How To Save At Least One Million Dollars in eDiscovery Costs
Five Guiding Principles For Finding The Hidden Money In The Mattress
By Adam Chardukian
Over the last five years, the eDiscovery industry has experienced dramatic disruptive change. One of the biggest changes has to do with client expectations. eDiscovery costs have skyrocketed for a variety of reasons. But clients seem to believe that technology, not people, now do most of this work. They expect to pay less because they seem to believe that AI and technology systems have made eDiscovery cheaper. This has produced substantial downward pressure on eDiscovery hosting and processing fees.
But the reality is quite different. At a time when clients expect to pay less due to technology advancements, the costs associated with owning and managing eDiscovery environments are actually going up—substantially. People, process and technology expenses increase every year. This has led to a gap that can feel insurmountable. What do you do when clients expect you to charge less but your costs are going up? My answer? Sharpen your pencil. In this article, I’d like to present five guiding principles for cutting at least one million dollars in eDiscovery expenses—without sacrificing performance. These principles arise directly from what we’ve learned while conducting hundreds of audits of eDiscovery environments worldwide. If you want to save at least one million dollars, and possibly a lot more than this, these ideas could be very useful.
Thousands of organizations worldwide now own and manage eDiscovery environments. These entities are of several types:
- Law firms. Many of these organizations are in the business of adjudicating client matters soup to nuts. eDiscovery is an essential part of helping produce the most favorable outcomes for clients. But eDiscovery is one of several types of professional services they offer. Most law firms do not have a core competency in managing eDiscovery environments, although they are experts in using the systems to find responsive documents and advise clients.
- Service providers. These organizations specialize in providing eDiscovery services to a range of client types: law firms, corporate clients, government regulators and others. They are often exclusively in the eDiscovery business. Their competencies in managing eDiscovery environments run the gamut.
- Major accounting firms (Big 4). These professional service providers offer eDiscovery services to clients because of their trusted advisor status. Over the last several years, many major accounting and consulting firms have begun to offer eDiscovery services because their clients already trust them to navigate sensitive matters. In nearly every instance where we’ve conducted audits, these companies have a deep bench of Forensics & Discovery experts. However they typically leave the management of their eDiscovery environments to their Global IT teams or, they have created a “Shadow IT” department out of pure necessity.
- Government regulators. These entities bear the responsibility to enforce legislation, investigate infractions and hold accountable organizations who violate regulations. Regulators operate with a mandate to protect the interests of the general public. eDiscovery empowers them to fulfill this mission. Regulators often lack the skills to manage eDiscovery environments effectively in-house.
- Corporations. Some corporations, particularly those with a history of litigation events, have chosen to bring eDiscovery partially or fully in-house. This allows them to control costs and to keep their private data more effectively under their purview. Based on the results of our audits, I would estimate that most corporations lack the experience to manage eDiscovery environments cost-effectively.
All these types of organizations have likely witnessed the rising costs of owning and managing eDiscovery environments. What’s contributing to these rising costs? Here are a few of the major drivers:
- Technology obsolescence. eDiscovery environments are incredibly technology intensive: servers, desktops and laptops, storage systems, disaster recovery systems and networks. These systems are often pressed into service 24 hours a day, 7 days a week for years on end. Because of this hard use, they wear out much faster, in many instances, than comparable technology systems in other parts of an operation. Many owners of eDiscovery environments now expect to refresh their primary systems every three years instead of every five-to-seven years.
- Proliferation of ESI. Electronically stored information (ESI) has grown exponentially over the last several years. Have you heard of zettabytes? They are equal to a trillion gigabytes. Some analysts predict that the 64 zettabytes of data generated in 2020 will escalate to 180 zettabytes by 2025. Virtually all of this data is subject to review on eDiscovery platforms. Data-set size is now a major driver in eDiscovery events, driving up costs.
