Automating Work Requires a Surprisingly Human Touch

Kevin Pray
January 19th, 2023 

Are you ready for robots? If your boss told you tomorrow that the business had purchased software to automate a portion of what you do, do you know that setting up the automation would likely require a lot of input from you? The infrequently mentioned truth about automation, and Artificial Intelligence in general, is that it must be designed by a human brain. 

The robots are coming. Legaltech is increasingly enabling the automation of work in ways that offer the possibility of improving life for legal professionals, but anxieties around the topic abound. These fears are often centered around how automation might take work away from people, replacing them in the same way that industrial workers have been replaced on assembly lines. A wave of recent headlines around tools like ChatGPT have stoked these anxieties, and it’s clear that human replacement is the big boogeyman in the room. This is a deeper topic to explore in another article, because here we will address a much more common and mundane problem around automating what you or your office does. 

Breathe easy, the robots are not coming for you, but rather, with you. Automation will not replace people, so much as assist them in doing more with the same effort they have always contributed. To caste it in terms of manual labor, think about a device that would help you lift more, but could not itself perform the lifting, the carrying, the placing of a box in the proper place on a shelf. Artificial Intelligence does not really exist yet except, perhaps, within some of the most advanced computers or research facilities. A better term for what is often marketed as “Artificial Intelligence” is actually “Machine Learning” (ML), which is exactly what it sounds like: the machine will learn from you how to do certain things. But, and this is significant, it will only learn to do the work well if it is taught well. Conversely, automation can be taught a lot of bad habits, if it is taught poorly. 

Attitudes have changed. The fact is, despite being an industry known to innovate slowly, reluctantly, and stubbornly, legal professionals are embracing software automation at a speed which indicates that some mental adoption barrier has been overcome. Legal professionals have realized that profit margins can be increased by reducing the amount of human labor used for certain kinds of work, largely administrative tasks, which support the profit-making efforts of the business, but which are not directly tied to profit. So, the odds are increasing that you will experience work automation soon, if you haven’t already. And if you are one of the people whose work will be affected by automation, to be blunt, who will have tasks automated, a lot of your knowledge is going to be needed. If you are already using automation of some kind, we at Contract Logix have seen that you are likely unhappy with it, that you may be feeling as though it is not delivering on the promise. This is because what the automation is doing for you was not designed with enough knowledge of how the work gets done. 

Let me break down a few principles: 

1. Implementing an automation tool of any kind that improves your business takes careful thought about what work is being replaced. This critical step of change management is often overlooked, or at least not given the proper importance. The software must be told what to do and how to do it, this is the “machine learning” part. In Contract Logix software, this teaching is often done during the implementation of the software at your business, designing triggerable jobs, and sequences of tasks, called Workflows. 
2. What the software will do is what a person used to do, and the software must do it at least as well as the person who did it previously. The reason many software automation solutions fail is because the software does not do the job as well as the person who did it previously. 
3. If the system was implemented without understanding how people did the work, and specifically what was valuable about the way people did the work, then the software will never deliver a better outcome than the person could have. It may deliver a cheaper outcome, a more efficient outcome, but eventually problems will arise because something the human worker did is unaccounted for. 

Just to focus narrowly on our domain, contract automation, one of the most important parts of implementing contract lifecycle management software with automation is understanding how your business functions without it. At its most basic, we encourage you to ask yourself: 

• Who in the entire organization is a part of the process of handling even a single contract? 
• What does each person involved do to interact with the contract? What happens to cause work to come to that person, and what happens to cause them to pass on work to someone else in the organization that is related to what they did? 
• What rules, often only in that person’s head, are followed to complete the work, given various circumstances? When do they notify an attorney about something, as opposed to when they notify accounting? 

Discovering the answers to these questions involves a human-centric exploration process, to inquire around your business or department, and ask what may seem to be questions with very obvious answers. Doing so involves untangling processes that people do without thinking about them, are often taken for granted as simple, and often have never been explicitly communicated outside of a few small conversations. Does your Chief Legal Officer know what the paralegals actually do? 

But do not be afraid of this effort. Our modern workforce has handled these exact issues since the beginning of the industrial revolution, and you do not need to reinvent the wheel to figure this out. There are several patterns to how a person or a group of people cooperate on a task that were mapped out as we switched to industrial manufacturing, one or several of which will apply to how you work. For example: 

• Repetitive Workflow: this is what we think of as assembly-line work, where a person stands in one spot and does the same thing, say tightening a bolt (let’s call it “Task A”), over, and over, and over again. Other parts of the work, say tasks B, C and D, are done by someone else, but each person only does one type of task.

• Discrete Workflow: this is often how a legal assistant works. This person will do several related tasks in a row before handing off the deliverable to someone else, or before changing what tasks they do. They may do tasks A, B, and C all at their workstation before handing the work off.

• Job Shop Workflow (aka, the Artisanal Workflow): this is how a lot of attorneys work. The person doing the work is a specialist who knows how to do a large part, if not all, of the job, and must do the work themselves to be able to apply their knowledge. This is how almost everything was manufactured before modern assembly-line science.

• Process Workflow (aka, Batch Workflow): this is how a lot of paralegals and some attorneys work. The person performs a narrow range of tasks but will do all similar tasks at the same time before moving on to less similar tasks. A good example is prepping mailers to send out too many clients. You might fold all the documents and stuff them into an envelope first, several dozen times, then address all the envelopes at one time, then stamp all the envelopes at one time. In contract management, you might review all the contracts of a certain type for a certain issue, before moving on to other issues of a specific type that you are looking for in the same contracts.

People around your organization are already using these workflow styles successfully. Each person may use a different workflow that suits them best, and miscommunications about when or how something gets done often stem from a lack of understanding about how one workflow differs from another. If you are going to replace human actions with automated actions, understanding how people work in this way will make it much easier to figure out what the software needs to do to reproduce or more likely facilitate the human effort. 

In summary, automation and AI can deliver huge benefits to in-house legal teams when it comes to things like process workflows. If you want to learn more about this topic or how Contract Logix’s contract management software can help you in automating work and improve your contracting processes, please reach out to discuss your needs. 

Looking for more articles about Contract Management? Check out our previous article “8 Steps to Effective Contract Management for Insurance Companies“.

Menu