Thank you for Subscribing to CIO Applications Weekly Brief
Rpa Sweet And Sour On Your Plate
Gabor Szentivanyi, Vp, A Global Diversified Specialty Chemicals Company
After running the RPA initiative for 3 years in a multicultural environment, a short summary I can give on RPA is that only those should use RPA who either cannot change their business processes, or who do not plan to review their current ERP landscape or collapse diverse ERPs into one global, fully automated solution.
RPA is the dead end of thinking – the tool to turn to when you must forego process realignment discussions. This usually happens when departments only use, but no one really owns processes (I bet you have heard: we used to do so!), but the momentum moves the company forward. Obviously, there are huge number of exceptions, but the real, big and juicy win sold by RPA companies only come when you get to this situation.
RPA is an inexpensive way to cover these gaps, gives short term shine for project managers and process owners, but on a longer term, in the global picture, it only drops a patchwork on dysfunctional operations.
Re-aligning ERP to processes (or the other way out) is a complex, costly, long lasting and politically dangerous process. Thinking long term is the right solution, but nowadays, most of the 2nd and 3rd level managers target short term wins – RPA is a great tool for them. Unfortunately, the short-term victory is hiding the overall – usually unnecessary – complexity on operations, saving time which doesn’t really provide tangible dollar savings. Especially if it is delivered without proper governance, it will be the MS Access of 2022, pushing organizations back to the “no one owns, no one knows, no one maintains” period of mid 2000s, risking and outdating pieces of important processes.
RPA is great to absorb the real repetitive activity of individuals or smaller workgroups, but you constantly need to double check your ROI. Moving unwanted processes from one department to another doesn’t add organizational value, the time spent to automate something might cost you more than “best” sourcing it and just get it done until you re-engineer the underlying processes.
The best potential use of RPA is to automate never changing activities like certain SOX controls, or those smaller activities which are done in every site and country in your organization, so one small automation effort gets multiplied with places it is used, thus saving significant amount of time. Still, you have to update/align automation periodically to the constantly changing environment, which on a long term erodes your ROI. And get ready, once you have automated something and the initial happy time is over, no one will remember that it still does require maintenance.
As a summary, RPA is a risk, but a shiny, easy to sell risk, which gives short term victories. Up to 0.5% of the overall IT budget can rightfully be spent on RPA to add value avoiding irrelevant repetitive activities, but any wider use would only push you backwards on the long run.
Sweet and sour, isn’t it