- Proliferation of device types. Along with growth in data, we’ve also witnessed a tremendous growth in the number and variety of devices producing data. iPhones were introduced in 2007 and smartphone technology suddenly became affordable for the masses. Every year these devices have become more powerful, allowing them to generate more data. Add to this the exponential adoption of the Internet of Things (IoT), which some analysts project to be more than 20 billion devices. Most of these device types not only produce data, they also can store it. Device types are also a driver of eDiscovery costs.
- Technology advancements. Because of these technology advancements, eDiscovery application developers have felt the pressure to continually upgrade and improve their software. Each update has a potential domino effect, necessitating upgrades to other parts of the eDiscovery infrastructure, like OS versions for all virtual servers for example. Staying on the cutting edge with eDiscovery applications can be an incredibly expensive proposition.
- Human costs. Because eDiscovery environments have become more complex, the human costs associated with architecting, deploying and maintaining these systems have also gone up. We estimate that a typical (if there is any such thing) eDiscovery environment for the types of organizations I noted above require seven highly-paid professionals, at a minimum. Larger organizations with round-the-clock requirements can expect to employ many times that number. It is true that many companies find ways to do more with less. But doing so does not allow them to stay on the cutting edge, putting them at a potential disadvantage to competitors.
So if you were wondering why that eDiscovery line item on your budget keeps growing every year, there’s your answer, at least in part. But the question remains: what do you do about it? I have some good news here. My organization has conducted hundreds of eDiscovery assessments and we consistently find ways to cut costs. Not one of our suggestions degrade performance, puts you at security risks or defers necessary upgrades. These ideas are about working smarter and practicing good stewardship of our clients’ resources. Here are five guiding principles:
- Leverage automation.
- Practice judicious outsourcing.
- Rationalize your operating environment against your actual needs.
- Improve data lifecycle management.
- Document application and environment performance benchmarks.
If you’re wondering just how much money these strategies can actually save you, please allow me to relate a quick story. We recently applied these principles for a global consulting organization. Our audit uncovered numerous areas where they could apply these principles. The net-result for them is that their annual eDiscovery operating costs dropped from $20 million to around $8 million. Will every organization that applies these principles realize a 60% reduction in costs? That’s hard to say. But I think it’s important to note that our average “found money” per company is over one million dollars. For many clients, the savings are far greater.
There are several workflows and operations that can be automated within most eDiscovery environments. This can yield substantial savings, while increasing your productivity and billable activity. For instance, have you heard of Rampiva? This company provides a workflow automation tool that can both speed up Processing—allowing you to begin Review much faster—and reduce the number of people dedicated to Processing, or conversely increase the number of projects and volume of data your processing team can support in a given timeframe.
For example, we recently worked closely with a client to implement Rampiva. This organization had 12 full-time (well-paid) people handling Processing on a consistent basis. After the implementation, they only needed four people to handle routine Processing. This allowed their people to focus on complex jobs and work on higher-level and usually billable tasks, which empowered the client to earn more revenue from existing staff. This initiative also impacted productivity. Whereas before the implementation, only 20% of cases were available for review within 24 hours, after the implementation nearly 80% of cases were ready for review within 24 hours.
We estimate that this client saved at least $500,000 annually. While there is a cost to deploying the software, the savings that can be realized over several quarters far outweigh the up-front costs. What might this look like for your organization? That all depends on the number of matters you are handling and the staff you’ve dedicated to Processing today. But let me give you a fairly common scenario. Let’s assume:
- You have three people completing various tasks including staging of data-sets, project setup, initiation and monitoring.
- You have one or more eDiscovery application experts involved in QA of the processed data-sets.
- You have a governance expert overseeing the operations to ensure compliance to standard operating procedures.
Let’s also assume the cost of these five highly paid employees is somewhere around $750,000. Automation could help you reduce the costs associated with Processing down to around $250,000. With an automation tool like Rampiva, you could quite possibly re-assign three of these five people to more profitable tasks, rather than trying to hire additional staff in today’s tight market for eDiscovery talent. Meanwhile, you get happier clients. Because Rampiva breaks up your data processing into smaller units, processed data is available on review platforms much faster than the traditional manual way.
