Robotic Process Automation is one of the fastest-growing digital transformation technology with an estimated market size of $1.3 Billion in 2019(Gartner). While the growth rate and adoption of RPA are quite high across industries, many organizations are struggling and not able to realize the value out of this automation journey. As per a recent report by leading management consulting firm, only 3 per cent of the organizations have reached any form of scale with more than 50 robots in service. Below listed are critical success factors to run the RPA program on the right track.
1. Prick the bubble, put aside the hype
One of the biggest challenge of getting started and executing RPA transformation is the misinformation and hype created around it. The aggressive campaign run by many of the software vendors has added fuel to this fire. But to see things clearly, we have to get rid of the smoke. RPA is a non-disruptive automation technology which can be implemented easily compared to other automation technologies and shows result in a short time. RPA is not a panacea for all process automation need and it does not show overnight results. The basic principles of IT transformation apply here as well, and it need to have governance, structure, bot design, build and rigorous testing to get it right and see the much sought after value of FTE savings.
2. Design robust bots. Don’t rush to see the value
The most notorious thing about RPA we hear is that the bots are unstable! RPA bots are very sensitive to the deployed environment and underlying application. Since this is a front-end automation and bot mimics the exact UI steps done by a human, even minor environmental changes can cause bot to fail. Add to this a rushed build phase of the project where requirements are not captured effectively, design is done in a hurry and developers are under pressure to deliver the bot in a short time. The result is an unstable bot. All of these can be fixed if the bots are designed right considering all technical and business aspects of the process. Using the right UI elements, leveraging the Xpath formula effectively and building correct and straightforward logics to tackle complex screen navigation, leverage open APIs wherever possible are few of the key factors to build robust bots.
3. Start small and plan big. Measure, learn and improve continuously
Intending to maximize cost savings, often organizations start automation program with a very ambitious and unrealistic target. Midway to the program, many roadblocks start appearing, and they realize that the project targets are far away from reality. It is advisable to start the automation journey on a pilot basis with 5-10 processes, define tangible success criteria and measure results on regular basis. It is imperative to analyze the sprints, define clear KPIs and metrics to measure the automation results, learn from past and continuously improve in subsequent development cycles.
4. Combine with the right tools. Standalone RPA is much less value compared to when integrated with ML and AI
In the past few years, there has been rapid development in the area of Machine learning and Artificial intelligence. Smarter solutions enabled with NLP, computer vision, predictive analytics, chat bot or virtual assistants etc. are trending. While RPA itself is very effective for process automation, one could see much more value when combined with machine learning. With ML, the scope of processes considered for automation will widen significantly. Some of the examples would be emailed text analysis based on NLP, unstructured data processing, automatic document classification with OCR and NLP, Contract document analysis etc.
5. Don’t mess up the existing controls for compliance
The need for data security, government regulations, consumer privacy, financial transparency has made risk and compliance an integral part of all business process. Today every leading organization has incorporated controls to mitigate various types of threats in their business processes. When automating these processes, one should be cautious of the existing controls and ensure that the To-Be bot process does not skip or compromise any of these controls. It is recommended to perform a detailed risk analysis of each process during the discovery phase and document it, at the same time, also consider the need for new controls in the post automated process
Vijesh Nair has over 11 years of experience in ITES and IT consulting space. He has rich and diverse experience in process automation, process optimization, process re-engineering and enabling it through digital technologies such as Robotic Process Automation (RPA),BPMS and Machine learning. His role involves leading large digital automation initiatives to reduce manual repetitive work and deliver value in terms of FTE savings, efficiency and quality.
