A recent case in one of my favorite fields of substantive law, Qui Tam, provides a good illustration of a court struggling with e-discovery vendor prices. United States ex rel. Becker v. Tools & Metals, Inc., No. 3:05-CV-0627-L, 2013 U.S. Dist. LEXIS 46529 (N.D. Tex. Mar. 31, 2013). District Court Judge Sam A. Lindsay in Dallas had to determine the reasonability of the prevailing plaintiff’s attorney fees and the reasonability of the charges of their vendor, UHY Advisors. UHY is a large accounting firm who has what they call an eDiscovery & Digital Forensic Team. The issue was not scope and entitlement, as you would see under a 28 USC §1920 court costs award. The plaintiff was entitled to reimbursement of costs under the substantive statute. 31 U.S.C. § 3730(d)(1)-(2) (a qui tam plaintiff “shall . . . receive an amount for reasonable expenses which the court finds to have been necessarily incurred, plus reasonable attorneys’ fees and costs.“) For this reason the primary issue here was reasonability of the costs award.
Judge Lindsay had no problem with analysis of reasonable attorneys fees, an issue with which he has long experience, but the issue of e-discovery costs reasonability was new and vexatious, much like the lawsuit itself. That may be one reason Judge Lindsay was clearly upset by all of these motions and ended his opinion with these pointed comments:
This litigation has become protracted; however, at some point all litigation must come to an end. Considerable but scarce judicial resources have been expended. … Accordingly, except for a clerical error, a matter relating to any supplemental fee request, or an intervening change in the law, the court is not inclined to discuss or write further on the issue of attorney’s fees and costs in this action. The parties are aware of their appellate options, which they should exercise if they believe that the court has ruled contrary to law, but litigation will end at this level. In this vein, if any motion is filed for reconsideration or to amend or alter the judgment, and the court finds such motion is without merit or good cause to be lacking, the court will impose monetary sanctions against the offending attorney or party. The parties are strongly warned not to test the court patience in this regard.
Qui Tam, Abraham Lincoln, and Me
Before I go into the problem of determining vendor price reasonability, and how Judge Lindsay and his magistrate Magistrate Judge Renee Harris Toliver handled it, a little background on the esoteric field of Qui Tam law is in order. The law is ancient in origin, going back a thousand years to the early days of common law in England. The first qui tam law allowed a person to turn in another for poaching game on the King’s land and get paid a bounty for the action. Over the centuries this morphed into a lawsuit where a private citizen could sue in the name of the government for any fraud against the government. If they prevailed, they would receive part of any recovery, and the government would get the rest.
This part of our common law was reinvigorated in the United States by one of my favorite lawyers, Abraham Lincoln. He had legislation passed based on the old common law that encouraged whistleblowers to come forward with knowledge of fraudulent billing of the federal government during the Civil War. There was a lot of that going around at that time (still is) and President Lincoln was desperate to try to stop it.
Today the federal and state laws implementing Qui Tam are called False Claims Acts. It is one of the most complex and bizarre ares of law in the U.S., which, I suppose, is one reason I specialized in it before going full-time e-Discovery in 2006. It is almost as complex as my other area of speciality, the dreaded field of ERISA litigation. (Beginning to detect a pattern here?)
Without spending too much more of your valuable time on Qui Tam, suffice it to say that these are very complex cases, with many factual issues involving fraud, sometimes including intentional criminal activities. These are law suits where an individual sues in the name of the government (ex rel.). The cases are filed under seal and the defendants are not even told of the law suit until after the government has a chance to investigate the allegations. The cases often remain shrouded in grand jury secrecy with related criminal investigations by U.S. Attorneys, the FBI, etc. (That is one reason I cannot speak about most of my experiences for both sides in this sometimes exciting area of the law.)
If the government thinks a private whistleblower’s case has merit, they may take all or part of it over and prosecute the case for the relator (the private plaintiff). If the government wins the relator gets a share of the recovery, typically from between 10% to 15%, plus an award of the relator’s reasonable attorneys fees and costs, which bring us back to the question of reasonable e-discovery costs in United States ex rel. Becker v. Tools & Metals, Inc.
What is a Reasonable Charge for Non-Legal e-Discovery Services?