The human costs of your operations might not look exactly like this. You may have a lot more or a lot fewer people involved in data Processing and staging. But my point still holds. If you haven’t yet figured out how automation can improve your operating environment, I would advise you to carefully think about this. In my experience, automation can cut aggregate Processing costs by 60%-70%. It can also free up staff to work on projects that can make you money or expand your service offering portfolio.
There’s yet another reason to consider automation—better ROI on your eDiscovery application licensing fees. Here’s what I mean. Many organizations pay an annual license fee for their primary eDiscovery application. But not all of these companies get value out of that license 24 hours a day. The reason that many companies don’t utilize their licenses around the clock is because they operate on traditional business hours because staffing across three shifts is cost-prohibitive.
However, if an automation tool can be working on your behalf, without the need for human intervention, you are getting better value for the dollar. Your eDiscovery license can be working for you while you’re asleep. This can produce a lot of benefits including hitting deadlines with greater ease, lower pressure on staff during working hours and better utilization of resources. That’s the power of automation.
eDiscovery is an industry where there are very specific and unique skillsets required to architect, deploy and maintain environments. But the need for those skillsets is rarely full-time. We find that the skillsets needed to keep eDiscovery environments operating at peak efficiency include:
- eDiscovery application experts. These applications have become increasingly more complex. The newest versions of the software often come with some great new features. But getting the most out of these applications requires expert guidance from people who really understand them.
- Data architecture experts. Data architects handle a wide range of requirements for managing data effectively. This includes translating business requirements into database and data repository specs, defining data frameworks and best practices for data handling across the EDRM, designing reference architectures and data flows and collaborating with all stakeholders for data-intensive processes.
- Workflow experts. All eDiscovery operations follow a phased and sequential workflow. The ability to hit hard deadlines is often a function of precise and effective workflows. These experts design standard operating procedures that expedite productivity, while also understanding how certain workflow operations will affect the underlying platform.
- SQL DBAs. These experts enact the policies of data architects and solve real-time performance problems, including load balancing, running database integrity checks, clearing database logs, preparing data sets for Processing, migrating large volumes of data and unclogging data “pipes” in your environment.
- Networking experts. These professionals ensure that systems are fully connected and that speed, uptime and configuration policies are closely observed. When networks go down, so do review platforms. They are also experts in internet service providers to ensure adequate bandwidth and performance based on environment load.
- Security experts. These individuals protect the integrity of data on review platforms and throughout the EDRM lifecycle. They are experts in understanding where an organization is at-risk of breaches and they consistently train and monitor staff to ensure data integrity. They are responsible for setting and maintaining security architecture standards, encryption methodologies, perimeter controls, and IAM policies. As a differentiator, these security experts must inherently understand how the implementation of standard security policies and controls can specifically impact the performance of an eDiscovery environment.
- IT experts. Virtualized systems have become the norm for eDiscovery environments. This means IT professionals must now demonstrate a mastery of how to do more with less, how to tune and manage these systems for optimum up-time and how to prevent catastrophic system failures.
All of these skillsets are available to “rent” versus to “own.” After conducting hundreds of audits of eDiscovery environments, I’ve come to believe that there are much better economies of scale available to renters than owners. Why would you own, as a full-time employee, a security expert, data architect, application expert or the like when you really only derive benefit from their work a few days out of the month?
Most people I’ve spoken to answer that question this way. “I want full-time employees because I never know when I might need them and when I need them, I need them right now.” I understand this mindset. I believe it’s more about peace of mind, believing you have talent standing by and ready, than it is about the reality of the performance of these people. I say that for two reasons:
- These professionals will get sick, need vacations, have family emergencies and the like. This means that they, in fact, may not be available when you actually need them.
- These professionals, as much as they have training, are probably limited in their problem-solving skills to the experiences of working for just a handful of companies. This means they may not be performing for you more effectively than other professionals who solve those same problems for hundreds of organizations.
