LEARNING FROM CLIENTS TO IMPROVE DECISION-MAKING IN THE LEGAL INDUSTRY
Practicing lawyers and those who manage law firms make decisions and judgments with incomplete information. When clients seek legal advice to guide their business decisions, lawyers frequently respond with, “It depends.” Managing a law firm to grow profits requires many trade-offs in allocating resources, assessing risks, and making choices to implement strategy effectively. Both the business and the practice of law require assumptions and dynamic information. Both entail the use of appropriate heuristics, data, and analytics.
Faced with so many choices and much uncertainty, is there a way to improve how we make decisions? Can we foster better results if we go beyond intuition, personal experience, and cognitive biases to use the right kind of information? In law practice there are many possibilities: the lawyer’s experience; empirical information; the client’s risk profile and tolerance for complexity; the lawyer’s judgment. But are they sufficient?
Managing and allocating resources, staffing, lateral hiring, in-sourcing and outsourcing, business development, and pricing involve similar challenges. Should we benchmark against others, predict the future, or use statistical models from which to draw inferences? Have we framed a valid hypothesis to ask the right questions in using statistical models? Is there a better place to start than lots of Excel spreadsheets and expensive systems?
Look Outside the Legal Industry
Lawyers are excellent at spotting issues, identifying risk, and marshalling the power of precedent. However, the legal industry lags behind other industries in the use of rigorous methods from the world of empirical analysis, statistical modelling, econometrics and data-driven decision-making. These are not merely technology solutions; they are a different way of solving problems in uncertain and evolving situations.
Experience in other industries provides guidance for the legal profession. Artificial intelligence and machine-learning methodologies have transformed decision-making practices and improved shareholder value for companies in consulting, health care, food manufacturing, biosciences, automotive and, agriculture – to name a few industries – and they are available to the legal industry as well. Predictive models assist in answering questions that impact legal and business risk and value. In the legal industry we refer to this as “business intelligence.” Consider the following “use cases” in the business and practice of law.
Just as the results of R&D efforts create value in many industries, lawyers are the most valuable asset for a law firm. Is there a way to predict success beyond the current methods? There are many choices to be considered in making decisions about who should be hired, at what level (partner or associate), and when. Once a hire is made, what kind of integration efforts does the firm make to support laterals? Are there programs to foster skills development in critical substantive, business-generation and other core areas? What impact do such factors as office location, virtual work and demographics have on success? These and other questions are asked every day; but they may all be the wrong questions because they reflect the biases of the firm’s managers. Machine-learning models can add value to the law firm by getting a better return on its investment in its principal asset – its lawyers.
What is the best mix of lawyers and other professionals for staffing a matter? Does the type of matter, location, client, existence of a client team or relationship partner, or the amount at stake make a difference?
Staffing decisions are often based on who has the right skills, whether they are available and whether they meet the client’s expectations about the profile of a team. Although these qualitative choices are important, do they necessarily create the best outcome in levels of client satisfaction? If a staffing model is adjusted, what are the trade-offs in quality, risk and profitability, and pricing structures?
Pricing Matters and Portfolios of Matters
The professional-services consulting industry has developed predictive pricing models that are relevant to the legal industry. The sources of the data may be different, but both industries have fragmented, incomplete systems as well as qualitative data sources that are capable of analysis. A model starts with historical information, and once put into service the model updates continuously. By learning from ongoing experience, the model adjusts and offers insights to provide real-time guidance.
Most cases settle or there is a payout after trial and appeal. Lawyers and their clients face the challenge of assessing whether there is an optimal amount of legal work and associated fees and expenses. Is there a certain point at which further work on a matter will not produce a better outcome? Are there certain cases that should be handled differently because of their attributes?
Artificial-intelligence-based agents are able to operate on data from both a client’s docket and reported cases (whether they are court documents or announcements in the media) to provide guidance to lawyers about pursuing or settling a case. These AI agents improve over time, based on newly-arriving data and past experiences. A client could look at all its cases of a certain type (all products cases; all employment cases) to make decisions about when to settle a case. In the case of contracts with suppliers, the analysis could expand to assess whether there are attributes that indicate potential disputes (e.g., medical conditions) and whether business people involved in the arrangement might be counselled to conduct themselves in a manner that mitigates risk.
Clients are harnessing available data to make more effective decisions and improve their financial results. Typically the data is unstructured, incomplete, and pulled from many different systems within the company. Publicly-available data may also be used to enrich the analysis. The legal industry is well positioned to learn from the experience of its clients.
When there is data, even when there are gaps and the data is “dirty,” there is information to be found. Using well-established tools from mathematics and combining it with extraordinary increases in computing power, companies are already achieving and maintaining a competitive advantage. The lawyers who take advantage of these techniques will be better positioned to achieve a sustainable competitive advantage and add value to the clients they serve.