If your RPA vendor or consultant says this… Walk Away!

At a recent conference I was speaking with a few folks about their current progress on Robotic Process Automation (RPA) initiatives. Unfortunately, many that I spoke with were frustrated. They had several reasons, but the top 2 were:

  1. All vendors seemed to want to point them at “order to cash” as the starting point but that is either not where they wanted to start or there was no real direction after that.
  2. No clear definitions or guidance on defining feasible candidates for RPA.

Of all their concerns, there were a few that were more disconcerting for me.

The first was that some RPA vendors and consultants told them that there was not a need to spend time capturing their processes first. My questions back to them was, without a clearly defined process in place, how do you:

  • know what you are automating?
  • know how to automate it? (since you must give very clear instruction to your RPA bot)
  • know what resources (people or technology) is involved?
  • know if the effort and cost to automate is worth it?
  • establish baselines to know if the automation provided the improvements or efficiencies you were looking for?

Bottom line, you cannot remove the P from RPA and expect to be successful. Process is both figuratively and literally RPAs middle name. If you hear your vendor or consultant say this… walk away.

The second was that it was ok to automate a “bad” process. Part of this goes back to the first issue of not clearly documenting your process first. You will not know if your process is bad if you don’t know what it is. Then, if you have defined your process and know it to be inefficient (or bad) why would you not fix it first then look for ways to automate it.  The response from the consultant was that with all the time you save by having automated the process, you will now have time to work on improving the process and making it more efficient. This does not compute, and simple math will prove this.

Many RPA vendors and consultants tout a 100-200% increase in productivity from their automation projects. For this example, let’s say you have a process that kicks out 50 defects or errors per month. Currently you spend some amount of time, for easy math let’s make it 10 mins, per error fixing it.  So that equates to 500 minutes or 8.33 hours per month fixing issues in the current process.  This is considered inefficient and the fire drill each issue created was a nuisance to your staff. To resolve this, the vendor or consultant suggests automating the as is process, touting you will have the ability to fix it with all time you will gain by automating it. IF we had 50 defects before, and we increase efficiency and productivity with automation as the vendor or consultant claims by 100%, you are now staring at 100 defects per month with 16.66 hours needed to resolve issues. I ask you, where will your team spend all that new found time? Looking for efficiencies and improving the process, or putting out the forest fire you just lit? If the vendor is fortunate enough to get you a 200% increase, I dare say your business may not be able to handle it.

If you hear a vendor or consultant say “it is ok to automate a bad process” remember math … walk away.

For more great articles on how and when to look at RPA and Workflow Automation, click here.


Related Content


Streamline Compliance and Mitigate Risk with Process Mining and Process Design 

In today's complex and highly regulated business environment, compliance management and risk mitigation have become top prior...

What is Business Process Management Anyway, and Why Do Companies Need It? 

By Kim Scott     There’s a lot of information to be found on the internet about what defines Business Pro...

Predictive Analytics: Digging Deeper to Get Things Done

An Intellyx BrainBlog for iGrafx In our previous chapter on automated decisioning, we discussed the importance of bringing...