The Cost of eDiscovery

As we approach the two(2) year anniversary of the December 2006 changes to the Federal Rules of Civil Procedure (FRCP), it has been interesting and somewhat educational to follow the tremendous confusion and the dramatic changes in the cost of processing Electronically Stored Information (ESI). As I have been claiming, the changes to the FRCP and the subsequent requirements to utilize new technology to meet these changes, is a classical example of a market paradigm shift causing chaos and therefore the opportunity for the free market system to run wild. And, as George Socha and Tom Gelbmann report in the 2008 Socha-Gelbmann Survey, commercial expenditures on Electronic Data Discovery (EDD) topped $2.7 billion in 2007 , up 43 percent from 2006, and that they predict that they will grow by 21 percent, 20 percent and 15 percent in 2008, 2009 and 2010 which equates to a $4.5 billion market at the end of 2010. As a point of comparison, Gartner Says "Worldwide IT Spending On Pace to Surpass $3.4 Trillion in 2008".

Standard EDD Fees
In a traditional electronic data discovery (EDD) project, costs break down along the following lines: (1) The cost of collecting data; (2) the cost of processing data from its native format into a format that can be loaded into an electronic document review system, and: (3) the cost of reviewing the data. Please note that according to the Fourth Annual Fulbright Litigation Trends Survey, document review accounted for 30 to 50 percent of all litigation costs. So, the current trend is to reduce the amount of data that has to be reviewed and to also reduce the actual review costs.

Manual Data Collection and Computer Forensics: The cost of manually collecting data and performing computer forensics is billed at an hourly rate plus expenses. In 2007, rates ranges from $350 - $750 per hour and have dropped to $250-$500 per hour or even lower depending upon the complexity of the project.

EDD Processing: EDD Processing or what is more commonly known as electronic data discovery (EDD) includes loading the data onto an EDD platform, extracting and "flattening" attachments and embedded files, converting the data to a readable format, reducing the volume of the data by culling it down based upon key word filters, de-duping and then creating a load file to enable the data to be loaded into an electronic review software. In 2007 EED processing costs were $1,500 - $2,000 per Gigabyte (GB) and have dropped to less than $1,500 per GB depending upon whether or not the clients request key word filtering, de-duplication or other specialized processing. In addition, in the last six (6) months, some service providers and vendors are charging less than $1,000 per GB for what they are calling "quick peek" EDD processing which is nothing more than flattening and converting the Electronically Stored Information (ESI) into a format that can be read by one of the electronic document review platforms.

Manual Document Review: Document Review is the "time honored" ritual of paying a lawyer or paralegal and hourly rate to literally review every document associated with a matter to determine pertinence, privilege and responsiveness. In 2007, if done manually in the United States by a licensed lawyer, hourly rates for review ranged from $250-$500 per hour or higher. However, in response to the markets demand to reduce the cost of review, many of the larger US based law firms are setting up off-shore operations in India and Malaysia to take advantage of the lower cost of labor and there are also now several third party off-shore document review organizations that are putting downward pressure on hourly fees with rates as low at $35 per hour. In addition, there are now several software solutions that contend that they can supplement human document review to even further reduce the costs.

Recently, several service providers, law firms and other interested third parties have introduced the concept of per-unit pricing for document review as opposed to the conventionally accepted billable hour. So, as with any market going through a paradigm shift, creative people will step in to fill a demand or void. Brett Burney does an excellent job of outlining some of the current trends in reducing the cost of document review in a June 23, 2008 article on the Law.com site, titled "Subdue the Costs of Document Review".

Document Review Platforms: There are several very poplar legacy client/server document review platforms such as Concordance and Summation and several newer Online Review Tools (ORT) such as iConect, RingTail, ImageDepot and Lexbe. The costs to host data on these platforms runs about $50 per Gigabyte (GB) with $100 - $200 per GB setup and load fees. However, several of the new Software-as-a-Service (SaaS) based ORT's such as ImageDepot, have recently run promotions to offer hosting and review of unlimited amounts of data for $1 per month for 5 months with a twelve (12) month commitment.

In keeping with the standard dynamics of new technology being introduced into a market in a paradigm shift, these newer ORTs are using SaaS technology that is very scalable and very inexpensive to run and can actually make an acceptable margin at these lower prices. It will be interesting to watch how the legacy vendors respond. The new lower price points will force them to either develop their own SaaS based tools, acquire one of the newer SaaS based tools or go out of business.

Solutions that Reduce the Amount of ESI
As stated, the easiest way to reduce the overall cost of EDD is to reduce the amount of data as early in the process as possible. Not surprisingly, several advanced technology vendors have emerged to fill this demand, including Clearwell Systems and Kazeon.

Clearwell Systems: Clearwell System's E-Discovery Platform, for instance, is an appliance-based system that identifies duplicate email and document attachments. It filters email according to domain, sender and receiver to further winnow the amount of data that needs to be reviewed. Beyond that, it handles reporting and hosting (in your client's data center, of course) and offers privacy controls.
Standard pricing for Clearwell Intelligence Platform is $65,000 per 100 GB of managed data. Compared with processing fees of about $1,500 per gigabyte of data, Clearwell Intelligence Platform costs clients about $650 per gigabyte to reduce the data required for review. By reducing the amount of data to be reviewed (which, again, is more expensive per gigabyte than the processing work), companies can realize significant savings -- a minimum of $200,000 per 100 GB, assuming 1% is actually reviewed.

