Why do organizations fail to scale RPA in order to capture its full impact

As long as a companies does not know its processes, it can't automate them.

In a digital economy were the "segment of one" and the "lot size of one" become the norm each business process becomes unique.

The question therefore is: "Is a company able to maintain business process patterns that can be automated?"

In my experince there are a few key drivers of failure for RPA scale-up:

1. Setting up a shaky foundation (under dimensionned/inexperienced team, poor quality RPA software, inappropriate infrastructure)

2. Treating RPA as the stand-alone universal panacea (and not making the most of it in conjunction with other technologies - such as OCR/ICR, AI, etc)

3. Poor choice in choosing the processes to be automated

4. Considering RPA as an unattended solution, that requires no maintenance after go-life point

5. No suficient effort directed towards stakeholders involvement and buy-in (which is a commom problem in most process improvement exercises)

So, once you made sure that you secure the right RPA software (with the least gliches and best support) and you take care to avoid the key failure reasons stated above, you should be on your way to a successful RPA implementation.

Dear Ionut,

My experience with RPA from a vendor side (we have implemented RPA in a multinational company). Several are the challenges which may lead to uneffective scale after go - live and that's why we've made an explicit recommendation to the Customer's team.

1.The first begins from the start of the project - optimization of the process before automation. Make it simpler, decrease the number of the exceptions. Otherwise, after go-live some of the exceptions might be a real stoppers

2. The implemented solution should be maintained by the customer. When they have changes in the processes / integrated systems they should involve the support RPA vendor team

3. Employees should learn how to interact with the robots - what is the meaning of the messages that the robot sends and what is required by them. The employees should stick to the new standards of the process when they perform manually the exceptions . They should do it like the robot does. Otherwise, when the robot should interact with data with different standard he will not succeed.

4. Last but not least - developing one process does not mean the team should not think of the other dependant / linked processes. Implementing a scalable solution requires knowledge about all related processes not only about the process in the current scope

How to overcome :

Choose a vendor not only with RPA development knowledge but processes re-design and improvement as well.

Know what future state of your company you want. Think big and analyse processes deep. The analysis should be at least long as the development itself.

Train the customer's employees incl. IT and give a recommendation how to understand the robot's logic, maintain and scale by themselves and when to contact their support (RPA vendor)

Hope this would be helful. Staying in touch,

Milena Ribarova (Consultant and RPA vendor)