Automating business processes using RPA: is it really a good idea?

Robotic Process Automation – a robot that reproduces human movements in order to carry out repetitive tasks – has a bright future ahead of it. I will begin by telling you a story, which I am sure many of you will be familiar with, at least in part.

A clear observation

As a sales representative, I have a privileged position that puts me in contact with many organisations (including public and private-sector companies of all sizes), in many business sectors. Everywhere, including in companies where I have worked, I have met employees at all levels of the organisation who carry out time-consuming, repetitive tasks, seen as having little added value.

They may involve transferring data from one system to another (Excel to Salesforce, or SAP to a business application, for example), a series of steps to publish a report or monthly analysis, or retrieval of information on the web or in emails, in order to format, store and process it.

I am specifically talking about very simple things here. This time spent represents a significant portion of people's working day (some studies estimate the figure to be as high as 30%!). Such tasks therefore prevent people from concentrating on high-added-value activities (those that are directly related to their role). In some cases, people are even required to perform them outside of their "normal" working day, on evenings and weekends. According to Gartner, by the end of 2022, 85% of organisations will have deployed RPA in one form or another?(there are several forms, with varying levels of sophistication).  

Why the enthusiasm surrounding RPA?

Most of us have come across macros and macro experts, which have been around for a long time. Generally, macro developers have expertise in a specific process and look for a simple way to transpose this expertise into a tool.? For me, new RPA platforms offer three main advantages:

• They are agnostic in terms of the technology environments involved in the process to be automated and have a wide variety of "connectors" to all of these technologies;
• They are not very intrusive. RPA projects use the security environment already present in the information system. Each software robot can be considered as a user, to which rights are granted, in the same way as any other human user of the information system;
• The development studio is mainly based on graphic components. More often than not, this involves configuration rather than development work, strictly speaking.

On the other hand, a certain number of criticisms can be made in relation to an insufficient consideration of information system urbanisation, as well as the risks of falling into the negative effects of "shadow IT".

These criticisms are valid, but it is also true that IT departments are unable to address 100% of business needs at the same time. Furthermore, not all automation projects necessarily generate a return on investment if the aim is to address them at the core of the IS.RPA can be an extremely interesting tactical response in this context. In order to be convinced of this, simply proceed in a gradual, incremental manner.

In order to make initial projects a success, begin by identifying simple workflow processes (a few basic series of tasks), which require a lot of man hours, and which are repetitive over time (daily or weekly frequency for example).

Such projects will bring an extremely rapid ROI.Building on this experience, you can then move towards more sophisticated scenarios, with very high added value, but which require a greater command of the technology used and involve more complex governance (cross-department, etc.). At the end of the day, RPA should be chosen with a pragmatic outlook.?And you will need to maintain this pragmatism throughout its implementation and use.