The CLD Playbook: Drawing Time And Value From Negotiated Terms

Machine learning for contract negotiations? It's a serious idea that can save money in the long run.

Last month’s article discussed how, by shortening the negotiation cycle around contracts, the corporate legal department (CLD) could positively impact revenue for the business. The question remains, “How does one shorten the deal cycle?” I suggested tracking contract requests, developing metrics, and analyzing performance to identify “hot spots” for remediation.

This is exactly the approach taken by your technology providers when trying to improve performance of a website or software application. Think about a website like Amazon for instance. As you browse for items, the website serves up a list of products and other suggestions based on your location, past purchases, and other details. For any given page on Amazon’s site, there are multiple requests that are sent behind the scenes to “fetch” data from the Amazon servers. The servers then take some time to process each request and return the data back to the browser.

When looking at performance for the website, an engineer would examine several items:

  • The number of requests (or round trips) to the server
  • Average time to process each request by the server
  • Average time to send data requests and receive data back from the server (roundtrip time)

Through the analysis, one may find differing root causes for performance issues. If the servers are slow to respond, the solution would typically be to add more (or higher powered) servers. On the other hand, if the network is unreliable or slow, the solution would be to minimize the number of requests, as there is a certain degree of overhead with each request.

To bring this back to corporate counsel and contract negotiation, try substituting the following:

  • Server = Corporate counsel
  • Roundtrip time = Elapsed time to schedule and hold a meeting with corporate counsel(s)

Sponsored

Now we have a framework by which to “performance tune” the contract process. If we further assume the hypothesis that corporate counsel is oversubscribed, we can quickly reach the conclusion that speeding up contract performance can be achieved by reducing the number of roundtrips — thus alleviating or minimizing the delays in getting two sets of counsel scheduled and moving them to agreement.

One way to reduce the number of roundtrips is by eliminating negotiation over contract terms that do not add value. In order to identify which negotiations don’t add value, one needs only to look at the body of contracts negotiated over the past one or two years. This is where machine learning can add significant value.

Contract terms are a prime example of unstructured text that can be turned into actionable data. A machine learning tool can break down each contract by clause so that the clauses can be analyzed for variability. The tool could also be trained to take this information one step further and identify trends in how certain terms changed by the time the final deal was signed.

This may still seem a bit abstract, so let’s look at a more concrete example using a Limitation of Liability clause. I’ve included a (simple and redacted) clause for reference purposes:

EACH PARTY’S TOTAL AGGREGATE LIABILITY UNDER THIS AGREEMENT FOR ANY COST, DAMAGE, LIABILITY, CLAIM OR EXPENSE OF ANY NATURE WHATSOEVER (INCLUDING ATTORNEYS’ FEES), SHALL NOT EXCEED TWENTY FIVE THOUSAND DOLLARS. 

Sponsored

For this language, the contract caps liability at $25K. Let’s assume that, in the first draft before negotiation begins, your CLD’s contracts always start with this number. After pulling data with a machine learning tool, you might find that 75 percent of the deals that your CLD signs contain different language, and they never actually settle on a deal that contains that original $25K. Further analysis of the data could reveal that the final liability language usually takes the form of one or two different variations — perhaps capping at the actual fees paid over the past 12 months, or some multiple thereof.

With this information at hand, you have determined two things:

  • The original language that you would consider your “standard” for a Limitation of Liability term is not in fact standard — it’s actually an exception to the rule.
  • You have information around what the one or two versions of that language — i.e. the variations that end up in the signed contact — really look like.

This can serve as the basis for developing a playbook for your CLD’s negotiated terms. Collecting information about your terms can help you identify the ones that are negotiated the most. By equipping your stakeholders with language that is much closer to the final versions of those terms, you can give your business the ability to use those versions without having to go back and talk to legal again — drastically reducing the amount of time and latency associated with negotiation.

This analysis can be extended across terms to identify which ones are most frequently negotiated away from the standard form and provide remediation paths for each — either by changing the form itself or providing a playbook for alternative language. Once you have worked out a system for collecting this information on your contracts and have developed a playbook for hotly negotiated terms, the firm’s matter management system can track time savings on contracts to see how the system working and look for more ways to optimize from there.

Leveraging machine learning to reduce time spent on tedious tasks presents an opportunity to give value back not only to counsel, but also to other stakeholders within an organization. As resource optimization continues to increase as a focus within the CLD, I believe that we will continue to see more organizations that can successfully leverage machine learning tools to drive efficiency and deliver tangible value.


May Goren Photography

Dean Sonderegger is Vice President & General Manager, Legal Markets and Innovation at Wolters Kluwer Legal & Regulatory U.S., a leading provider of information, business intelligence, regulatory and legal workflow solutions. Dean has more than two decades of experience at the cutting edge of technology across industries. He can be reached at Dean.Sonderegger@wolterskluwer.com.

CRM Banner