Software that finds documents based on tools other than search words

Analytical discovery software determines the relevance of documents in a collection based on the contents of the documents rather than on the presence in them of specific words or phrase. According to discovery consultant Conor Crowley in Met. Corp. Counsel, Dec. 2010 at 15, three different capabilities are offered are on the market. Some software gathers similar documents into clusters, some make binary relevance determinations, and some rank the documents they find according to an algorithm for relevance.

According to Crowley, the newer generations of analytical software use sampling and iterative learning. The software becomes more accurate as human beings outline what it should evaluate and then repeatedly assess what it finds and refine the filters, rankings, terms and other elements of the software’s functions. Once tested and taught, the software applies that learning to cull relevant documents from the entire set.

An example of such software appears in the following article, by Randall Burrows of Xerox Litigation Services. He explains that CategoriX embeds learning from linguists and statisticians to most efficiently find the right documents from the training set. The result of improving the software over repeated runs – fine-tuning what it will seek in the larger documents set – is a “ranked value of where all those documents stand relative to that original training set.”

Spearman’s correlation: an example from RFP proposal evaluators and their ordered ranking of law firms

The best known type of correlation is known as Pearson’s correlation, which tells how much one series of numbers varies as another series varies (See my post of Nov. 9, 2009 #3: varieties of correlation tests for various circumstances.). For example, a general counsel could calculate the correlation between months elapsed of lawsuits and total payments to law firms.

Another situation where correlations can be useful is if two or more people rank something and you want to assess the degree of correlation. For that, the Spearman correlation can treat the ordered data. As explained in Tony Cilly’s book, 50 mathematical ideas you really need to know (Quercus 2007) at 144, if several lawyers rank order a group of law firms that have submitted proposals, Spearman’s correlation coefficient describes the level of agreement among the lawyers.

Let’s say there are six lawyers and seven proposals. Each lawyer gives a 1 to the best proposal on down to a seven for the worst. Email me if you would like the formula. Meanwhile, I will walk you through it.

Take one lawyer’s rank number for a firm and subtract the other lawyer’s rank, then square the result. The sum of those seven squares is multiplied by 6 to complete the numerator. Let’s say that sum is 30. You then divide 180 (6*30) by the denominator: 6 multiplied by (6 squared minus 1) or 6 times 35 equals 210, which is .8. As the last step, subtract that from 1. At 0.14, the Spearman correlation of those two lawyers is low. Clearly they have different criteria in mind, or different weightings, or something else pushed them apart.

You can also do this with figures in different realms, like pages of patent applications compared to translation costs. To do this, you sort pages and assign ranks and sort costs and assign ranks. Then calculate the Spearman correlation for the two rankings.

Rees Morrison’s Morsels #143: posts longa, morsels breva

If only a lawyer’s sexual orientation was not even worth mentioning! The General Counsel of Constellation Energy, Charles Berardesco, is profiled in Diversity & The Bar, Sept./Oct. 2010 at 46. The piece makes much of his being gay and public about it (See my post of March 17, 2006: gay and lesbian lawyers; Jan. 14, 2007: Accenture’s survey of law firms; Aug. 4, 2007: NYC’s Office of Corporation Counsel; and April 16, 2009: the Corporate Equality Index.).

Review every word of every contract? In a profile called “Quotes to Practice By,” from the ACC Docket, Oct. 2010 at 22, we read this gem from a general counsel: “To this day, I still review every word of every contract to make sure it really needs to be there.” Surely you jest?

Deluge of articles and student notes on offshoring. Cassandra Burke Robertson, A Collaborative Model of Offshore Legal Outsourcing, Case Western School of Law Working Paper 2010-35 (Nov. 2010), cites in two footnotes a total of eight law review articles and six student notes. The 59-page paper cites many others but I did not count them. Most of the articles concern ethical issues perceived to be associated with the provision of offshore legal services. http://ssrn.com/abstract=1705505

“Paraprofessionals” as a term and some examples. A benchmark survey for technology companies used the term paraprofessional. The four types they reported on were patent agents, contract negotiators, stock plan administrators, and other paralegals. I could imagine litigation support analysts in the category as well as librarians. It is the level of member in a law department with more responsibilities than admins (secretaries), receptionists, and file clerks, and data entry staff.

More benchmark data on total legal spending as a percentage of revenue. On the website of ALM Intelligence, one of the sample pages for its survey of law departments shows median total legal spend (TLS) as a percentage of revenue for four revenue brackets. Less than $100 million in revenue was 1.9%. Between $100 million and $999 million it was 0.330% while $1 billion to $4.9 billion dropped a bit to 0.327 and above $5 billion even further to 0.148%. The trend down with size is confirmed once again.

