How to scale RPA and Intelligent Automation in your organization

Publish date:

Many companies struggle to implement RPA and Intelligent Automation on a large-scale and as a result are failing to maximize the benefits. In this article, we look at pitfalls and best-practices to keep in mind to create qualified automation demand.

Over the last few years, RPA has helped many businesses create new efficiencies and quality improvements within their organization. However, many companies still struggle to implement RPA and Intelligent Automation on a large-scale, and as a result are failing to maximise the benefits.  Often the reason for this is an ineffective or even missing process for demand generation and process documentation as part of the RPA Operating Model.

This article focuses on generating demand and creation of a back-log and provides you with insights on how to execute a large-scale process assessment successfully.

Why should you start initiatives on large-scale automation potential?

We see four different situations that lead to large-scale process assessment on automation potential:

  • Initial detailed business case for RPA
  • Lack of good process candidates even for Proof of Concepts
  • Scaling Intelligent Automation technology after successful Proof of Concept
  • Roll-out of Intelligent Automation technology after successful implementation in one function (e.g. Finance or HR) or location

What is the best starting point?

There is one assessment approach which provides you a fair amount of high-quality automation candidates with reasonable effort.

Process Map RPA

A starting point should be a process map which contains information on the criteria to be used for later prioritization (e.g. information on FTE or average handling time). For many clients this means to plan some extra time to build their information foundation. In Germany, worker council alignment may be required at this stage.

Once you have your process data straight, this data will be compared with heat maps and best-practice benchmarks (from Capgemini). The output will be a pre-selection of processes which should be feasible for the kind of automation you are planning and which have a reasonable automation business case on the process level.

It is recommended to stress-test the list of appropriate processes from the business and process owners as many cases are business pain points and do not represent a sound business cases due to single-man solutions which are not scalable.

How to do a shortlisting workshop?

In the shortlisting workshop, you will build on the pre-selection and will take into account further information, such as quality and strategic aspects. The result of this workshop will be a shortlist of processes where you see a good business case for automation. To run these workshops effectively, you will require a team of experienced facilitators and a documentation toolset that enables fact-based decision-making on what processes will be part of the deep dive phase.

How deep is a “deep dive”?

It is critical in this phase – even more so than in the shortlisting workshop – to have an experienced team that’s able to ask the right questions and find potential show stoppers. Equally imperative is to carefully select processes to perform deep dives on. Only processes which can also be implemented within the next few months should be assessed, as processes may always change.

What are pitfalls in the decision process?

To guarantee a smooth and effective decision making process, several potential pitfalls need to be considered. First, if only low-level business is involved in the assessment process, we have experienced that the fear of automation will lead to shortlists including merely processes with low automation potential. Therefore, it is mission-critical to include all business layers to gain a holistic view and achieve a high and realistic automation potential. The second common big mistake is not to involve IT early on – underestimating the role of IT and its vast influence on the success of automation. This can lead to processes being selected for automation which are about to modified by IT, leading to substantial waste of resources. Another danger regarding IT is that RPA automation is seen as competition by the “traditional” IT. That could lead to an opposing stance of IT hindering the progress. Thus, the IT perspective and IT itself must be incorporated in the whole assessment process. Finally, selecting processes without in-depth RPA expertise can lead to a shortlist with low automation potential since domain knowledge in RPA is required to be successful.

How to move on?

So how can you get started with scaling RPA in your organization? Begin by investing in a process map and take your time to build a solid information basis since it will be the foundation for all following steps and will hugely influence your success. Once you have your backlog defined, do not forget to apply the minimum viable product (MVP) approach during the implementation. Meaning don’t aim for the 100% automation of a process right from the start but stay with the 80/20 rule. Keep in mind that an agile approach will help you to scale your automation level.


This article was written by Stefan Burghardt.

Weitere Posts


Multichannel Personalisierung im Aftersales

Thomas Hegger
Date icon Juni 22, 2020

Personalisierte Kundenansprache mit Fokus auf individuelle Bedürfnisse als Hebel zur...


Inventive FRC – Credit Risk: Basel IV: Lohnt sich die Verwendung des IRBA noch immer?

Grigori Tymchenko
Date icon Juni 8, 2020

Auf internen Ratings basierende Ansätze werden von großen Bankinstituten wegen ihres...