
David, thank you very much. David Stanton is an eDiscovery partner at Pillsbury. Thanks for joining us this week.
Our guest next week will be Kevin Clark, who is Litigation Support Manager at Thompson Knight. Thank you all very much. Again, David. Thank you.
Thanks George. Thank you very much. I appreciate the opportunity.