Why do the work twice?!
Why do the work twice?!
This happens more often than people might think. Teams are so used to proving that something works manually that they sometimes create an enormous amount of manual work before they allow themselves to automate it. They build the first version, then the second, then the twentieth, and only when everyone is completely exhausted do they start thinking about how the process could have been designed differently from the beginning.
I don't think the answer is to automate everything immediately. Sometimes you absolutely need to understand the manual process first. You need to know what the designer does, where the problems appear, what information changes and which decisions require a human. But once those rules become clear, continuing to perform the same actions manually just because that is how it has always been done doesn't make much sense.
This is one of the lessons we have taken from our own projects. Automation works best when you design for it from the beginning, rather than taking a very large pile of manual work and asking the computer to somehow clean it up afterwards.
The best automation is not a faster version of a bad process.
It is a better process altogether.
One of the more interesting things we have seen was a workflow where hundreds of assets were being manually created in Figma before the automation process even started. The intention was understandable, the team wanted to prepare and test everything first, but at some point we had to ask ourselves a simple question: if we already know this is going to become an automated system, why are we manually creating hundreds of examples before building the system?
This happens more often than people might think. Teams are so used to proving that something works manually that they sometimes create an enormous amount of manual work before they allow themselves to automate it. They build the first version, then the second, then the twentieth, and only when everyone is completely exhausted do they start thinking about how the process could have been designed differently from the beginning.
I don't think the answer is to automate everything immediately. Sometimes you absolutely need to understand the manual process first. You need to know what the designer does, where the problems appear, what information changes and which decisions require a human. But once those rules become clear, continuing to perform the same actions manually just because that is how it has always been done doesn't make much sense.
This is one of the lessons we have taken from our own projects. Automation works best when you design for it from the beginning, rather than taking a very large pile of manual work and asking the computer to somehow clean it up afterwards.
The best automation is not a faster version of a bad process.
It is a better process altogether.
1 Guest(s)
1 Guest(s)