by Saurabh Sharma, on Sep 30, 2020 1:05:42 PM
Achieving success with process automation initiatives calls for synergies between People, Process, and Technology facets, which is clearly missing in many enterprises. Implementing RPA as a point solution for task automation without proper business-IT alignment will ultimately lead to struggle in scaling. For larger implementations, enterprises must set up a center of excellence (COE) and have a proper governance framework in place to ensure that RPA initiatives remain on track and deliver business value.
Friction between competing facets continues to hinder scalability
The first wave of RPA adoption was driven by the need to reduce the costs and errors associated with human involvement in mundane, repetitive tasks in back-office functions that were difficult to automate otherwise. RPA is capable of performing a large number of user interface (UI)-based tasks in a predictable way, and in the case of an increase in the number of tasks, enterprises have the option of deploying additional robots. However, many enterprises have struggled to scale beyond the first dozens of bots owing to the friction between competing facets.
RPA initiative leaders who do not spend enough time on analysis and optimization often think of RPA as a hammer and see most of the business processes as a nail to hit. This means that the automation of an inefficient business process with RPA is unlikely to deliver significant gains. Unoptimized, fragmented processes result from process and system variations. Process modeling and optimization can help reduce fragmented processes and ensure rapid and sustained gains from RPA implementation.
Process automation initiative leaders often opt for automating processes in an "as-is" state via software bots to realize the low-hanging fruit in terms of time and cost savings – these are quick wins that help in developing a broader case for investment in RPA implementation. While RPA can deliver a quick return on investment (RoI), it will not always reduce process inefficiencies. This is an area where process re-engineering and optimization (via a process modelling tool or on a white board) can deliver significant value. RPA initiative leaders should keep one point in mind – do not forget the “process” in process automation. In case a process is automated without optimization, there is likely to be a need to change the automation flow later on, once changes are introduced in the existing version of the process. This iteration can be easily avoided by bringing business and IT stakeholders together to develop what could be an “ideal first version” for automation.
However, technology is not the main culprit here. People and processes have a decisive role in determining the success or failure of an RPA initiative. As indicated by Ovum, RPA skills shortage, poor change management, lack of IT ownership, ill-defined success criteria, and disregard for infrastructure management considerations are some of the factors that can lead to failure. Indeed, results from a global RPA survey conducted by Deloitte in May 2019* revealed that only 8% of surveyed enterprises have scaled RPA beyond 51+ automations.
Then it might be the case that IT function did not plan for the infrastructure requirements for scaling RPA and failed to take into account the difference between RPA and traditional IT deployments. In this regard, having the flexibility to deploy on infrastructure-as-a-service (IaaS) provisions ensures that not much time and energy is spent on provisioning infrastructure capacity as per the increasing scale of implementation.
As and when RPA bots (i.e. digital workers) are brought into the operating environment of an enterprise, it is necessary to plan for change management and to alleviate fears in human workers as to how to work with a software bot. A good case in point is of a large enterprise wherein designated names where assigned to bots and each bot had employee and email ids, and represented as a dedicated digital worker within their specific business unit.
Business and system execution failures are quite common events calling for effective exception management. Business exceptions occur in cases where data or inputs are not in line with business and application criteria or when established business rules are broken. System exceptions could arise from application crashes, dramatic changes in the UI, or a slow-to-respond application. Needless to say, such exceptions call for a proper exception management mechanism to ensure faster issue resolution or forwarding to humans for their efficient intervention.
A holistic strategy with business-IT alignment is indispensable
It is not uncommon to see enterprises struggling to achieve success owing to a fragmented strategy and a lack of a holistic vision for end-to-end process automation. Managing relationships with multiple automation tool vendors and professional services providers is itself a difficult task, not to mention that such an approach may lead to discrete objectives and islands of implementations that fail to scale beyond a few bots.
Business-IT alignment is a key requirement, as is the case with every strategic initiative. Without IT involvement, there is risk of creating islands of automation infrastructure and implementations that do not align at a strategic level. Data security and governance is a key concern if individual RPA initiatives are run without IT’s oversight (i.e. shadow IT). Automation initiative leaders should establish an automation CoE for systematic adoption at enterprise level. Development of RPA skills and preparation for change management are other key focus areas for adoption at scale. The center of gravity of RPA implementation can be closer to the business side (i.e. line-of-business or LoB-led with IT’s oversight), however, without effective involvement from the IT side, the chances of scalable and sustained bot operations are significantly reduced. Moreover, there is a need to look for options for tech integration with intelligent document processing (IDP) and artificial intelligence (AI) or machine learning (ML) capabilities to automate a greater proportion of more complex business processes.
*Source: Deloitte Insights, Automation with intelligence, 2019