No epiphanies give birth to law department management’s “big ideas”

“Any seemingly grand idea can be divided into an infinite series of smaller, previously known ideas.” This quote from a book by Scott Berkun, The Myths of Innovation (O’Reilly 2007) at 7, set me to thinking about innovations in law department management and the shoulders of giants they were built upon.

Convergence seemed a worthy idea to deconstruct even though many other innovations would serve as well. The wizards of Wilmington did not summon the idea out of the clouds, unparented, whole cloth and all. Far from that. The pieces they assembled were many.

Agreements, since the terms under which the Primary Law Firms (PLFs) agreed to represent DuPont had to be spelled out;

Bill review procedures, since law departments from time immemorial have looked at bills;

Concentration of spend was hardly a novel idea, although its application in the law department domain may have been a new import;

Evaluations of law firms, a tool as long known by law departments as it is fecklessly used;

Internal relationship lawyers who looked out for the primary law firms;

Matter management systems,which the department must have used to identify and classify incumbent law firms and thereafter to provide the data that let them know whether they had succeeded (a precursor to dashboards). I joined CompInfo in 1984 and it had been selling systems to law departments for several years, along with INSLAW and LawTrac, among several other competing vendors;

Outside counsel guidelines, which were either revamped or created or wrapped into the PLF agreements;

Requests for proposals, which the DuPont law department under Tom Sager presumably used at some point along the march to reduce the number of firms; and

Volume discounts had been around previously (See my post of April 9, 2009 #4: Aetna long ago obtained volume discounts from law firms.).

Beyond aggregating and carrying out familiar ideas, the incremental improvements originally added by those who fashioned the DuPont Model were, to my knowledge, (1) some interaction and expertise sharing among the PLFs, (2) massive amounts of public relations, and (3) some insistence on collecting work product from firms. As I write this I am not sure whether the LEDES standard pre-dated the DuPont initiative or indeed whether DuPont insisted that its PLFs use task-based coding.

Nothing I have written disparages the ground-breaking effort of DuPont Legal. Execution is everything, to state the truism, and to select and integrate the nine ideas listed above – and other previously exposed ideas – remains admirable.

Nothing fishy about Poisson distributions as used by law departments

Three times I have referred to a statistical function called a Poisson distribution, yet I have never explained the actual computation (See my post of Jan. 20, 2006: one of many kinds of distributions of numbers; Aug. 16, 2006: predicts likelihood of event during a given time period; and June 15, 2009: relation to queuing theory.). Nor did I mention that it is important to understand that a Poisson distribution implies randomness in the underlying events.

Here is what I learned from StatTrek http://stattrek.com/Lesson2/Poisson.aspx. I will apply it to a hypothetical, EEOC charges filed against your company each quarter. Let’s say over the past few years on average the company has prevailed in 6 of them per quarter. Further, assume that dismissal of the charge is a success and anything else is not a success and that you want to know the likelihood that in the coming quarter you will succeed on 7 charges. (Perhaps your performance bonus depends on that?)

The forbidding equation for a Poisson distribution to calculate a probability is P(x; μ) = (e-μ) (μx) / x!. In the EEOC scenario described above you would read it as “The probability that exactly 7 charges are dismissed during the next quarter where the average has been 6 per quarter is equal to 2.71828 (e is the base of the natural logarithm system, and if that is unclear to you, ignore the explanation but use the approximate value, which is raised to the negative power of 6), multiplied by 6 raised to the 7th power (the average number of dismissals per quarter multiplied by itself seven times) divided by 7 factorial (7 times 6 times 5 times 4 times 3 times 2).

The handy calculator on the StatTrek tells me that the probability is 13.8 percent that you will prevail on precisely 7 EEOC charges. You can also find out various cumulative probabilities. For example, the probability on these facts that you will prevail next quarter on more than 7 EEOC charges is 25.6 percent.

A compilation of 11 countries that offer LPO services

Cassandra Burke Robertson, A Collaborative Model of Offshore Legal Outsourcing, Case Western School of Law Working Paper 2010-35 (Nov. 2010),presents a thoughtful discussion of the topic. One point that is clear, from her paper and other sources, is that many countries besides India have seen the arrival of LPO providers.

