Document Automation + the Division of Legal Labour

What should lawyers do with their time? Just client work, or do they need to invest some of their own time in making themselves more efficient? If that’s of interest, read on.

Artificial Lawyer recently received the news that bot builder, Josef, had created a web-based no-code doc automation system that works with Word templates (see end of article). The idea is that this system will be so easy to use that it will allow lawyers themselves to create their own automated documents, rather than having to rely on others to do the technical leg work for them.

But the software development step is not the bit that most interested this site. The thing that really struck Artificial Lawyer was how this connects to the division of legal labour, and what the right balance is when it comes to what lawyers should actually be doing.

Tom Dreyfus, Josef’s CEO, says this in a recent blog post on the doc automation system:

‘If you ask a lawyer – in-house or at a law firm – what they know about document automation, you might well be met with a blank look. Some of them might have filled in a questionnaire to generate a document or two. But almost none of them have built an automated document template themselves.’

Dreyfus asks and answers his own question then, about why this is so. The short answer is that doc automation is not always simple, especially when you need coded elements and fiddly methodologies that soak up a lot of time, or have formats that don’t easily mesh with how lawyers really work, i.e. smooth integrations with Word.

But again, this is not the most fascinating part. The bigger question is why should lawyers do anything other than be lawyers?

Which in turn begs the question: what is ‘a lawyer’ and what are the outer limits of their employment duties? What should we expect ‘a lawyer’ to do, and what should we not expect them to do?

I.e. just because a lawyer can do X, should they do X? Here’s some thoughts.

The Division of Legal Labour

In many ways the above question goes to the heart of everything that is happening now in the world of legal technology.

Legal technology today is asking: can we relieve lawyers of the mundane, the process tasks, the admin work, the secretarial chores, the basic functionary tasks, the operational, the repetitive, the automatable….?

Can we allow lawyers to focus on what they are meant to be good at?

And what is that? For this site, here’s a few things that seem to be what a lawyer is meant to be good at (and the list is surely longer….) – thinking for you on matters where you are not an expert; leveraging their knowledge not just of the law on a factual basis, but adding in their actual human experience of how things work in the legal world and thereby adding value that no-one else could; taking responsibility for providing a solution to your problem and staying with you until it is resolved properly. And of course being someone you can trust; who at least can show some element of human understanding of our personal situation; and hopefully someone who also doesn’t charge you what looks to be an unnecessarily large fee or work in a highly inefficient way.

But, how is a lawyer to be this professional superhero when they are weighed down by a hundred mundane tasks, all of which take time? And – for now – time is money, hence they end up making themselves far more expensive than they need to be.

(And, of course, if lawyers can quickly and easily speed up the process of doc creation then this is one way of changing things.)

But, maybe they don’t want to be these super-efficient, yet also very human professionals? Let’s get to the heart of the matter: do lawyers enjoy being immersed in process work that soaks up their time (and indirectly their clients’ money)?

Artificial Lawyer recalls a conversation with a painter and decorator many years ago. The decorator was asked if he ever found it a bit dull, painting wall after wall in white paint – which constituted the vast majority of the work he did for clients across London? To which he replied: ‘No, I enjoy it. It’s a bit like meditation.’

In this case, the monotony of the process was seen as good, at least for this particular person. But then, the painter didn’t need to interact much with clients or engage much with anything other than the paint pot, the paint brush and the wall. His life was all very Zen.

The value in their work was literally their ability to uniformly and consistently apply paint to surfaces with a kind of robotic persistence.

Each to their own, and value is in the eye of the beholder, but still….a lawyer’s day job is more complex – or at least we’d hope so. And if not, then why on Earth is it a regulated profession that costs a fortune to qualify into?

‘And the point is…?’ you ask.

The point is that one of the key purposes of technology is to relieve humanity of labour that we’d rather not invest ourselves in. From washing machines, to spell checking software, the basic logic is the same: save time, save personal energy, do something more valuable for you and society, and hopefully whatever else you do instead will be more enjoyable too.

Very Zen…

Which leads us to something of a quandary. We want lawyers to be efficient and provide value, i.e. not waste their energy and experience doing process tasks. But, we also are not totally clear about what they personally should be doing to alleviate this inefficiency.

I.e. we want lawyers to slice off that wasted activity, but do we want them to be responsible for removing it? Or is that someone else’s job? I.e. now a law firm needs to pay someone else to do these things that reduce inefficiency, which creates cost to the business, which now must be paid for by the client.

If we take the lawyer automating their own docs example: although the lawyer can now do it, should they do it? Do we need to expect lawyers to be responsible for personally reducing their own inefficiency, or should we instead expect their employed support team, or external vendors and consultants, to do it for them?

