Whit Andrews

[climbs out of deep hole where blogging was impossible due to unknown reason possibly involving radiation or other Neal Stephenson-inspired obstacle]

Well.

After some weeks of intensive interviews, conversations, writing, peer review and sweeping of emails from the inbox to a Special Folder, I am pleased to report that Gartner’s report on the Google Search Appliance has finally landed on Gartner.com. Writing a report like this entails assembly of the information we get from inquiry, an ongoing process of outcalls (I frequently seek recontact with people I talk to in the course of my work to find out How Things Went), and sometimes interviews. The Google Search Appliance is right now the most popular for-pay search product enterprises can buy (although Microsoft’s SharePoint family of search products is not to be ignored), and it appears in my inquiry more than any other product.

I reached out to Twitter to find people who might not ordinarily be Gartner clients to gain access to their insights, and am indebted to many such who are anonymous per their request. Also, though, Dave Carruthers, Dave Isenhower, Jey Jeyarajan at the Toronto District School Board all offered insights and let me use their names. I sent cookies if people let me; if you helped and I didn’t mention you here, speak up and I’ll send cookies and amend the entry.

The research is for our paying clients, of course, and I recommend they invoke the link and read the note if they can. In short, I’ll say this: Google’s execution is catching up to its ambition here, and in its key market of simple installation/simple use, it’s the solution to defeat. There are a lot of nuances, and no one should ever make a decision on a single sentence. Many times I recommend against it for particular reasons, just as I recommend in its favor — and no matter what, it has competitors to consider as well. Call us or ask for inquiry; I’m looking forward to talking.

It’s rare we get a chance explicitly to revisit research, but we do so here in a note we just published about trends in e-discovery as seen in the LegalTech lens.  Behind the paywall, I regret for casual readers.

If you have some experience with the Google Search Appliance, or the Mini, then I would be grateful to talk to you – or trade emails – about your experience. If you’re willing, I might blog the exchange, at least a bit – or not, if you’d prefer not. I would incorporate it (without names or tales) into a research note I am writing about the appliance.

If that’s something you’re willing to do, please reply and tell me if you’d rather talk or get my questions via email, and whether you would prefer I not mention the conversation on my blog, or if that would be OK.You can email me at Whit.Andrews@gartner.com

I’ve been trying to hash out the value of twitter. Good example is Rob Robinson’s ComplexD twitter stream, which is prolific but engaging. Easy to scan it regularly and see what’s in it; quick clear statements of value on the other side of the tiny click. Eventually one is likely to say, Hmm, Who is this masked man? And then he has a sales lead. Wonder how twitter will change.

Howdy y’all. I’m off to San Francisco in May to keynote (has such a nice ring to it as a verbified word, “keynote”) the Mark Logic User Conference 2009. Mark Logic’s all about snipping content into little bitty pieces tricked out with metadata and then putting them back together like Mister Potatohead or Spore (which are pretty much the same thing, except in Spore there’s more gnashing of teeth and negotiation with faraway creatures for empire). Hope to see you there. Hashtags at noon?

If you don’t believe content gets reduced to components, don’t look at what the New York Times is doing, because it’s important not to jostle mistaken expectations or they curdle. We’ve written a pile of research on content components.

Crouthamel: Katrina was a test for us. We had thousands of pieces of litigation, but with the decreasing storage costs and improved search tools, it makes sense to keep these in a general system. In our core business units, we archive 100% of their email.

Some documents must be held 25 years. Going to that, none of our outstanding litigation was in excess for that.

Woods: We have some very senior trial lawyers who don’t know how to turn on their computer. I think that State Farm gets thr long-term evidence management. We talk to them about how to manage problem of e-discovery. You have to be able to bring it inhouse — two years ago, no one could have predicted what would happen with subprime. You have to be flexible enough in your process so that you know how to deal with the 20% of your cases that will drive 80% of the risk.

The whole concept of search and retrieval, and what the lawyers did — you’re going to have to do what Tim did, which is use methods that you can explain. State Farm is taking HR pages and putting them in archive, because they have so many employees that they know they will be part of litigation. A reasonable process, discharges my responsibility as a lawyer.

State Farm is using a multi-matter management environment, something I’ve been hearing much this week. (CaseCentral, the keynote sponsor, is the supplier.)

There’s a lot of discussion on the screen now that would not be very interesting from a blog-reading perspective now. Instead of scribbling down what they’re saying, instead, what I would point to is that much of what’s being described in the idea of a platform for multiple cases is not new from a broad perspective. Tt is, instead, a savvy business plan in which what CaseCentral — and Inference, and Iron Mountain, and other vendors I’m sure I don’t know — to outsource rich, complex business processes into an environment where multiple (legal) enterprises are ALL outsourcing these processes.