Let me give you a comparator. Let’s assume that this team of seven professionals costs you 1.5 million dollars a year, all-in (salary, benefits, office space, etc.). By comparison, you can rent those same skills for about half that cost. Better yet, a great managed services provider with technology expertise in this space will give you actual availability of skillsets, not just perceived availability, by offering service level agreements (SLAs) with guaranteed response times, target resolution times and potentially even service level credits against your monthly fee for missed SLAs.
For example, a great managed services provider can guarantee a response within 15 minutes to support calls. These organization also typically have a large number of professionals on their bench who can be called in to resolve problems quickly. These professionals usually have worked with hundreds of other companies, not just a few, so they know how to solve problems quickly and thoroughly. In my experience, a great managed services provider can give you better results for far less money than you’re likely paying today.
And I want to say this for the record. I do not believe it is wise to bring in managed services providers and then fire your employees. Great employees are hard to find and it’s not a good idea to simply sack them. Instead, I recommend that you find better roles for them that allow them to make an economic impact on your business. I have lots of ideas about how to do this, but that’s not what this article is about. If you want to know more about my ideas, just reach out to me.
If you want to save at least one million dollars in eDiscovery costs, I recommend that you carefully analyze your actual technology needs. Why do I say this? In the vast majority of environments we’ve audited, we’ve found two things:
- Most environments are over-built. This happens for a variety of reasons. Sometimes, technology was purchased to provide for future, but not-yet-needed, capacity. Sometimes it’s purchased because budgets are use-it or lose-it. Some purchases are made because people are not aware of better options or because they were given erroneous recommendations.
- Most companies had never conducted a full-blown total-cost-of-ownership analysis. This means that no one can actually tell, in real dollars, how much the organization is spending annually on eDiscovery management.
How much can this exercise save you? Let me give you some examples.
- Some people seem to believe that eDiscovery works best on Solid State Drives (SSDs). These types of drives, while very fast, are also very expensive. We believe they should be reserved for I/O-intensive operations. We recommend that no more than 10-20% of your storage environment should be on SSDs. This alone can save hundreds of thousands of dollars in unnecessary expenditures.
- SQL servers are another area to closely examine. We recommend RAM over Cores. Why? Microsoft licensing is based on Cores. But due to eDiscovery workflow, which is heavy data operations versus complex SQL operations, additional Cores offer little upgrade in performance while driving up costs. In the eDiscovery realm, a 32 Core system with 512 GB RAM performs nearly the same as an 8 Core system with 512 GB RAM. But 32 Cores versus 8 Cores represents about $6,000 per month in additional and unnecessary SQL licensing expense.
- Speaking of SQL licenses, we continuously see SQL servers incorrectly licensed based on their Core & RAM allocation and the associated workload and tasks being performed. SQL Standard licenses usually cost hundreds of dollars a month while SQL Enterprise licenses cost thousands of dollars a month. In our experience, SQL Standard is more than capable of handling many jobs. We find that many organizations overpay by up to tens of thousands of dollars per month for SQL licenses.
- Disaster Recovery (DR) is another place to examine. In the majority of cases, we believe DR environments should be specified to meet approximately 25% to 50% of your production environment’s compute facilities. We often find that companies make unnecessary promises to their clients and these increase costs. For example, they build an environment to be 1:1 specified against their production, when 75% of their production environment consists of inactive matters that have not been touched for months, sometimes even years. Intelligently designing your DR environment to effectively handle the percentage of active matters and reviews at any given point in time, as opposed to the maximum possible load, could save you hundreds of thousands of dollars a year.
These are just a few areas that we routinely identify as major cost drivers. I did not dive into the selected technology stack, as this is usually a foregone decision once we are brought in. But the tech stack also has a dramatic impact on cost and can be gradually remediated as you scale your eDiscovery environment over time. For more information on this topic, I recommend our CTO’s whitepaper on The Case Against Hyper-converged Infrastructure in eDiscovery.
Data storage is expensive. Our audits show that many companies have 90+ day old case data sitting idle on costly storage units, wasting space and diminishing performance. Or they pay a service provider $5 per gigabyte per month to store it. A data lifecycle management plan can deliver huge savings, as low as $.04 per gigabyte per month. We recommend that storage plans should be tiered and based upon business requirements, with SLAs that clearly define restore policies, timelines and pricing. What might this look like?