She mentions that a PWC LPO/Outsourcing Survey published in January 2010 refers to legal outsourcing providers in Sri Lanka. Including the behemoth, India, that brings to 11 the number of countries that have been mentioned as providing low-cost law-related services to international clients (See my post of Jan. 27, 2006: Accenture and Mauritius; Nov. 27, 2007: Israeli offshoring; April 13, 2008: Kuala Lumpur, Malaysia; March 6, 2009 #3: South African LPOs; June 17, 2009: China and the Philippines; Nov. 10, 2009: New Zealand; and Feb. 9, 2010: French speaking lawyers looking to Romania and Morocco.).

If someone has a newish idea, here are ten shots likely to be unleashed by doubters

One of the myths exposed in Scott Berkun, The Myths of Innovation (O’Reilly 2010), is that people love new ideas. They don’t, especially if the idea challenges their accustomed way of thinking or their power. Berkun lists (at pg. 57, with more at 90) ten negative things often slung at those who put forth a new idea (See my post of Oct. 12, 2010: how to defend your good ideas from attack.).- I have paraphrased eight of them as if a senior staff meeting were considering offshoring legal services.

“This offshoring will never work.” “No client will want their work done in India.” “The theory may be fine but it can’t work in practice because of this, that and the other problem.” “Our clients and law firms won’t understand it.” “Getting these legal services done isn’t a problem.” “These legal services are a problem but no one cares.” “Getting these services done is a problem, our clients care, but we can solve it better another way.” “This is a solution in search of a problem.”

I will add two more: “What other Fortune 500 companies identical to ours have years of experience with offshoring these services and found it to be terrific?” And, “We already tried it, remember, and it bombed.”

Berkun observes, wisely, that emotional and psychological resistance affects opponents more than the merits (at 61): “Innovative ideas are rarely rejected on their merits; they’re rejected because of how they make people feel.”

Eight myths of innovation from Scott Berkun, illustrated through tiered discounts on billing rates

A thoughtful and useful book, Scott Berkun, The Myths of Innovation (O’Reilly 2010), should correct many of the misimpressions we hold about much-vaunted innovation. To summarize them, this post takes discounts from a law firm that increase as fees increase. Let’s apply the myths to that innovation, since it had to have happened for the first time somewhere.

It is unlikely that some mid-level lawyer leaped up, shouted “Eureka!” and announced the epiphany: “Let’s set tiered discounts.” Nor was there a smooth, progress onward and upward to that novel mandate or an established methodology followed that produced the new concept.

Other myths punctured by Berkun deflate also. Other members of the department probably disagreed with the new idea, found it objectionable, unworkable, even unthinkable. And, the idea for step discounts did not emerge from the fertile mind of a lone lawyer, toiling away in isolation until the grand conclusion was unveiled.

Lots of contending ideas tussled with the discount scheme, and perhaps (or probably), a deputy general counsel or two opposed the new-fangled idea. Meanwhile, a better idea to control costs, such as fixed fees or selection of firms, may have languished while the tiered-discount idea prevailed.

Berkun’s book has a couple of other myths, but the preceding eight impressed me as correct. More important, if we understand the misconceptions we may be able to come up with useful new ideas and turn them into actual successes.

An “attorney efficiency ratio,” with some commentary

At the recent Consero 2010 Corporate Counsel Forum, the slides of one panel included material on how to select a preferred provider network. Included in the mock analysis of several incumbent law firms that hoped to be chosen were the number of full-time equivalent attorneys that had charged time to the company’s matters and the total number of attorney’s that had charged time. The slide then showed the result of dividing FTE attorneys by charging attorneys, called the “attorney efficiency ratio.”

Hence, a firm where two FTE attorneys charged time (presumably based on some average number of hours charged in a year, such as 1,900) but a total of 24 attorneys billed to the client’s matters, the attorney efficiency ratio was given as 0.09 (9%).

The ratio has merit to the extent that it focuses on whether a firm uses a core group of attorneys to perform most of the work. But the ratio suffers to the extent that a firm draws on specialists to contribute their part to analysis and solutions of legal problems.

Seventeen software applications commonly used by law departments, in three levels by specificity of applicability

At the recent Consero 2010 Corporate Counsel Forum,a slide by a consulting firm arranged 17 software applications used by law departments. The arrangement presented a core group of four applications: matter management, document management, records management, and corporate secretary. Arrayed around them were nine more that primarily serve the internal legal staff: discovery management, intellectual property management, legal portals, risk assessment, collaboration tools, electronic billing, document assembly, and transcript management.

The outermost ring had four company-wide applications that find much use in legal departments but are not unusual to them: accounts payable, corporate intranet, Microsoft Office applications, and e-mail/calendaring. This three-level schema makes sense to me.