Artificial Lawyer recently published an article by Karl Chapman, the CEO of Kim Technologies. The Democratization of Document Automation: We have Reached the Tipping Point.

I'm sure that Kim is an excellent piece of technology, and anything that can make it easier to create automated templates at lower cost is a good thing. But this article significantly oversimplifies the work required to get good outcomes from an automation project.

The core premise of the article, and Kim's approach to document automation, is that your existing MS word documents "are your requirements, and there is no need to do a template project or transformation program. "You can forget lengthy 'requirements gathering' projects that focus on requirements rather than on what business problem you are trying to solve. Just use the documents you already have!". "These documents are the way you do business today."

"In the words of our CTO – It's like a top restaurant being able to sell its cuisine at fast food prices!"

Sounds great, right? Imagine if it was that simple!

It's not our experience though.

It's been said so often that I doubted it needs repeating, but it seems it does.

If you automate a bad process, you get an automated bad process.

Most existing business processes are bad processes.

Processes need to be improved before they're automated. You're missing out on most of the benefits if you don't.

The article does say to focus on what business problem you are trying to solve, but you won't find that from looking at individual documents for your requirements.

If you want to get successful outcomes, you start your automation project by thinking carefully about what business outcomes you want to achieve and what you need to do differently to get better results than you are getting today.  

Technology will enable you to work differently and create different outcomes, not just do the same things faster.

With clear outcomes in focus, you can work backwards from there to design your new process, and the role that people will play, accordingly. Remember, it's about people, process and technology, not just technology. 

Any documents should be written to reflect the role you need them to play.

In many cases, you'll find that you no longer need a document at all because you can automate the process (e.g. an approval or a conflict of interest declaration) instead.

In other cases, when you look at each document, you can see you really have several different versions that could, with a little effort, be consolidated into one with standard options.

Sometimes you'll see that there are actually many documents throughout the whole process that, when viewed together, could be made into a package of documents that consistently use the same data so they can be created without needing to enter information more than once.

You should also look at what people are doing in Outlook and Excel, which often holds the process together and needs to be improved at the same time to get the best overall outcomes.

Before you jump to your solution, you need to spend the time to speak to your colleagues and build a consensus about what they need and will require before they'll change the way they work. At best, you're missing the opportunity to improve things for them too. At worst, they'll just refuse to use it.

If you just automate what you already have and don't think about it, you could actually make things worse. You could overwhelm other teams in your own organisation by sending them poorly thought out documents through a fire hose, or even worse, do the same to your customers and suppliers - Who will then come back sucking up time asking for clarification or expensive negotiations. Or, even more expensively, they'll misunderstand what you wanted and deliver something that doesn't achieve your goals.

Imagine if anyone can create automated templates with no training required to get other people to work how the template author wants them to, even though the template author hasn't really thought about it. It's like letting everyone create their own SharePoint libraries and Teams sites. How well does that work?

The article says "What if the same tool takes, with only a few exceptions, your existing formatting and layout of your documents and automatically replicates it. Same with your brand guidelines and layout. With very few exceptions, the integrity of the document styles and format remains intact."

Sounds great, right? But rarely do we ever come across any documents with high quality, consistent document styles and formatting. They've often been nicked from another organisation (easily visible in the document properties), and the styles corrupted through years of copy and pasting and "save as".

Before we automate documents, we often have to put the content into a brand new document template to ensure that only the correct styles are available and used consistently.

If you just automate an old template full of style issues, you'll guarantee that everyone who ever creates that document again will suffer the same problems.

Again, I'm sure that Kim is a fantastic piece of technology. And I'm sure that if Kim's customers think about the outcomes they want to achieve and then design their documents and overall process to ensure they achieve those goals, it will deliver excellent results. But in my experience, it's not as simple as suggested. As my old teacher said, "The only place' success' comes before 'work' is in the dictionary."