- Premium hosting for active matters ($$)
- Long-term storage for inactive matters ($)
- Cold storage for archived/completed matters ($)
- Removal of matters that are no longer viable
How much can this actually save you? We recently conducted an audit for a client and discovered that they had 63 terabytes of data with a service provider. They were paying the provider $5 per month per gigabyte. Our analysis indicated that 20 terabytes had not been touched in years. That data could have been deleted. Instead, that company was paying $100,000 per month for valueless data and had been doing so for months on end.
eDiscovery application vendors publish performance benchmarks based on laboratory operating conditions. These benchmarks are usually derived from best-case-scenarios and everyone knows this and doesn’t expect to realize that same performance in-house. However, our assessments have shown the many clients are actually seeing performance rates that are so far below the benchmarks that something is definitely wrong. In more than one instance, we have seen actual performance at 50% to 75% of the promised benchmarks.
If you are not tracking your application performance against the benchmarks, I recommend that you begin to do so right away. I say this for three reasons:
- In some instances, application vendors can show you how to modify your practices to improve your results. But you won’t know you need to do this if you don’t document your performance.
- The preferred solution for under-performance often seems to be—throw more hardware at the problem or purchase more licensing from the application vendor themselves. That’s often an unnecessary expense. Before you buy more hardware, document your application performance and engage your application vendor.
- You can gain leverage in negotiating your licensing fees if you can show substantial differences between the benchmark and your actual performance and this can save you a lot of money.
Let me give you an example. Let’s say your vendor states that one worker can process 5GB of data per hour per worker. Let’s also assume that your actual performance is 2.5GB per hour per worker. Let’s also assume that you have contacted the application vendor and followed their counsel and yet performance has not improved. What can you do? This puts you in an excellent position to negotiate licensing fees to be in line with the promised benchmarks. Depending on your annual licensing fees, the savings from this could be measured in the hundreds of thousands of dollars.
Finally, I recommend that you benchmark application utilization and environment throughput around the clock, whether or not you run a 24×7 operation. This can give you insights into where you can increase operating leverage. How so?
- If you identify times of heavy and lighter load from reviewers, you can adjust internal Processing and administrative operations to ensure minimal impact on system responsiveness.
- This can help you track and understand individual staff productivity across shifts.
- To prevent the unnecessary purchase of expensive new hardware, you can better balance workloads 24×7 with automation tools or additional staff resourcing.
A 24×7 utilization benchmark gives you the full picture of your environment’s operating capability. Then, and only then, can you make fully informed and data-driven decisions as it relates to future investment.
CHIEF OPERATING OFFICER, GEORGE JON
Adam is the Chief Operating Officer of George Jon, Inc. As the primary author of GJ’s Strategic Roadmap, he ensures that the organization’s value propositions are aligned with market needs and are positioned for continuous growth. Adam delivers industry-specific knowledge gained over 10+ years of management consulting experience – directing multiple large, digital transformations; establishing enterprise-wide operating models; and conducting due diligence for technology platform selection – to clients around the world. His expertise is rooted in target operating model design, business strategy development, financial analysis, product management, and Lean Six Sigma implementation.
George Jon (GJ) is an eDiscovery infrastructure, product and process specialist, delivering performant, scalable, fault tolerant environments for users worldwide. GJ works with global corporations, leading law firms, government agencies, and independent resellers/hosting companies to quickly and strategically implement large-scale eDiscovery platforms, troubleshoot and perfect existing systems, and provide unprecedented 24/7 core services to ensure optimal performance and uptime.
George Jon’s (GJ) conclusions are informed by fifteen-plus years of conducting enterprise-class eDiscovery platform assessments, application implementations and infrastructure benchmark testing for a global client base. GJ has compiled extensive quantitative and qualitative insights from the research and implementation of these real-world environments, from single users to multinational corporations, and is a leading authority on eDiscovery infrastructure.