Most courts have great difficulty determining whether e-discovery vendor charges are reasonable. After all, most law firms and litigants have the same problem. So it should come as no surprise that judges do too, most of whom have never had any experience retaining vendors before they went on the bench.
Most of my readers, however, have had plenty of experience with vendors and costs. So before I tell you what the actual charges were here, and what the judges decided, try coming up with a ball park figure yourself of what a reasonable charge would be. The information provided in the opinion is pretty sparse, but it should give you enough information for some ball park estimates. Please remember, these are all just vendor charges, and do not include any attorney fees or accounting fees; just non-legal e-discovery services.
Before you guess a number, consider this additional background. The lawsuit started on March 25, 2005 as two cases that were later consolidated into one. The cases were against multiple defendants involving complex issues related to false billings for tools used on military aircraft, including the F-22 and the F-35 fighter jets. The primary culprit, TMI, filed for bankruptcy. The CEO of TMI, Todd Loftis, was charged criminally for the scheme, found guilty, and sentenced to seven years in prison. The case continued against deeper pockets who were alleged to have recklessly supervised the inflated charges of their subcontractor, TMI.
The government intervened in several of the counts, but not all. After the primary defendants survived a motion for summary judgment, the case settled on April 2012 for Fifteen Million Eight Hundred Thousand Dollars ($15,800,000). The two relators, John Becker and Robert Spencer, received a Two Million Dollar share of that settlement. Judge Lindsay also awarded another $1,675,323.28 to them for reasonable attorney fees. He also awarded the relators’ costs, almost all of which were e-discovery related. (The intervening government did not receive a fee or cost award.)
To summarize, we know this was a very complex case involving many entities and witnesses that went on for more than seven years and ended up settling for over $18 Million. Have a guess yet as to the total e-discovery expenses? Would you assume more than ten million dollars? Five million? Less than that? Two million? Less than one million? Make a guess, and then read on for descriptions from the opinion on the e-discovery billings. See how it changes your estimate. Don’t worry, the results are anonymized. So just take your best shot, play the game, and vote on this poll.
e-Discovery Charges in Becker
Here is the first fact that should impact your analysis. Although the case went on for seven years, a majority of the costs at issue were for services performed by UHY from February 1, 2010, through May 31, 2011, just a year and a quarter. No doubt you are now adjusting your costs estimate downward a bit. Now consider these additional facts from the opinion:
- The e-discovery charges included “processing and uploading” data and the “creation of a Relativity index.” Yes, that means Relativity review software was used.
- The plaintiffs’ purchased six user accounts.
- The data was hosted by UHY from February 2010 through September 2011.
- The hosting fees were billed at a rate of $39 per GB, per Month.
- There was an additional charge to repair broken and corrupt files received from defendants and to reprocess, upload, and create a search index for the corrupt data.
- UHY used two examiners to image dozens of hard drives of TMI at TMI’s offices.
- UHY provided a project manager dedicated solely to the case.
Now what do you think UHY charged the plaintiff relators for all of these services? Ok, I know what you are thinking. How many gigabytes were involved? How many files? How many custodians? Give us at least some idea of the volume of the data so we can make an a more educated estimate; after all, you did tell us the gigabyte hosting charge of $39 per month. Sorry, apparently Judge Lindsay did not consider that information important enough to include anywhere in his 52 page opinion, which included 17 footnotes.
Now you see part of the problem? We are, however, told that dozens of hard drives were imaged. Further, a close study of the opinion, and a little reverse engineering, does show that the size of the data hosted was just over 500 GBs (19,739.30/39); 506 GBs to be exact.
So, again I ask you to formulate an educated guess as to what UHY charged its customer for all of these services? And maybe even go on to guess to yourself what amount the judge approved as reasonable? Don’t look below the poll or you will see the answer and spoil the game.
Conflicting Evidence of Reasonability of Data Processing Charges
In total UHY charged its client about $900,000 in this case, so if you guessed around a million, you were about right. I’m sure many of you guessed a much higher amount. The entire amount the vendor charged in this case was not awarded for a variety of reasons that we will get into. But first let’s break these charges down into their component parts, as Judge Lindsay did in his opinion, stating with the data processing and Relativity creation fees.
UHY charged $299,710.34 for processing, uploading and creating the Relativity database for 506 gigs of documents. UHY charged an additional $38,116.38 to repair broken or corrupt files produced by defendants and reprocess them, for a total charge for the processing and uploading of $299,710.34. That works out to about $592 per GB. Does that seem reasonable to you? Remember this was incurred in 2010 and 2011.
