RPA - Let's stop putting lipstick on a pig

I suspect that Process Improvement methodologies increase productivity and save money more than RPA ever does. At General Electric Six Sigma famously gained $700 million in corporate benefits in just two years, in the first five years the benefits were in the order of $10billion. Here is a good article about it: Why Six Sigma Certification and Training Reduces Company Costs (sixsigmadaily.com).

If all the main RPA software vendors added up all the benefits from all their customers globally, would it come close to this $10 billion benefit in one US company?

Is it time to stop talking about RPA and related IA technologies without also always talking about Process Improvement? My view is that RPA works best when it is an enabler for a Lean Six Sigma (LSS) team and using RPA without process improvement just leaves us 'putting lipstick on pigs' by automating tasks that are part of broken business processes.

As part of the ‘Managing Change’ pillar in my last article (How to scale an RPA capability (also titled 'It's not rocket science!' ) | LinkedIn) I talked about the importance of both ‘Process Improvement’ and ‘Standardisation’ to be able to truly scale RPA.?The banking customer I mentioned found far greater benefit when it redesigned a business process for the robots rather than automating as-is.

Another favourite customer of mine was one of the UKs largest online retailers.?Their RPA COE was based within a Process Improvement team where everyone was Six Sigma qualified.?Their use of RPA as part of LSS methodologies was eye opening to me about how our world should look.?For one project they designed a brand-new anti-fraud process utilising robots that saved the company ￡13 million per year. ??One RPA project, led by Six Sigma, a ￡13 million saving.?For this retailer RPA was part of Process Improvement rather than being something separate, last I knew they had 190 robots doing the work of 600 people.

I also remember a business process we were going to automate for a telecom customer, while looking at improving the process for automation, something the business-based COE always did, they discovered the downstream need for the process could be provided in a better way so this six FTE process was no longer needed at all (the offshore team that had been blindly working the process had not been needed for years).?If we had automated the process as-is without putting it through a process improvement filter we would have automated a process with zero benefit.

We all have stories like these in our back pockets.?We also know stories of RPA project failures where we were trying to automate flawed processes.?If all experienced RPA people have such stories then why is my Linkedin feed that is mostly RPA and IA contacts and companies almost totally devoid of any process improvement content? ?Why are we not articulating LSS + automation as what we do rather than just pushing the tech?

There has been a large increase in Process Mining being discussed as part of Intelligent Automation in the last 12 months which is an encouraging move in the right direction. UiPath have their own Process Mining tool, Blue Prism partnered with Abbyy Timeline,?Celonis seems to be getting far more traction. ?This is really promising but a word of warning, these process mining technologies need to be seen as a process improvement enabler, not as a straight to RPA shortcut.?It fills me with dread the idea of these mining tools recording broken business processes as-is and spitting out inefficient automations.?Process Mining belongs to LSS Analysts, not RPA developers. ?The route to automate with these tools should be Mining->Process Improvement->RPA (if robots are part of the re-designed solution).

My recommendation is that every Intelligent Automation business analyst on the planet becomes a LSS trained (I did a designing for Six Sigma course to help me as an RPA consultant 15 years ago).?If we are in the productivity improvement and transformation business we need to be process change experts.?The same is true of the LSS experts out there, they need to also be RPA solution design experts because RPA should be a key part of the business processes they re-design.

RPA should be part of the Process Improvement industry and we should talk about it as such.?Let’s stop putting lipstick on pigs.