Column

Quantifying the Value of Legal Information

[Download the accompanying spreadsheet here]

Quantifying the value of legal information is difficult: the most valuable commodity in a law firm is the knowledge in the minds of the people who work there, and in the written information firms produce and acquire that elucidates their work. In the event of a bankruptcy, it’s possible that the only assets left to settle debts is the art on the walls, because the value can’t be recovered from the people’s heads when they leave — I always look at the art in law firms. The value of this information is more difficult to quantify than Apple’s computers and the coffee Starbucks sells. We can, however, make some assumptions about a hypothetical law firm and extrapolate some information about the value they deliver. I have done some math about the value that can be expected from a given scenario, and I am attaching a spreadsheet, so you can play with my numbers.

In my hypothetical law firm there are 100 lawyers, and they each manage an average of 10 files per year, so the firm as a whole manages 1000 files a year. For this discussion I am assuming a civil litigation practice, as civil litigation matters have clear monetary outcomes where it is possible to win or lose. The fact that many litigation matters don’t have perfect winners or losers and that there is often a middle result can be disregarded here, as long as the averaged results match our assumed variables (which you are welcome to make whatever you want on the spreadsheet). I have purposely chosen to use a medium sized firm in this example to show that it isn’t only large national or international firms that can benefit from investing in better information practices – larger firms will have larger numbers. Feel free use whatever numbers you prefer, but the larger the number of matters you can average your results over the more accurate your results are likely to be given variability in actual outcomes.

The average value of a matter under litigation is assumed to be $1,000,000 for a win and -$1,000,000 for a loss, with 60% odds of winning and 40% odds of losing. Based on these assumptions the expected value of a single matter is $200,000. This number is derived from the possible outcomes multiplied by the probability of each and then added together. This outcome is mathematically relevant, but it may not in fact be a possible outcome at all (the expected value of a die toss is 3.5), but averaged over a larger number of iterations it becomes closer to expected outcomes. In this case the results are averaged over the 1000 files the firm’s lawyers handle annually.

In the movies, there are lawyers who “have never lost a case”, which I suppose means that they know perfectly what cases to settle. In this context those (fictional) lawyers would be said to have perfect information (also fictional). Having perfect information is still only worth so much: it doesn’t mean the firm’s lawyers would always win, rather it would negate the downside of losing, as they would always know how to best proceed. In this case there would still be 60% chances of winning, and a win would still be worth $1,000,000, but if we know we will lose we can reduce the chances of ending up with the negative result of -$1,000,000. With perfect information our expected value for the matter is $600,000. The difference between the original expected value of $200,000 and our new value of $600,000, or $400,000, is the maximum amount information could ever be worth to one of the firm’s clients in a single matter in this scenario.

Given there is no source of perfect information, what can we do?

It is possible to collect more information about a situation and improve our knowledge about what will happen. This can include things like getting an expert opinion, which is of course what the firm’s clients are already doing, looking at prior history and extrapolating, or conducting research in secondary sources of information and the primary law. This can increase confidence in predicting the outcomes of matters, thereby allowing the firm to make better recommendations. For the purposes of this model I am assuming the following probabilities:

  • Looking at the firm’s existing data, they can say that the probability of a win after they predict a win is 90%.
  • The probability that the firm will predict a win is 57%

Using Bayes theorem (which you can read more about here if you care to) this leads us to a new set of probable outcomes, and the new expected value of the matter is $457,143. This means that the value of the better predictions in value gained or losses avoided are $257,143. This is the maximum value of the information delivered to a client regarding a particular matter with additional information at this level of certainty. If information costs any more than this, the client is better off not paying for the information and playing the odds.

Now it may be possible to look at ways to improve the quality of information the law firm delivers. This could be done by doing things like hiring researchers, training people better, funding additional information products to improve research, or knowledge management, by spreading knowledge and best practices to raise the average. If a firm could increase the odds of providing correct advice by 1%, the expected value of all the matters the firm handles goes up too.

Given these assumptions, improving the certainty of information by 1% increases the expected value of clients’ matters across the firm goes up ~$11,400,000 per year.

The model of the billable hour decreases the impetus to make process improvements because in theory the firm is paid the same amount regardless of outcome, but some of this value can be expected to be retained by the law firm through mechanisms as bills being paid faster, repeat business, justification for higher rates, alternate billing, and better delegation. In alternative business models or contingency billing the connection between better information and firm profits would likely be clearer. In this instance, it seems reasonable to make the assumption that the firm would be able to retain 10% of this increased value or ~$1,140,000 per year.

Increasing revenue by $1,140,000 per year for a 1% improvement seems like a business opportunity worth exploring.

Note: For the purposes of creating a model for discussion that can fit here and for simplicity I am disregarding the effect of the impact a skilled litigator has on the likelihood of a win and the change in the probable outcome which would affect value. This can to some degree be accounted for in the spreadsheet by changing the probability of a win. I have tried to account for that by weighting my variables on the side of the firm winning, but it doesn’t greatly affect the final calculations in this model unless the likelihood of winning or losing becomes very large.

Thank you to Gregory Werker, Xavier Beauchap-Tremblay and Rachelle Bastarache for giving me feedback on this column. I appreciate your insight.

Start the discussion!

Leave a Reply

(Your email address will not be published or distributed)