Migrated from eDJGroupInc.com. Author: Barry Murphy. Published: 2010-03-03 07:01:49Format, images and links may no longer function correctly. It’s funny how life always brings you back to the elementary school playground. One kid tells the others, “anything you can do, I can do better.” Well, the same is true in the eDiscovery software world. Every software vendor thinks it is that much better than the others. After years as a market analyst and a brief stint with a software vendor, I’ve almost become cynical.
The reality is that each software package has its relative strengths and weaknesses. But, marketing and sales departments are compensated by how much product they are able to move, not on how accurately they portray their offering to the market. Good marketers are able to identify hype cycles before they peak and tune messaging to follow the wave. We’ve seen the results of that with Early Case Assessment (ECA) – virtually ever vendor has some kind of ECA product or partner offering. The present hype cycle ready to take off is around software for processing.
The providers of processing software have been around for quite some time. The products, however, have had difficulty gaining mainstream traction. This is due primarily due to immaturity in the buying community – not many organizations have a great infrastructure for proactive eDiscovery. And historically, most organizations deal with eDiscovery on a matter-by-matter basis, letting outside consultants or law firms bring in their own tools for eDiscovery.
The trend to more and more in-house eDiscovery activities is on. The most recent Fulbright and Jaworski Litigation Trends Survey (the 6th Annual) found that almost half (48%) of respondents plan to in-source at least some eDiscovery activities in an effort to reduce costs. As more and more organizations take control of eDiscovery, the issue of scalability and performance will move to the forefront. No longer will processing software simply manage small matters; rather, it will need to provide insight into huge stores of enterprise information.
Quick reality check – this evolution will happen slowly. In the next 18-24 months, the purchase of eDiscovery-related software will more often than not be led by legal departments (and thus driven by the requirements of lawyers rather than technologists) The requirements of IT, however, will become more and more important. Figure 1 below denotes how this evolution will play out.
As IT becomes more involved in purchasing and managing eDiscovery tools, software providers need to message to the technology concerns. One such concern is about how processing software scales to handle large volumes of information and how quickly the software can tear through that information. Why? IT is sick of collection nightmares that tie up staff and keep them at the beck and call of legal over and over and over again. In order for IT to gain control of eDiscovery, it is necessary to have some kind of tool that allows for simple, fast collection and movement of information to legal for review (oh, and if that tool has an interface for legal to control as many activities as possible…even better).
So, as IT gains more say in the purchase of eDiscovery tools, vendors are tweaking their messages to appeal to the technologists. Seeing tag lines like “massively scalable” and “the industry’s most scalable” in reference to eDiscovery software is not uncommon. What is uncommon is having a way to compare apples to apples and truly understand which tool will be able to handle your information. Recently, vendors have started throwing down the gauntlet, making announcements about how much information their tools can process. Figure 2 below details the claims of the some of the leading tools that clients ask me about often. Note that some vendors say “processing” while others say “indexing.” There is a difference between the two and my personal opinion is that vendors are very careful when they use one word versus the other. Indexing is not complete processing (though it does make the information corpus searchable).
Recommind is one vendor for whom I could not find a specific claim, though there is a reference to “scalability to hundreds of terabytes of data and millions of users.” What this chart portrays is the difficult task that the vendors have in proving their scalability versus each other. While there is the need for the vendors to message their scale, there are few ways to prove it. It’s not like there are a ton of independently-created data sets that contain publicly available information that does not hold any potentially private or confidential data with which to practice on.
What is a buyer to do? Well, do not necessarily be impressed by fancy terms like “grid architecture.” There are varying approaches to scalability and just because a vendor is “architected for scale” doesn’t mean the product will scale in reality (I have hundreds of users that will testify to just this). First of all, specify how important scale and performance are for your situation. While eDiscovery may be a pressing priority for you, it’s possible that you will be indexing only off-hours and don’t need a tool that can index 1 TB / hour. If scale is of the highest priority, the best thing to do is demand good customer references. Some of the best “customers” of these software providers will be EDD processing houses. The EDD processors often have experience with multiple tools and large data sets. And, when talking to customer references, be sure to ask all the right questions (exception rates, speed, Q/A experience, deployment time, etc) – don’t simply ask if the customer is happy because this call is the time to drill in.
Have an experience with processing software you want to share? Fill in the form below to contact an eDiscoveryJournal editor.