The Specialisation of Legal Labour

A commercial law firm is a smorgasbord of different types of legal labour, ranging from hand-holding clients, to negotiation, to litigating, to legal research, to doc review, to knowledge management, to project management, to fee prediction, to business development, to public relations, to the raft of tech-related aspects of the job…..and the list goes on. Some tasks are done by lawyers, other by ‘associated professionals’, i.e. the support team.

But, they all have one end goal: to make money for the owners of the business via the sale of legal products and services. (For inhouse teams it’s different: to help the owners of the business to keep making money via the management of its ongoing legal issues. But, both revolve around making money for the business.)

And a good way for a business to lose money is to be inefficient and have people doing tasks they are not best suited to. Hence, every organisation can show you an organagram of staff, each with job descriptions that should – ideally – fit together like a jigsaw.

In reality this organagram in a law firm, i.e. the structure and distribution of legal labour and the specialised tasks within it, is not so tidy. Some lawyers are great at business development, others are a liability when in front of clients; some lawyers love to explore what tech can do, others want it ‘just to work’.

Real life is messy. Roles get blurred. People take on tasks they need not do, others get forced to take on tasks they don’t even want. When organisations perform badly one cause is that too many people are doing the wrong tasks and should be more focused on other things.

And so, where does document automation sit within all this?

You Are What You Do

As Artificial Lawyer has said before: you are what you do.

So, going back to Tom’s point above about rarely meeting lawyers who have created a doc automation template themselves, the point is perhaps: does it matter?

The answer has to be this: if creating a doc automation template can be made so easily, and be cost effective and efficient, using a certain software approach, then why not let lawyers do it?

Lawyers increasingly are their own personal assistants, largely due to technology, for example. They can zap through legal research that they might have in the past asked a team of junior associates and the law library to help with, also.

Why do they take on these tasks now, when they didn’t before? Because it’s now simply easier to do it themselves largely due to technology having removed inefficiency.

In fact, adding extra labour into the process can make it less efficient, not more efficient.

The point then is this: as long as the activity doesn’t interfere with their core role, then lawyers knocking out their own doc automation templates makes sense. But, if it gets to the point where it does impinge on their valuable time, it should go the way of all other legal labour not best suited to a lawyer. I.e. to someone else.

The challenge for any business is deciding what is the right level of ‘stretch’ within each job role. The reality is that this is always a moving target. Roles also keep evolving. The division of legal labour is in flux.


The key takeaway is this: just because you can do it, doesn’t mean you should do it – that is what the specialisation of labour across the whole of society is all about.

But, likewise, if you can do it, add value and it not impinge on your total output of value to the clients and the business, then why not do it?

Artificial Lawyer says if lawyers want to build their own doc automation templates then go for it. But, be clear that you’re driving up overall efficiency, rather than taking on tasks that will slow your organisation down. Much will depend on how long it takes….and that will mean experimentation, which opens up another front in the division of labour discussion.

(To be continued….)

Josef’s doc automation solution and the two joined problems it solves, by Tom Dreyfus, CEO:

‘Document automation involves the production of a document (or document set) that is populated with data drawn from an external source, like a database or a questionnaire. Historically, there have been two main forms of document automation.

The first is web-based document automation, where you draft a document online. The user experience is often smooth and modern. So far so good! The problem, though, is that you can’t simply upload or download a Microsoft Word document while retaining the formatting. So although web-based document automation can be used in other industries, it’s not much good for a busy lawyer whose documents need to be perfectly styled.

The second option is Word based document automation. While using a Word document as your template is a great way to protect your styling and formatting, it presents a different problem for a lawyer. To populate a Word document, the lawyer would need to learn to ‘code‘. See the example below for how complex even a very simple automation function can get. 

Here’s the problem in a nutshell: on the one hand, web-based document automation provides a great user experience but can’t handle complex Word documents; on the other, Word-based document automation offers a terrible user experience (for anyone other than a coder), but protects the formatting and styling in the source document.

This was the problem that we set out to solve at Josef. 

We wanted to build a web-based document editor that provided a beautiful user experience, without compromising the styling or formatting of a Word document, no matter how complex. An editor that allowed a lawyer to upload a Word template and apply logic or insert responses without using a single line of code. An editor that was so simple to use, that any legal professional could take the Word templates that they work with every day and transform them into dynamic, bot-powered automation tools.

In other words, we set out to take the best of web-based document automation and combine it with the best of Word-based document automation to create a beautiful document automation experience for the modern lawyer.

And that’s exactly what we did.’

You can find out more about Josef here.

1 Comment

  1. Really interesting thoughts! I see learning skills in automation and document design as a critical part of future lawyering… so whilst the first few attempts might take away some lawyer time, I think lots of lawyers need to understand the process behind automation and document design, as a new form of precedent creation… 😀

Comments are closed.