The vendor supported the reasonability of these charges with an affidavit by Douglas Herman, Managing Director – eDiscovery & Digital Forensics Practice at UHY Advisors. In the court’s words:
Herman states in his declaration that the fees charged by UHY are consistent with market rates and fees charged to UHY’s other customers based on the level of service it provided to Spencer over more than 28 months.
The defendants attacked the affidavit basically saying it was self-serving, that the vendor had a vested interest in the outcome of the decision. They also submitted an opposing affidavit by an expert in the field, none other than George Socha. What do you think George opined was a reasonable fee to process, upload, and create a Relativity review platform of 506 gigabytes? Let’s play one last guessing game and vote here, but again, don’t look below the poll or you will see the answer.
Here is what the court said about the Socha affidavit:
Lockheed presented affidavit testimony of George Socha, the president of a company that provides e-discovery consulting and expert services, who opined that a reasonable fee for the services provided by UHY would be $20,000.
This works out to be $39.53 per gigabyte. But did anyone guess that low? Did you think the vendor here over-charged the customer fifteen-times too much ($300,000 instead of $20,000)? Do you think that $280,000 of their $300,000 charge was unreasonable?
Apparently the affidavit and defendant’s other arguments did not convince the Magistrate Judge Renee Toliver who awarded plaintiffs’ $174,395.97 for these services, which works out to $344.66 per GB. In other words, Judge Toliver reduced the vendor’s bill from $299,710 to $174,396, a reduction of $125,314.
Judge Lindsay agreed with his Magistrate and rejected defendant’s opposition as “purely speculative” and “not supported by the evidence.” Seems a bit harsh to me considering you had affidavits from two respected experts, but here is Judge Lindsay’s holding:
Moreover, Lockheed’s contention that the actual cost of uploading and creating a search index “may have been substantially less” than the magistrate judge’s $174,395.97 estimate is purely speculative and not supported by evidence. Further, a court necessarily must rely on its measured discretion when presented with conflicting evidence. Frankly, this type of determination does not lend itself to mathematical precision; and the court finds the magistrate judge’s decision is within the bounds of reasonableness.
The Judges’ reasoning for the $125,314 haircut they gave the vendor here is quite curious. They did not challenge at all the reasonability of the vendor’s charges. Instead they reduced the cost award because they held that the plaintiffs should never have paid anything to the vendor to repair the broken or corrupt files produced by defendants. Instead, they should have asked the defendants to produce the files again, but this time have the files produced right. In other words, they attacked the necessity of fixing corrupt files, and reprocessing them, not the reasonability of the expenses. Again, this shows a fair degree of naiveté in my opinion, as there are always corrupt files in any production, and it is highly unlikely a request to defendants, including a bankrupt entity whose CEO was in jail, would have succeeded. Still, that was their reasoning.
Based on this holding it seems to me this case now provides legal authority to argue that a producing party has a duty to fix all troublesome files they produce to you. What a host of arguments those demands will make as each side blames the other on technical issues.
The judges then went on to figure that this unnecessary repair cost justified what they called an approximate one-third reduction in the vendors total processing and uploading costs. (It was actually a 42% reduction, not 33%, but perhaps I quibble.) Here is Judge Lindsay’s holding:
Based on the foregoing determination, the magistrate judge deducted the $38,116.38 allegedly incurred to repair broken or corrupt files produced by Lockheed. She also reduced the amount sought by deducting for reprocessing costs. Because Spencer’s billing records did not segregate the costs for reprocessing and uploading the data and creating a searchable index, the magistrate judge estimated that one-third of time billed by UHY was for reprocessing, one-third for was for uploading, and one-third was for creating an index and concluded that Spencer should only recover two-thirds of the time billed by UHY totaling $174,395.97 for uploading and creating an index. The magistrate therefore reduced the total amount sought ($299,710.34) by $125,314.17, which is slightly more than 30 percent of the total amount sought. Based on the court’s review of and familiarity with the record, evidence, and applicable law, and the magistrate judge’s determination that Herman’s testimony should be discounted rather than excluded, the court finds no error in the magistrate’s determination.
