Robotic process automation Robotic process automation

Among all the technologies that are currently trending, firms investing in digital transformation have mostly explored robotic process automation (RPA). However, many are disappointed when the expected return on investment fails to materialize. Indeed, RPA may not necessarily be the goose that lays the golden egg. We explore the reasons why this is so and what can be done about it.

For decades, evolving technologies have attracted massive investment. In many cases, the payoff has proven more than worthwhile in terms of competitiveness, impact and added value. RPA is not a new technology as its premise started in the late 90s, but recently, it has emerged as a buzzword in the context of digital transformation as it is frequently associated to new technologies such as blockchain, artificial intelligence or machine learning. From basic e-mail replies to more complex reconciliation processes, RPA refers to the automation of a process that was previously performed by humans, bringing benefits such as speed, accuracy or efficiency. But is it really all that simple?

A good way for firms to stay competitive is cost reduction. It’s one of the main reasons demand and desire for automation keeps growing, and explains the impressive increase in companies’ investments in RPA. In September 2019, 53% of respondents to a survey admitted having started their RPA journey and this number should rise to 71% by 2021. Forrester predicts that by 2023, more than USD 12 billion will be spent on RPA services. Service providers promise a return on investment of 30% to 200% in the first year. However, cost savings are not the only benefit of automation; companies are also pursuing higher quality services, faster payback and improved productivity by reducing the amount of manual work. For specific tasks, the ROI is valued between 600% and 800% when merging robots and artificial intelligence.

In general, service providers agree that 30% of daily tasks can be automated, which could potentially lead to significant cost reductions, efficiency gains and a reduction of human errors. However, a 30% decrease in manual work does not translate into cost savings of 30%.

Although the rush to RPA can be easily observed, the possibilities of such applications are just beginning to unfold. Many projects have been completed to date, however success is not the only story. While recent research shows that the expected payback on RPA is less than 12 months, other studies show that this is not always the case. Indeed, between 30% and 50% of RPA projects fail. The reasons are manifold, and include underestimating the complexity of the task, overestimating the cost reduction or lacking a global governance and strategy. Some even go as far as to believe we could be in the middle of a bursting bubble, where the hype and investments far outweigh what is really possible.

The lack of global governance and strategy for automation is probably one of the main reasons such projects tend to fail, or at least fail to generate the expected ROI. The main element to consider is that even though robots are, ultimately, a bunch of automated actions performed by scripts, they still have to be designed, implemented, maintained, monitored, and upgraded.

This, however, does not mean that companies should not invest in RPA. On the contrary, automation is undoubtedly where most if not all firms should head towards – as witnessed during the global health crisis we are going through and its impact on daily operations. However, the investments should be made intelligently, considering current and future state, impacts on the workforce, benefits, etc.

The most innovative and entrepreneurial of firms have been successful in developing automation by taking the topic on internally and setting up task forces to go around the company, automating one process at a time. This approach is successful not only because employees know their processes more than anyone else, but also because it creates a positive dynamic, increases interactions between departments, and limits that “fear of losing my job” feeling (some estimates anticipate that over 100 million of the workforce could be replaced by robots by 2026). However, a study from McKinsey suggests that only 5% of jobs can be fully automated. Most likely, RPA implementations will result in movements within the organization: automation will indeed lead to employees being able to spend more time in value-adding, client-facing and front-end activities. Companies that address automation in this way see an increase in employee satisfaction and, ultimately, job retention, which may seem paradoxical in this context, but has proven extremely positive.

Ultimately, it would be a mistake to think that RPA projects can be done entirely without external support. Indeed, the more complex the tasks and dependencies are, the higher the risk of failure of such implementations. Therefore, when assessing whether such project should be conducted internally or not, focus should be put on qualifications and experience. In the end, a failure – even if the project has been conducted internally – represents a significant loss of time and money. As such, firms should examine whether they truly have the capabilities to implement such RPA projects themselves, especially when they are highly complex.

Before running to RPA, companies should spend time evaluating the impact and challenges. This will allow them to decide what their overall automation strategy should be, knowing that more than ever now is the time to act. Finally, having a long-term vision of what the firm wants to accomplish with such projects is necessary. For instance, a long-term project involving RPA on control automation should be considered in light of the latest technologies such as artificial intelligence or machine learning, taking into account the criticality of staying in control. Investing in RPA is beneficial if done right, and having a mid-term vision on what firms want to achieve will support sustainable solutions.