What Repetitive Tasks Are Worth Automating First?

Hi ​@renvik04 

we don’t automate single tasks but whole processes.

We look at the following citeria to decide whether a process is worth to be automated:

• business case
  • time saved per excecution
  • frequency of process excecution
  • estimated development effort
• other aspects
  • process stability: how frequently does the process change? process changes require rework
  • number of applications involved: the fewer applications involved, the lower the complexity
  • application stability: how frequently do the involved applications change? applications changes require rework

Hi ​@renvik04 

I echo all of ​@Stefanie comments! 

I’ll add from our team, we also have a risk and cost avoidance valuation involved, especially for higher profile processes. We look at the rate that RPA could reduce risks and take into account if the cost of risk (fines + rework) exceed the cost of automation. Sometimes the ROI isn’t positive for automation to develop the process, but if automation can meaningfully reduce human error risk, we’ll move forward with automation anyway.  

We ask ourselves and the business units:

• Financial/Error Cost Risk
  • What’s the cost of a human error? Can RPA eliminate or greatly reduce this cost?
    • Are fines involved if a human error occurs?
• Compliance and Audit Risk
  • Is automating this process the result of a failed audit?
    • This question is helpful to understand if the process should be fast tracked to top of our backlog. Additionally, if the automation has been recommended due audit findings, we will automate the process even if the ROI is not a net positive. 
• Customer Risk
  • Would automating this process reduce errors or issues that end up affecting customers?

Looking forward to following this thread!