The Role of Machine Learning in Contract Analysis

If you think contracts are long and unwieldy, here’s proof: according to an article in the New York Times, the average length of computer software contracts is 74,000-plus words, or, as the professor who combed through all the agreements to calculate that average pointed out, “basically the length of the first Harry Potter book.”

While Harry Potter might make for some fun reading, anyone who manages contracts for a living will likely tell you the pages and pages of fine print involved in most corporate agreements are decidedly not. And analyzing these lengthy volumes has historically been a cumbersome, manual task that – due to the potential implications of details as small as a comma – can take hour after painstaking hour. But contract analysis is hugely important work: this close scrutiny helps businesses to strengthen standard contracts, reducing their risks and improving potential upsides, and helps businesses to evaluate the terms of a specific contract, understanding exactly what they are obliging themselves to do and what they will get in return.

The Downside to Manual Contract Analysis

The greatest challenge that arises with manual contract analysis is the time factor. Combing through huge volumes of text can take hours, if not days. For long, complicated agreements, the time and resources involved in doing so can net out to costs in the hundreds of thousands of dollars, according to some estimates.

When contract reviews and analyses hold up new agreements due to staffing bottlenecks and other delays, the result can be slowed revenues and profits, and other problems that ultimately erode the bottom line. The detailed nature of contract analysis can also be vulnerable to human error and bias, meaning the results of an analysis may not provide their expected business value.

How Machine Learning Can Help Improve Contract Analysis

The rise of artificial intelligence offers tremendous potential for contract management and analysis. While the terms artificial intelligence (or AI) and machine learning are often conflated, they’re not quite the same thing – and so in order to understand the role of machine learning in contract analysis, it’s important to first understand what the two concepts mean and how they differ.

Artificial intelligence refers to the full spectrum of intelligent machines and machine functions that are capable of performing tasks usually done by humans. This includes technologies such as voice recognition, process automation, and voice-activated smart home controls. Machine learning, meanwhile, refers to a computer’s ability to use algorithms and models to learn new functions. It can also work together with other AI capabilities, helping the machine to learn from each new result it encounters and improve its own performance – rather than relying on code updates to progress. And therein lies the most immediate opportunities for employing machine learning for contract analysis, as well as some of the greatest potential for the future.

If you think about the various stages of contract lifecycle management, some of the opportunities for machine learning start to become clear:

Contract Drafting and Negotiation

The contract drafting and negotiation stage of the contract lifecycle relies heavily on the negotiators’ sophisticated understanding of the full risks and opportunities that exist within a given contract, and their ability to craft a document and come to an agreement that optimizes the chance of a successful relationship. Having an ongoing contract analysis process in place is key because it allows your business to learn from previous wins (and less successful outcomes) in order to refine standard agreements and better address outliers based on actual experience.

In the future, machine learning will be able to assist in this respect – perhaps even automating the creation of some contracts – by leveraging the results of previous contracts in order to optimize contract language and terms. As a machine learning system ingests more contracts and analyzes more results, it will be able to improve its output, creating stronger contracts over time – minus the lift that would be required to conduct ongoing, systematic manual contract analyses. Furthermore, in reviewing contracts prior to signing, machine learning may also help businesses to more effectively spot risks and red flags so that they can be mitigated before the deal is sealed.

Storage and Organization

You can’t conduct effective contract analysis if you don’t know where your contracts are (or if you know where they are stored but don’t know what types of agreements they are or what they contain). Keeping your contracts organized is important for all aspects of contract management, including contract analysis. But entering every new contract into a database is another hefty manual task because it requires you to read through each agreement in order to tag it appropriately. Automation technologies such as CLM software are already capable of crawling documents and tagging them based on their contents, but even these results can often use some refining. Machine learning, however, can improve upon the results of automated tagging so that categorization becomes more accurate and sophisticated over time, enabling better analyses.

Post-Contract Analysis

Contract management is cyclical in that the ultimate performance of a given contract should inform similar agreements in the future. This is why it is important to spend some time reviewing results before a contract is archived at the end of its term. Machine learning offers the potential to extract better insights from these reviews, and then to make suggestions that will help drive greater benefits from future agreements.