Why do most Robotic Process Automation projects experience difficulties when trying to move beyond the Proof of Concept or Pilot phase? While most organizations are advised to start with automating the ‘low hanging fruit’ first, the truth is that it can create traps that will impede your ability to achieve RPA at scale. In fact, scaling RPA into the organizational structure is fundamentally different from implementing a conventional software product or other process automation. It’s critical to take a holistic change management approach that is focused on aligning, people, processes, and structures to the strategic business goals of the enterprise.
Organizations typically make a set of consistent choices on how to carry out organizational processes and how to achieve goals. These elements are interrelated and keep each other in balance when processes are carried out. As a result, a lightweight IT solution such as RPA doesn’t naturally fit into the organizational structure which could be a significant barrier to achieving RPA at scale.
To successfully scale RPA, organizations need to overcome two major risks caused by their current organizational structure:
Traditionally, technology solutions are implemented and governed by IT stakeholders. However, because of the lightweight nature of RPA tools, it’s usually driven by the core operating business functions like HR, finance, or procurement. There is a common misconception that RPA does not require any involvement from the IT team, but this is far from true.
As organizations try and move beyond the pilot and achieve RPA at scale, the project requires IT knowledge for tasks such as hosting, governance, support, security, scalability, and assurance. But IT still needs business knowledge to identify the right processes, model these processes, and set the vision for the technology.
But this means that achieving RPA at scale fundamentally contradicts the traditional organizational structure that separates the roles and responsibilities of business and IT groups. Scaling RPA will require new alignment mechanisms that redefine the roles and responsibilities of IT and business stakeholders.
Implementing a traditional IT solution requires an enterprise-wide approach because its use will affect many other organizational systems and processes. But as we’ve identified, RPA is a lightweight solution that can be initiated and configured by the business with limited involvement from IT. The absence of an enterprise-wide approach means organizations will run the risk of local, ad-hoc automation solutions where no end-to-end enhancements are achieved or even considered.
This means that as organizations attempt to achieve RPA at scale, they’re faced with the challenge of sorting through the web of bots and determining how each one fits within the enterprise. Scaling RPA will require that all stakeholders involved to agree to the benefits and uses before implementation to build organizational support and enhance operational effectiveness.
Because RPA is more of a lightweight and flexible solution, it’s easy to start by automating the ‘low hanging fruit’ within one operational department. But by purposefully selecting processes with a high potential for automation and low complexity, teams unknowingly buy into a false expectation that all future automation will perform the same way. And as the organization ramps up their use of RPA, they are executing on false pretenses because they have not determined if there is a demand for RPA in other departments or if it is truly a value add.
Without visibility into organizational departments and a deeper understanding of added value, organizations may face pushback from employees, and limited acceptance of RPA, hindering an enterprise-wide roll-out.
One of the most common challenges that organizations face when trying to achieve RPA at scale is simply the rush to get started. As they rush to digitize their processes and automate them, many realize that implementing RPA is simple, but by no means easy. Many reach the same roadblock – what are the right processes for RPA and how do we optimize them?
There’s no doubt that this is an extremely important question to answer. In fact, some research suggests that a poor choice of processes during the initial pilot is the leading cause of project failure. And because we know that 30-50% of initial RPA projects fail, it’s vital for organizations to ask themselves:
To choose the right processes for automation, consider the following:
Processes that change frequently aren’t good candidates for RPA. Because RPA usually interacts with user interfaces and minor changes to those interfaces add complexity to deployment which may even lead to a broken process. In addition, changes upstream and downstream, even during bot configuration, can significantly delay bots being put into production. A Forrester report noted that because of frequent changes, some organizations may not meet their expected ROI because of the added infrastructure and maintenance costs.
Processes that have a low business impact are probably not worth the time and effort to automate. It’s important to consider processes within an enterprise context prior to development to determine the actual business impact it will have. Of course, you want to be able to say you were able to build 100 bots in less than a month, but if the proper analysis is not complete and if bots are not modeled and optimized from end-to-end, your 100 bots will end up costing you in hours of rework, errors, and general maintenance. In addition, without considering processes end-to-end, you’re more likely to build redundancy into your production environment. As a result, you’re likely diminishing the value of the bots built.
RPA is a generic solution. It’s only one part of a very complex system. You want to make sure that you have an integrated solution that is able to design, model, and optimize processes for large scale, end-to-end automation. Because most organizations have hundreds, if not thousands, of processes that beg for automation; defining, managing, and monitoring them with a single, generic RPA tool simply is not enough. Forrester recommends a mix of intelligent automation technologies to avoid common automation challenges and drive an effective and sustainable automation program.
We’ve outlined many strategic recommendations on how to structure your RPA initiative and achieve RPA at scale, but what about the actual technical side of things? Below are some technical pitfalls that most people run into as they try and scale their RPA program: