Robotics process automation: You might be doing it wrong

Robotics process automation (RPA) has the potential to save companies significant amounts of money by shifting mundane, repetitive tasks, such as data extraction or data translation, from humans to robotic software agents. In fact, the McKinsey Global Institute estimates that half of the activities that workers are currently paid $15 trillion in wages to perform globally could be automated with current technologies.

An example of RPA would be a robotic agent that automatically opens and reads word documents or spreadsheets, identifies and extracts the necessary data, and transfers the data quickly and accurately to the appropriate business processing system. When it works right, RPA can speed up business processes, reduce manual data entry errors and free up employees to be more productive. RPA is also the first step on the road to higher value activities like machine learning and AI.

Unfortunately, many companies that have launched RPA projects have been disappointed when the savings have failed to meet lofty expectations. Other organizations have achieved early successes with proofs of concept, but have been unable to scale beyond the 'low hanging fruit' processes.

In order to avoid the pitfalls that have prevented companies from realizing the full benefits of RPA, there are several key points to keep in mind, particularly when it comes to deciding whether to deploy assisted or unassisted RPA.

The first mistake companies make is jumping ahead and trying to grab quick and easy wins before doing the necessary strategic planning required to support a scalable, long-term and sustainable RPA deployment.

For example, companies need to make sure they don't try to automate processes that are inherently inefficient because simply doing the wrong things faster doesn't help the company achieve lasting business process improvement. Companies also need to analyze business processes in a holistic, end-to-end way, otherwise speeding up one part of the process might just create a bottleneck somewhere else.

For many companies, it's tempting to implement unassisted RPA in which the software agent operates without human intervention. That can lead to early wins, but it can also lead the company into a dead end where they run out of projects that lend themselves to unassisted RPA.

Unassisted RPA requires processes that can be completed with simple decisioning, but it turns out that these processes are few and far between in the real world.

The real value of RPA will only be realized when companies adopt assisted along with unassisted RPA, combining the speed and accuracy of robots with the complex decision-making and creative abilities of humans. By working in tandem, humans and robots can achieve the efficiency goals that companies are seeking. And RPA, if done with the requisite strategic planning, can serve as the foundation for the organization's business process automation journey to more complex, higher-value technologies like cognitive computing, machine learning and AI.

[Attachment] Rajeev Talwar is a digital professional with 17 years of experience in Digital, Robotics, Design Thinking, Direct Marketing, Social Media, Social CRM, CRM and client servicing for organizations like UK Cabinet Office, Deutsche Bank, Unilever, P&G, Novartis, Bristol-Myers Squibb, The Payments Council, Ministry of Justice (UK), Engie (Belgium) and Shell.