From a vendor’s perspective, the resulting mat of interwoven processes is pure joy. (Think eBay, myFamily,com, or Facebook.) The magnetic effect of one law firm enticing another to join to share the BPO environment, and then another and another, carries the potential of a magnificent humming field of togetherly working law firms throwing off money. From the law firms’ perspective, its a massively fabulously efficient internetworked environment of collected processes.

I survived the e-marketplace hype of the late 1990s and early 2000s. (My favorite was a marketplace for green unroasted coffee beans. Gone.) The terror that I have run into is that these things will vanish — as did so many in the last boom — leaving behind a rustling heap of unrequited processes. I have NO REASON to believe that CaseCentral will suddenly vanish; they’ve been around quite a while, and the same is surely true for Iron Mountain and Inference. All these companies have cases to make for business viability.

Two key concerns, here, then: Tipping points and viability.  These meshed processes just aren’t broadly adopted yet, and the strategic value of an outsourced legal provider is not far off the strategic value of a power supplier, a fuel supplier, or a core materials supplier, which is apropos of the above-linked entry. This sort of technology opportunity also represents a fearsome business challenge and risk. Just like e-marketplaces, it’s fascinating. Ebay’s still around, as is Amazon, and as are many of the vendors from that area. (My second favorite: Shoe.net. Also gone.)

Woods: Get on top of search and understand it, because that’s where this all is going.

John Woods, Hunton Williams

Tim Crouthamel, State Farm
Head of Litigation Support

Tom Thimot, CaseCentral
President and CEO

Hedges: Nothing would change, because you would still have 2% to argue about.

Smith: You would have to sell it to a judge. You would have to convince the judge.

Hedges: You would have to convince the former me about this. I was a judge, and I would appoint my own expert. Each side would have its expert, and I would bring my own.

To the extent that you can eliminate information that is clearly not valuable early, the better off you are. Imagine you have a pyramid: You can almost always talk people out of asking for all the documents at the bottom of the pyramid.

Tombleson: The judge in a UK case got into the granularity of the search terms. And the judge suggested that the word “delay” be included. Controversial, because of the huge number of false positives. [I checked a PST I started two weeks ago -- 37 emails out of 2,400, which would be 37,000 emails out of 2,400,000, I think, right? That's a judge in favor of paying reviewers.]

Paskach: None of the indexing/retrieval engines are perfect, just as no person is perfect. Yet people seem susprised that they might miss a few. What’s the best way to deal with that? How do you help people understand it’s an art?

Hedges [excerpts/gist]: I have colleagues think you never disclose anything to the other side. I happen to think in the Sedona Cooperation Proclamation that you have to disclose facts. There is no basis in the world not to disclose facts. Having said that, if you’re doing electronic review — why choose a tech, what did you do — Victor Stanley notices that this isan iterative process. I think transparency is very important.

Tombleson: It’s so important to have an audit trail.

Paskach: What about specialized methods like Latent Semantic Indexing, Linguistic Analysis? Anyone doing that who is willing to admit it?

Smith: We are using all terms and all technologies and try to reduce and pull things. “The biggest problem is ‘there’s cake in the kitchen’ that goes to everybody.” At the end of the day, anything going out to opposing counsel, want to have somebody’s eyes on.

Paul Tombleson, KPMG

Tombleson: We see them being used for speed of review. [Describes early case assessment.] Second is where not to use search terms, effective in investigations/foreign corrupt practices. Useful where might be code words.

Hedges: Search terms are becoming an obsolete technology. Lawyers are anal retentive. It’s a loit easier to use a word. Lawyers may not understand. Look at O’Keefe. Lawyers come in and argue about search terms, and Facciola says neither of you know what you’re talking about. Your laypeople, as am I.

Smith: I don’t have a lot of clients who want to be early adopters.

Witczak: Doesn’t matter what tools you use. You need to realize that the cheaper tools may be slower. the tools that say they can get through the data that much faster. You can’t look at the tool alone — you have to look at the whole process.

Hedges: If ytou’re doing it right, you are probably upfronting some discovery costs. These are costs that you will spend eventually. Much better to spend it upfront to get an agreement.

Paul Tombleson, Lead Partner, KPMG LLP

Mark Smith, Associate, Winston & Strawn

Ron Hedges, former U.S. Magistrate Judge

Gregory Witczak, Vice President, directory of eDiscovery, Lehman Brothers Holdings

Moderated by Chris Paskach, National Partner in Charge, Forensic Technology Services, KPMG