A recent article about the sharp decline in the number of equity traders at Goldman Sachs from some 600 in the year 2000 to just about two is shocking to say the least. With automated trading algorithms doing the majority of the work, there is little need for manual intervention other than the software engineers required to oversee the systems and troubleshoot when required. So, what does this mean? Are machines taking over?
Computers today are able to see, read, speak, learn, and improvise to solve problems and take decisions. In short, algorithms can perform almost every function that a human brain can, only in more efficient, reliable, and cost-effective ways. Complex algorithms, aided by machine learning, are taking over jobs that previously required the smartest of humans. Such is the supremacy of artificial intelligence.
Recent developments in the financial services industry lead us to believe that finance and accounting and compliance management are two functions that seem like ideal candidates for robotic process automation (RPA), primarily because both these functions involve rules-based work. It is needless to say that most finance professionals, including the CFOs, could easily be replaced by algorithms. All rules-based, repetitive and time-consuming activities within the finance function are being considered for RPA. Tasks such as expenses and payments approval and processing, costs and revenue tracking, budgets and forecasting, inter-company reconciliations or eliminations, and data consolidation for reporting, are unarguably set to be robotized.
But, are we applying RPA blindly?
Perhaps, yes! One very scary trend that were observing these days is that RPA is being applied without much thought. Several financial services firms are automating daily tasks just as they are performed currently. What this means is that even inefficient or non-value-adding sub-activities are also getting ported to the automated systems, without reasonable doubts being raised about their utility. Hardly any assessment is being done to ascertain if there are better ways of handling such activities. Such unchecked application of automation is creating a huge pile of unwanted tasks that would now get executed very fast, but will remain hidden from plain sight. Robots can and will perform those too, without an error, in a jiffy, and at low costs, resulting in hidden inefficiencies that would prove to be toxic in the medium-to-long run.
This is a dangerous trend. In my opinion, CFOs looking to go digital must first critically evaluate every activity and task within their organizations. They must first decide on what does not require automation, and can be completely eliminated instead. CFOs should also assess whether the adverse outcomes of automation outweigh the benefits. Only after giving it an elaborate thought will automation deliver on its promise.
What do you think? Have you faced a scenario in your organization where automation did more harm than good? What, in your view, are some of the aspects that merit consideration before embarking on an automation initiative?