While RPA can transform your operations and provide a very real benefit / ROI, enterprises still face challenges while scaling up their RPA initiatives. There can be multiple factors which can affect the up-take of a large RPA program and one of the key culprits that starts to emerge is the enterprise automation velocity. While working with multiple clients at various maturity stages of their enterprise automation journey, I have realised that the pace and cadence at which their process automation are developed, deployed and stabilised to start returning benefits, has been the issue more often than not.
Most enterprises have individual automation programs running, and a common view that a Head of automation / RPA COE lead or a Chief Robotics Officer has to his perusal to track the health of these individual automation programs is at times lacking. Now imagine when you are tracking a large number of such automation programs that are running and vying for your attention, this issue can get a multiplier effect and can create a whole lot of complexity. At times even leading to abrupt stoppage of an automation program. With an attempt to make this easier to comprehend and provide a metric / measure oriented view, I went about viewing this problem from a data perspective.
Automation by its essence needs to have a rapid turnaround time to develop and deploy. Velocity of process automation is one of the key areas to monitor and a key metric to identify which programs need scrutiny. I believe the automation velocity index is an important metric to track and correct. The below depiction is a version that I wanted to share about the mechanics of the enterprise automation velocity index and get a general feedback on- from other intelligent automation experts.
Most enterprises classify their processes as simple, medium and complex and so do the automation products in the market. As an example, let us assume that in the past your average time to develop and deploy process automation for a simple process is 2 weeks, for a medium process is 4 weeks and for a complex process is 8 weeks. Based on the number of automation programs that you are running, you can quickly calculate your enterprise automation velocity index. A possible way to do this with minimum fuss would be to calculate a weighted average score of each automation program and average it out for the whole of the enterprise. This would be : Sum of (No. of processes – Simple, Medium & Complex X average time to develop & deploy)/ (Sum of average time to develop & deploy for all processes)
e.g.: for Automation program 1 below: Automation velocity index: = ( 4*2 + 2*4 + 1*8) / 2+4+8 = 24/14 = 1.71
An ideal approach would suggest that the enterprise work towards an ideal velocity index – e.g. 2 in this case and evaluate all similar automation programs to be well beneath the index. Programs which are not meeting the index or are approaching the threshold are the ones to watch out for and may need course correction so that they don’t skew your enterprise automation velocity index. The index may also be able to identify the most ideal set of process automation that one automation program of work should be built up of, for your enterprise without impacting the velocity index. Of course the end goal is to keep improving on your enterprise automation velocity index which is primarily influenced by the time taken to develop, deploy and stabilise your process automation.