Home » Technical Topics » 3D Printing

6 Steps to Get the Best Out of Your RPA Implementation


Over the last couple of years, there has been a lot of hype around robotic process automation. This makes a lot of sense if you consider that in 2018 Gartner was already labeling it “the fastest growing segment of the global enterprise software market” (with a revenue growth of 63%).

Moreover, based on a Dave Vellante study conducted between April 2019 and 2020, RPA was the technology with the highest adoption rate, together with machine learning and artificial intelligence. Yet, RPA implementation has led to mixed results for companies across the world and across industries.

According to the Deloitte RPA Survey The robots are ready, Are you?, only 3% of the surveyed organizations have scaled beyond 50 software robots, which confirms that scaling to enterprise-wide use, despite being the envisaged completion of the automation journey, is still a serious burden. Another research conducted by EY corroborates this result and shows that as many as half the RPA projects initially fail.  

We can learn from these mixed results that it is mandatory to keep our feet on the ground and not fall for the hype, while at the same time hold on to the positive facts about automation. Achieving this balance is crucial now, in the climate of the ongoing coronavirus pandemic and its expected daunting economic effects.

The key to automation is to assist organisations to remain competitive or even improve across a rather uncertain business environment. And the first step in order to get the best out of your RPA implementation is to start well, meaning to pick the most suitable processes for automation. If you want to find out more, take a look at this article about 8 questions to ask about processes before implementing RPA.

How to make the most out of your RPA implementation

The list below aims to provide some guidelines for making the most out of software robots deployment. These are a few ways that can make RPA implementation a very good starting point for your company’s long and winding automation journey.

1. Start with a ‘bird’s eye view’ over the automation process

A centralised model is our recommendation because it ensures an integrated and coherent long-term vision of the road towards RPA deployment. A robotic Centre of Excellence (CoE) provides the institutional framework that fosters the development of the automation project beyond the implementation stage, towards scaling oriented deployment, while maintaining high standards across these different levels of automation maturity.

It is a multidisciplinary team, including IT, C-suite, marketing, and department representatives. Oversight from the CoE sets out a proper, holistic maintenance of the RPA project. Additionally, it can help you to coordinate the change of company culture which can help your employees to assimilate software robots into their perception of the workflow.

2. Make a goal oriented choice of the automatable processes

Questions regarding what you need automation for, for which processes, and in which departments should guide the selection of automation opportunities. Assuming that your overarching objective is to attain the RPA benefits in the least amount of time, your top candidates should be processes that have the following features:

  • Manual and repetitive;
  • Rule based;
  • High volume and high frequency; 
  • Mature and stable;
  • Have measurable savings;
  • Have readable inputs.

Running a cost – benefit analysis of automating the shortlisted candidate processes will help you make the final decision.

3. Use RPA performance metrics to track progress along the way

The metrics provide a quantitative demonstration of the financial, business and operational impact of robotic process automation on your company. Key performance indicators such as cycle time, tool utilisation or employee retention rate yield valuable information about the benefits of automation, which can be disseminated throughout the company and generate excitement and buy-in. 

4. Promote a productive collaboration between IT and business units (e.g. accounting/ finance, procurement, HR)

The underlying idea is an effective, goal oriented division of labour. People working in business departments are those who have the know how with respect to the processes that make up the workflow, and who are in charge with operations performance. Hence they are the ones to specify the financial objectives, and to make a corresponding choice of the processes to be automated. 

But make sure you don’t leave the “geeks” out! Different from the beginnings of automation, support from the IT staff is necessary for robotic process automation implementation that works, and even more for scaling successfully. Among others, they can ensure that software delivery is indeed consistent with the technical infrastructure of the company, secure bots’ end-to-end performance, manage cyber security, etc.

The latter is particularly useful in the COVID-19 context, when the sudden increase of the usage volume of online communication is likely to cause additional security breaches. 

Areas that will benefit from the two teams’ commitment to shared goals are, for example, bot updates and application changes

5. Initiate RPA delivery by thoroughly documenting the processes that you will be automating

The best way to do this is to create a Process Definition Document (PDD), which outlines how the robot is going to mimic the way humans conduct a particular process. Invite subject matter experts to contribute to the PDD. The Document is an illustration of the joint effort of people from business and IT teams

It’s as if business people would ask for bots that fulfill a certain function effectively, and the IT staff would quantify the effort required to build such a bot and, eventually, deliver it. The PDD provides a twofold description of the automated process, highlighting both the benefits and the costs (i.e., the effort) of having it performed by bots.

6. Constantly invest in skilling and upskilling your team of RPA developers

Since the RPA sector is relatively new, it is rather challenging to find experienced and knowledgeable people. However, you do need such people to get the best out of your RPA implementation. This is why we recommend that you train your staff along the lines of “life-long learning experience”, keeping them up to date with technological developments.


Sometimes RPA implementation does fail. Whether this is because you start the journey with unrealistic expectations, or because you allow yourself to rush to the expected benefits, thereby losing track of potential obstacles, or because many other reasons, things can go wrong. It’s crucial that you take this possibility seriously and, should it happen, you are set on learning from the mistakes you have made.

The bottom line, however, is that failure can be avoided. You can make savvy implementation decisions if only you keep your eyes open for pitfalls, such as improper management of employees’ resistance, wrong choice of processes to automate, or unstructured implementation teams, and adjust your strategy such that you overcome them.