Is your processing tool leading you down the road to malpractice?

Is your processing tool leading you down the road to malpractice?

Hypothetical question: What if the tool you were using to process documents for review was missing more than 90% of the content you were responsible for producing? Seems improbable, but just play along for the sake of argument. How would that affect the quality of your production? How would it impact your ability to conduct a "reasonable "search under Rule 26 of the Federal Rules? And who's going to shoulder the blame when an opposing party complains to the judge that, hey, wait just a second, there appears to be glaring hole in this production?

This would be a tight spot for any lawyer. Luckily, most people are oblivious to the acute inadequacies of the tools they are using, and many opposing parties are likely just as unsophisticated. Others may feign naivety for the larger good, as if to say, "We don't catch this stuff. But hey, nobody does. As long as we all keep quiet, things are good." The concept of mutually assured destruction applies.

The problem is that, as proficient technology becomes more widely available and standards for competence rise, these excuses (ignorance) and allowances (looking the other way) no longer fly.

It seems hard to believe that, in 2016, widely used eDiscovery and document processing tools are failing to capture large swaths of content -- read: potential evidence -- fed to them. But that's exactly the case.

To be sure, many tools at least flag the files they are unable to read in exception reports, those maddeningly long, inscrutable documents that identify the exact file that failed and, sometimes, the reason for the failure.

But, sometimes, large chunks of information just vanish without a trace. Evidence gone, with no record it ever existed in the first place.

Just... poof.

Consider the following real-world example, pulled from the Enron dataset.

Here is a spreadsheet titled "Gas DAily deals," which shows transactions of natural gas stocks. The file was originally included in the Enron dataset as an email attachment. Below, it appears in Logikcull after having been extracted from the email, indexed, imaged and so forth. You can click to enlarge.

Gas DAily deals.xls in Logikcull

Notice a couple things. First, the document is 484 pages long, as shown by the light blue box on the left. Second, it is formatted in the document view in a way that closely resembles the original, native file. That original document has been image and OCR'd, and all of its associated metadata has been preserved and captured so as to be made findable via text search.

Now look at the following document, processed by a popular eDiscovery vendor.


This is the same document, but, as it has been presented for review, is only 12 pages long. Not 484 pages. Twelve. In other words, the eDiscovery tool used to process the spreadsheet above captured approximately 2% of the information that was actually there.

Now, in this instance, the reason for the disparity in the presentation of the documents is fairly straightforward. Logikcull captures and extracts hidden fields. The other tool does not. But if this strikes you as a fringe case, consider that this lesser tool also has trouble processing basic file types, like Word and PowerPoint documents, as its exception report indicates. This could potentially be due to file corruption, file mismatch, the size of the document (it doesn't process Excel spreadsheets larger than 1,000 PDF pages, for example), or simply because the processing engine couldn't account for certain common file types. But in any case, about 6,000 of 40,000 total documents in the Enron corpus fed to this tool did not appear in the searchable review set.

The scary thing is that this is the industry norm, and nobody seems to care. But certainly, if a superior, affordable solution exists, attorneys have a duty to their clients and as officers of the court to find and use them. Courts have never demanded perfection in discovery, but they do demand diligence. Parties must, according to the Rules, make a "reasonable effort to assure that the client has provided all the information and documents available to him that are responsive to the discovery demand." Does taking a chance you're missing half the stuff that's actually there seem "reasonable"?

Certainly, then, using tools that fail to capture much of this information, or capture that data inaccurately, has deep, direct consequences on the ability of parties to fulfill their obligations under the law -- and, therefore, has implications for sanctions and malpractice. Such failures have cost high-profile attorneys and their clients dearly. Look no further than the closely watched Coquina vs. TD Bank case from four years ago, where a key document presented at trial by lawyers for TD turned out to be devoid of critical information appearing on the original version of the document -- which subsequently surfaced in a separate case. That mistake ended a career.

The importance of "processing" gets overlooked due to its highly technical nature and the way in which it's described in the context of the EDRM, as a chore that must be completed to get to the important stuff: search and review. But the fact is, processing is among the most essential to revealing what the evidence actually says, and the easiest to get wrong. That's why it took Logikcull years to get it right. You don't automate the 3,000+ essential processing steps overnight.

It it is simply unacceptable, given the advances in technology, that the vast majority of eDiscovery tools are routinely missing large chunks of potentially discoverable information. If you're reading this, now you know, and you owe it to yourself to vet the tools you're using, consider what they're missing and what that could mean for your company, your clients or your career.

Want to see Logikcull in action? 

Let us show you how to make Logikcull can help you save thousands in discovery.

Want to see Logikcull in action? Let's chat.

Our team of product specialists will show you how to make Logikcull work for your specific needs and help you save thousands in records requests, subpoenas, and general discovery.