Reasonability of Data Hosting Charges
That takes care of the processing and uploading fees, but what about the hosting and other vendor fees? The vendor here charged $424,289 for hosting for twenty months and an additional $15,750 for six user fees. The defendants argued none of these hosting and user fees should be awarded, not because the $39.53 per gigabyte, per month, hosting fee was too high a charge, or $2,625 user fee too high, but because the plaintiffs only used five of the of the 506 GBs worth of documents so hosted and accessed. A novel argument, but the court did not find it persuasive, holding:
Lockheed contends that UHY billing Spencer $440,039 for hosting of and user access to the documents produced in the litigation is unreasonable under the circumstances because Spencer used only five of these documents during the litigation and did not notice a single deposition.
The court disagrees with Lockheed’s contention that reasonableness is determined based on the number of documents used in the litigation. Due to increasing amounts of electronic documents being created by individuals and companies and the relative ease of retrieving such documents, discovery in complex commercial cases such as this has evolved into an extensive undertaking with parties producing and culling through large quantities of electronic documents to identify key documents that will ultimately be used to establish a claim or defense. Many of the documents produced and reviewed in such cases are never used in the litigation. This, however, does not necessarily mean that the documents do not have to be reviewed by the parties for relevance by physically examining them or through the use of litigation software with searching capability to assist parties in identifying key documents. Lockheed’s objection in this regard is therefore without merit.
These are wise observations by Judge Lindsay. Once again the arguments made by defendants here were really attacking the necessity of doing e-discovery at all, not the reasonability of charges. Of course you have to look at documents to determine the truth of what happened in a dispute. This is particularly true of qui tam cases where you are dealing with fraudulent billing.
The District and Magistrate judges did, however, reduce the hosting and user access fees totaling $440,039 to $271,110.23. Once again though they did not do so by examining charges other vendors make for hosting Relativity databases. They did not find the $39.53 per gigabyte, per month, hosting fee too high, nor the $2,625 twenty-month user access fees too high. Instead they reduced the award by reducing the time period that the data should have been hosted, and cutting out one of the five user fees.
The judges reduced the time because they held the plaintiffs should have closed the database as soon as a settlement was announced in principle, and not waited seven months until the deal actually closed. They cut one of the six user fees because they said the time records submitted showed that only five people in fact ever used Relativity, so the purchase of six user licenses was unnecessary.
Plaintiff’s protests that this was all hindsight and second guessing were rejected. In fact, Judge Lindsay found plaintiff’s argument regarding the danger of closing the Relativity platform before the settlement was finalized to be rather snarky and chastised plaintiff’s counsel:
The court preliminarily notes that it does not appreciate the unprofessional and condescending tone of counsel’s declaration. Further, while Spencer states in his motion that it would have been “time consuming and costly” to remove and reload the data if settlement was not reached and his counsel states, “[t]he estimated costs of nearly $100,000 for an unloading, storing and reloading process was a large obligation to incur,” both statements are conclusory and unsupported by credible evidence.
Reasonability of Data Collection Charges
The next cost at issue was for the vendor’s ESI extraction or forensic imaging of TMI’s hard drives and related travel costs ($104,716.10). Defendants argued it was unreasonable in comparison to the amount of money they spent for similar services. Here is how Judge Lindsay summarizes the parties interesting argument on this cost:
Spencer countered that the expenses were reasonable because the services performed by UHY were different than those performed by Lockheed’s vendor. Spencer contended that UHY provided two examiners to image dozens of hard drives at a fixed cost per hard drive for a total of $61,890; the cost of the associated media was $12,750; and related travel costs were $10,206.10. In its reply brief, Lockheed contended that Spencer’s approach was to copy everything regardless of whether it was relevant, whereas Lockheed’s vendor, based on information obtained from the Government’s investigation, conducted a targeted extraction of information from the TMI hard drives, which resulted in their imaging less than a dozen documents and avoiding extraction of large quantities of irrelevant documents belonging to employees who were not involved in the fraud.
The court agree with defendants on that argument and only awarded $20,000 for copying of TMI’s hard drives:
The Court finds particularly persuasive the Lockheed Defendants’ argument that Spencer could have conducted a more targeted search for relevant documents, thereby saving considerable time and associated expense. Spencer’s course of action was not prudent under the circumstances, considering that he could have conducted depositions to determine how best to conduct more limited discovery, rather than copy all materials even those that were highly unlikely to contain relevant information.
