27 Sep Software Automation Isn’t A Job Killer
Can software replace my job? It’s a concern with roots going back centuries to the arrival of the first automated power looms in 18th-century England, continuing into more modern times with the large claw-like robots that build automobiles. Today it seems there are hundreds of websites making their living pushing clickbait headlines describing the perils of the coming Robopocalypse of software automation replacing human workers, and I often see those fears bubbling to the surface when working with clients. Fortunately, in my experience, those fears are often overblown, and I’m not alone in thinking so. Wired’s James Surowieck agrees that we have nothing to fear.
The automation jitters usually start to come out during discovery. I talk with clients to clarify the business logic, workflow rules, or other automated features, and I’ll use the term “robot” to describe the software we’re building. After all, “if I can’t tell the robot what to do, how will it know how what to do?” It’s somewhere around here when we’re asking somebody to explain their whole job function that somebody on the client’s team will mutter, half-jokingly, “So I guess I better start looking for another job, huh?” It’s not hard to understand where that thought is coming from. After all, if we can completely strip away the human layer and turn their job into a series of rote, mechanical steps, what role is left for them? I quickly try to assuage their fears.
Hopefully, the robot will take away many of their current job functions, especially when we are talking about transitioning pure analog processes to digital. When this happens, we often see incredible gains in Return On Investment for human hours spent on a process. I think back to an accounting department we supported that printed out hundreds of invoices and time sheets every week to verify client invoices against employee time sheets. Leaving aside the costs in paper and toner and copier maintenance, 1 employee spent 20 hours a week on this task… and then a second employee had to verify it! Wasting another 20 hours! We built an exception report that accomplished the same task in fractions of a second, leaving only the outliers for the accounting employee to inspect.
Once we implemented the fix, no one lost their job, no one was demoted. In fact, the employees became more valuable to the company. They began to do more value-added work, those things that usually end up being a lower priority than validating and submitting invoices but that add more value to the company over the long haul. Most of this value-added work came in the form of asking questions, performing analysis, implementing process improvement or other lofty goals that are so often put on the backburner.
Why did John Smith always have wrong time cards?
Why did Mary Johnson always choose the wrong project?
Why did Acme company always pay late? How can we ensure that Acme pays on time?
Why are we spending so much money on paper?
Well, hopefully the last one solved itself, but the others were questions we always wanted to answer but never had time to dig into. By taking a process that consumed 40 hours of employee time per week, and reducing it by 75%, we freed up a substantial amount of time for these folks to do better, more important work. It’s not that invoice checking was not important, but it was incredibly tedious and something a piece of software could do as well as, if not better than, a human. Until Artificial Intelligence progresses, I doubt a computer is going to be answering those value-added questions anytime soon, which is largely the point of James Surowiecki’s article at Wired.
The whole point of software automation is for humans to do what humans do best and have programs do what programs do best. We want to automate tedious, monotonous tasks. We want to eliminate paper (for various reasons.) We want to be able to ask hard questions, perform analysis, and reach data-driven conclusions. We want to take those conclusions and re-engineer or augment our processes to make them better. If we don’t, we become stagnant, and we’ll never know why John’s time card is always wrong.
Can the software-pocalypse cause you to lose your job? The answer is, yes, it can. Have I ever seen it happen? No, and it is extremely unlikely, but even if this scary possibility is looming over you, there is a simple solution: “Make sure you provide value.”
If you were the main cog in performing one of these tasks, you should be an expert! You know the pitfalls, you know the gotchas, you know the “Oh craps!” You know what to look for. You have all of the ingrained domain knowledge that came with your manual process expertise. Take that knowledge and propose new ways of doing things. Use the new found time you have to suggest enhancements to the process. Implement or spearhead that pet project the CAO mentioned once in the break room. Show management that they made the right decision to automate and keep you in the company. And don’t forget, the Business Analyst is in your office, asking you how to make your company’s killer software, if you weren’t important you wouldn’t even be involved in the conversation.