Kazeon: Kazeon contends that they revolutionizes the way companies perform eDiscovery by using the Kazeon Information Access Platform software to intelligently discover, search & index, classify and act on electronically stored information. Kazeon provides a full spectrum of proactive and reactive eDiscovery solutions in response to litigation, information security and privacy, corporate investigations, regulatory compliance and storage consolidation requirements. The Kazeon Information Server software automates eDiscovery functions from identification, collection through processing, preservation, analysis and review for corporations, service providers, and law firms. Through the development of unique search & indexing, analysis and workflow automation technology, Kazeon has been recognized by various industry forums, including Gartner's eDiscovery MarketScope and the recent Socha-Gelbmann eDiscovery Survey 2008 where Kazeon was recognized as a Top 5 eDiscovery Software Provider. Kazeon has established partnerships with leading companies, including Fujitsu, Siemens, Google, Network Appliance, Oracle and Symantec.

Kazeon's Claim of $4.30 per GB Processing Costs Stirs Up the Industry
On May 6, 2008, in a press releases announcing that Attenex and Kazeon have formed an eDiscovery Alliance, the real news was buried in the boilerplate as Kazeon contended that they could process a Gigabyte (GB) of data for $4.30. This was actually not new news as they had previously contended in a press release titled "ESG Lab Finds Kazeon’s Information Server Delivers Fast and Cost Effective Information Access" this same financial offering in a slightly less alarming way by stating that the ESG Lab verified impressive price/performance with a single Kazeon Information Server appliance able to index 2,500 documents per dollar and a cost as low as $4,300 dollars per terabyte (which is approximately $4.19 per gigabyte).
However, even in the face of the overwhelming Blog posting response from the industry about this claim being completely misleading, Kazeon has not removed the claim from subsequent press releases and in fact has embraced the claim as the center piece of their marketing message. The best and most thoughtful response came from a posting on the eDiscovery 2.0 Blog by Kurt Leafstrand on May 6th, 2008 titled "eDiscovery Processing: You Get What You Pay For". Given the fact that the eDiscovery 2.0 Blog is written and sponsored by Clearwell Systems, a another player in the eDiscovery space, it may be a bit biased. However, Kurt did a great job of listing the various critical and required processing components that need to be considered when comparing the overall cost of Electronic Data Discovery (EDD).

Example of the Cost of an EDD Project (Using Standard Technology)
Given all of the above as a general foundation for the fees associated with an EDD project, following is an example project:
Parameters and Assumptions

1. Two (2) US based corporations
2. Two (2) Sites for ESI Collections
3. Approximately two (2) terabytes (TB) of data (all native ESI)
4. 2X expansion of ESI during the initial flattening and extraction of attachments and embedded data
5. 50% reduction during initial processing with key word filtering and de-duplication
6. Document Review required at multiple locations
7. Time from matter filing date to completion is estimated to be 3 years.
8. Onshore legal review at 300 pages per hour

Collection Costs
Collection of the data along with computer forensics work, travel expense, storage hardware and hourly rates is as follows;
Collection Fees = 10 hours x $300/hour = $3,000
Storage Hardware = $2,000
Travel Costs = $2,500
Computer Forensic Fees = 10 hours x $300/hour = $3,000

Total Collection Costs = $10,500

Initial Processing Costs (Flattening, Culling, De-duping)
4 TB of Data x $1,000/ GB = $4M (Reduced back down to 2 TB)

Hosting and Review Technology Costs
36 months of Hosting for 2 TB at $50/GB/Month = $3,600,000

Document Review
2 TB of data at 75,000 pages per GB = 150M pages
150M pages at 300 pages / hour = 500,000 hours at $150 / hour = $75M

Total Cost of Project
Collection = $10,500
Processing = $4M
Hosting = $3.6M
Review = $75M

Total Cost = $82,610,500

Alternative Approach
Obviously, $82M is cost prohibitive. So, where can we save from money? Following is an alternative approach:
Changed Assumptions

1. Use of new front-end processing technology
2. 90% reduction in data during culling and de-dupe
3. Use of off-shore review

Alternative Collection / Processing Costs
Although it may not be completely realistic, lets assume that both sites had technology to automatically gather, filter and de-dupe the data (e.g. Clearwell, Kazeon). And, there was no computer forensic work required. Therefore, the cost of processing 4 TB of data using the Kazeon model would be 4 TB x $4,300 / TB = $8,600 (reduction of data to 400 GB)

Alternative Hosing and Review Technology Costs
36 months of Hosting for 400 GB at $50/GB/Month = $720,000

Document Review
400 GB of data at 75,000 pages per GB = 30M pages
30M pages at 300 pages / hour = 100,000 hours at $40 / hour = $4M

Total Cost of Project
Collection / Processing = $8,600
Hosting = $720,000
Review = $4M

Total Cost = $4,728,600

Conclusion
ESI reduction through culling and de-dupe is essential to reduce the amount of data that has to be hosted and reviewed along with reducing the cost of actual document review is the key to reducing the overall cost of EDD processing.

Labels: Clearwell, Consulting Fees, EDD, eDiscovery, Electronic Data Discovery, FRCP, Kazeon, Pricing, Socha-Gelbmann Electronic Discovery Survey

posted by Charles Skamser @ 10:49 AM