In response to plaintiffs’ argument that “there was no one to depose because TMI was bankrupt and had been dissolved for many years and Loftis, TMI’s former president, was in federal prison,” Judge Lindsay pointed out that there are technological alternatives that make it possible to make a more targeted collection. Although I am not sure this really would have saved money here, and most experts would make forensic images when a fraudster going to prison is involved, still the $61,890 charge was very high. Judge Lindsay’s observations are interesting and should be noted and followed in most cases, if not this one:
[Plaintiff’s argument] fails to address why he was unable to conduct a targeted search of the hard drives through the use of key word searches or having UHY provide a map of the hard drives to assist in the preliminary identification of key words or areas that might contain relevant information. With the availability of technology and the capability of e-discovery vendors today in this area, the court concludes that it was unreasonable for Spencer to simply image all of the hard drives without at least first considering or attempting a more targeted and focused extraction. Also, lack of familiarity with technology in this regard is not an excuse and does not relieve parties or their attorneys of their duty to ensure that the services performed and fees charged by third party vendors are reasonable, particularly when recovery of such expenses is sought in litigation. The court therefore overrules this objection.
Although I do not think keyword search is the answer, especially at the collection phase, targeted collections is always in order, and attorneys certainly do have a duty to ensure that the services performed and fees charged by third party vendors are reasonable.
Reasonability of Project Management Charges
The vendor here added one more charge to their bill, $38,514.48 for project management costs. They were completely disallowed by the court primarily because there was no description on the bills of the services provided, nor any reason provided for why the project required a full-time project manager. With a little more thought on the vendor’s part here they probably would have received at least some award for project management services. We all know how important these services usually are.
Plaintiff’s vendor here charged a total of $880,705 for e-discovery services: $299,710.34 for data processing; $440,039 for hosting; $38,514 for project management; $61,890 for collection; and, $40,552.21 for sales tax. The court found those charges to be unreasonably high and instead awarded the plaintiffs $549,281. The arguments on reasonability seemed to have little to do with what you would expect, but instead were more like arguments concerning necessity of the services. Where expert testimony was heard on some of the charges, it seemed to have little effect on the court, possible because of the wide divergence of expert opinion, $299,710.34 versus $20,000.
Reasonability of vendor fees is a new wild-west issue and there are currently few guidelines for practitioners. If you are in the enviable position of the plaintiffs here, you would do well to retain an outside expert to assist you, and not just rely on your vendor. The court might, as in this case, tend to disbelieve their opinions as biased and self-serving.
Lawyers might also want to retain an expert to help at the outset of the case too, when negotiating prices with a vendor for a big project. Judge Lindsay is right, lawyers have a duty to ensure that the services performed and fees charged by third party vendors are reasonable. If you are a trial lawyer, and confused by the maze of vendor services and charges, get an expert to help you. Even the experts are sometimes mightily confused, thanks to vendor games, but e-discovery specialists have the background and experience to figure it out.
Electronic discovery experts can help trial lawyers to select and negotiate a better contract with vendors, and help to supervise their performance. The over-use of full forensic exams in this case is an example of where that supervision would have helped. An outside expert could have devised a more targeted collection scheme, either that, or clearly established the necessity. In my experience vendors often over-utilize forensic exams, after all they have an economic incentive to do so, and their risk mitigation arguments can be compelling. An outside expert could also have helped in the corrupt files issue in this case. He or she could have determined whether the vendor’s additional work was really necessary, or whether a few calls to the producing party could have made that unnecessary as Judge Lindsay seemed to think.
Discussions of reasonability in retaining a vendor, and for supporting an award of costs if you win, should reference what other vendors in the area customarily charge for similar services and should have per gigabyte and per file references. These should also be fair “apples to apples” comparisons. That is not easy to do, but it is possible. Finally, quality counts. Vendors are like lawyers in that regard. Do you really want to bid out a bet the company case to the cheapest professional willing to take it on?
There are lessons in this opinion for vendors too. Provide more detailed billings for any project management billings. Ask any lawyer how it is done.
Finally, a plea to vendors: make your billings simpler and easier to understand. Lawyers should not have to hire an impartial outside expert to figure them out. But in the meantime, until the industry heeds this call, smart lawyers in big cases will do exactly that.