December 2009 Blog Archive

We recently came across an article in an ARMA publication that sought to compare enterprise search with something called “e-discovery search”, an article which ostensibly came from an eDiscovery vendor…but clearly one of the countless that is not well versed in the ways of search technology.  ARMA is a highly respected organization, so any article within one of its publications deserved a review.  What quickly became apparent, however, is that the eDiscovery industry – or at least a large number of vendors within it – is still desperately seeking a clue about how search technology can and should be brought to bear when assessing, collecting, culling, reviewing and/or analyzing ESI as part of a lawsuit or investigation. 

Given that all but a few eDiscovery vendors rely on search technology OEM’d from small third parties – in virtually all cases rudimentary, keyword-only search – it is perhaps not surprising that such lack of knowledge and expertise abounds.  But given the importance of search in the eDiscovery process, we felt the inaccuracies too important not to correct.  The following are just 3 of the article’s most obvious inaccuracies/particularly erroneous statements which we will seek to correct from our perspective as a leader in the enterprise search and eDiscovery industries for most of the last decade.

Incorrect Point #1: Enterprise Search Is Far Less Demanding than “E-discovery Search”

This statement is always offered by vendors who have never operated in the realm of enterprise search, especially enterprise search catering to the world’s largest law firms, telecommunications providers and pharmaceutical companies.  When a senior researcher or law firm partner needs to find the most relevant precedent for their project, the most probative case or the most knowledgeable expert for their question, they aren’t looking for a “small number” of relevant documents, projects or people: they are looking for THE document, matter or person.  Hence the need for sophisticated search technology which can find the right person, project or document regardless of keyword, misspelling or other technical or human hurdle.  “Enterprise search” isn’t a simple Google query about a restaurant, a point which is commonly missed by those who don’t know much about enterprise search.

Incorrect Point #2: The Quantity of Queries is the Key

Another common misperception in the eDiscovery industry is that a good way to find more relevant documents is by simply doing more searches with ever-longer query strings.  To state it as simply as we can: search is the means to the end, not the end itself.  The end is finding the documents for which one is looking.  As the enterprise search industry has known for decades, keyword-only search is notoriously inaccurate; simply doing more and more keyword-only searches is like trying to push the round peg through the square hole harder – that peg still won’t go through the hole, and you still won’t find all relevant documents.  Even worse, this “blunt force” approach takes time…which costs money – lots of money – which clients do not like and law firms are coming under intense pressure to change.  A far better approach is to use more sophisticated (search) technology which can actually find documents before any queries are needed, and which can actually quantify search results before any searches are made.  As the saying goes, if one’s tool of choice is a hammer, then all problems look like nails.  Unfortunately, this outdated approach simply doesn’t work anymore in eDiscovery.

Incorrect Point #3: Attorneys Must Review ALL Documents

Perhaps nothing better represents antiquated, “linear” thinking (i.e. that linear document review works just fine) than the fallacy that attorneys must review ALL documents in a collection.  Simply put, this is dead wrong.  There is no statutory, common law or other requirement that attorneys review ALL documents in a collection – down to that last spam email, Christmas party reminder or fantasy football update.  Attorneys must simply review relevant documents as part of a document review and ensure documents aren’t incorrectly coded, e.g. with privileged documents.  But when vendors rely on a third party’s keyword-only search “technology” to find all relevant documents, as is the case with most eDiscovery vendors, it is a practical impossibility to take any approach other than traditional linear review where each and every document must be reviewed (remember: if one’s “tool” of choice is a hammer, then all problems look like nails).  And as bad as linear review is, the effect it has on outside counsel is even worse: rather than spending their time advocating on behalf of clients, the linear approach forces attorneys to weed through waves of irrelevant documents, driving up costs for unhappy clients.