Cutting Through the Hype: A Holistic Approach to Robotic Process Automation

Date:   Wednesday , January 11, 2017

A subsidiary of the Infosys, Infosys BPO provides a broad spectrum of integrated end-to-end business process outsourcing solutions/services that delivers transformational benefits to its clients through reduced costs, ongoing productivity improvements, and process reengineering.

The concept of Robotic Process Automation (RPA) is fast catching-on in the services sector as much as in the manufacturing sector. In this article we will look at the Business Process Management sector. Many organizations are looking to implement the magic of RPA across their processes to dramatically lower costs, enhance quality, enable rapid scaling, and improve processing time. While many have been quick to jump onto the bandwagon, there have been many instances where RPA has delivered only modest gains. And so, is all the buzz surrounding RPA just hype? Definitely not, but with a qualifier. RPA can deliver tremendous results - the key is to see it not as \'the\' tool, but look at it as one powerful tool that works best in conjunction with other tools. Let\'s look at some key challenges in automation to understand this better:

1) Existing Processes Are neither Harmonized nor Standardized

The scale and complexity of today\'s global operations demand nuances for processes for multiple reasons; be it by country, product type or lines of business. This in turn increases the degree of complexity in automation wherein multiple end algorithms are needed, given the combinations and unique process flows. An example of this complexity - a leading hi-tech company was faced with this challenge while automating its channel incentive payments. Every region had its own variations for the payment calculation and reconciliation processes, and the partner needs were different.

2) Upstream & Downstream Processes Need To Be Changed To Enable Better Interfacing

Many organizations look at automating only a few activities of an end-to-end chain without looking at the upstream and downstream steps. This leads to multiple challenges, given that the inputs are not standardized, which would again mean multiple combinations to deal with. Take for example, the claims processing unit of an insurance company that faced this challenge of complexity when they could not standardize the claims forms that they got from different third-party administrators (TPAs) and end customers.

3) Underlying Systems & Applications Are Out Of Bounds

Organizations still understand the implications of RPA and are often hesitant to allow automated scripts to access their systems and applications. Though RPA as a technology resides in the service layer and does not impact any data or applications in the underlying systems, there are still a lot of questions about whether the technology will impact the integrity of internal systems.

The net sum of all of these challenges is that when companies try to automate a 100 FTE process by around 80 percent, what they end up automating is only 15-30 percent of the activities. There are many tools that can deliver this kind of result e.g. a lean implementation or sometimes just through Kaizen. Then, what is the right approach to go about implementing RPA and reaping optimum benefits from it? Here are a few prescriptions:

1. Don\'t Look At a Process in Isolation

Organizations need to study the processes- both upstream and downstream, and improve them. If today, the inputs for a process come from twenty different vendors, each with a different input style, getting a standardized input will require policy intervention that mandates that vendors adhere to a common format. Unless you get all the vendors to agree, implementing RPA will be an extremely complicated exercise.

2. Examine the Constraints & Willingness of the Organization To Standardize The Process Beforehand

There are two ways to go about RPA. One is to automate first and progressively standardize later, so that there is only one level of change management to be dealt with. This method will however increase the degree of complexity of the automation effort. The second method would be to standardize the process completely before automating it - thus minimizing the complexity, but creating two levels of change management effort. The question of the optimal level of standardization needed before automation is a call the organization must take based on its internal culture of receptivity and adaptability to change.

3. Try & Build Automation Functionalities into Your Already Existing Tools

It is only after streamlining existing tools that you should finally get a software robot to run the process and realize its true impact. Automating steps in such a way that the processing occurs in the tools to the point that human intervention is minimal, is the best way to automate a process. One can then look at automating the residual steps through RPA. Take the example of a manufacturing company processing payments that were being made to its partners through legacy systems. It identified opportunities to automate certain steps into the payment engines while the residual steps were automated leveraging RPA.

4. Remove the Roadblocks to Access Systems And Applications

An automation partner\'s ability to access applications does impact the speed of execution. For captive organizations, this is not very complicated. However, for third party providers, getting access to the client\'s applications needs to be simplified. This would mean strong governance mechanisms involving the IT team of the client organization.

To conclude, everything you have heard about RPA is not just hype - it is a critical tool in the toolbox. However, unless you take a holistic approach and use it in conjunction with other tools, you will not be able to wield its tremendous power to the greatest advantage.