The service industry is evolving rapidly, and technology has become the biggest boon for it. Given this, Robotic Process Automation (RPA), which is the use of software programmes with artificial intelligence (AI) and machine learning capabilities, has fast found favour with companies looking to upscale their systems and improve processes, to provide a better experience for users and customers. It is highly advantageous to handle manual, high-volume, repeatable tasks which require fixed logic, such as queries, calculations, reconciliations, refunds, automated responses, and acknowledgements, through RPA. By facilitating increased efficiency, rationalisation of manpower, and a significantly reduced turnaround time for such tasks, it is able to respond faster, better, and more accurately.
Geeta Degaonkar, Senior VP – Shared Services, Thomas Cook (India), said, “We at Thomas Cook India decided to explore RPA and its benefits for our shared services. The aim of the shared services was to enable the consolidation and standardisation of HR and finance processes, and deliver an enhanced experience for our stakeholders. Our first objective was to identify the processes that could be integrated into the shared services setup. These included multiple units, which were accompanied by their own unique processes, highlighting the need for an operational back-end that has standard processes, responds fast to changing customer requirements, is more efficient, thus reducing overall cost of operations, and delivers accurate output impacting the business outcomes through measured metrics. This is where RPA was brought in where manual intervention in processes can be automated to make them more efficient and effective.”
Thomas Cook first conducted an end-to-end analysis of the process, in order to identify the best approach to implement RPA. However, the company’s goal was not just to implement RPA, but also transformation. This required it to look at all the processes from end-to-end, and carefully consider the aspects that needed to be fixed. As a result, the company was able to ensure that only those processes were automated which were efficient, and had a maximum impact on customer experience, and not those which were broken.
“This is because the automation of a bad process cannot improve it, but instead, only result in repeated faulty outputs, with no beneficial impact on the efficiency; hence, we excluded them. RPA can be utilised by most of the currently existing processes, including HR and finance, which will help create maximum impact for our internal and external stakeholders. We have implemented RPA on critical reconciliations to help us assess potential risks on open items on a real-time basis. Currently, we have three RPA-enabled processes and three bots in use, which have shown an improvement in efficiency by 40-50 per cent. By the end of this year, however, we aim to implement as many as 10 to 12 bots,” informs Degaonkar.
She further explains, “Our pilot project has also been quite a successful experience. Our selection method for the pilot project was designed to prioritise inputs on our customer-facing process and transactional process. Our Proof of Concept (PoC) was quite successful, following which, we then presented it to our senior management, conducting an end-to-end video documentation of how it helped maximise efficiency. This gave us tremendous confidence to move forward with the project.”
The alignment of the entire team within the organisation was another critical requirement, as RPA is a new technology. So, following the presentation with the senior management, Degaonkar interacted with the HR teams, IT teams, and internal operation managers, among others, showcasing the PoC and the efficiencies that could be gained by utilising it. This helped them bring in the necessary alignment within the organisation, along with gaining the support and backing of the senior management.
She says, “We also needed to ensure that our IT department was aligned with the infrastructure requirements, and other elements necessary for RPA implementation. This is because a large part of the existing infrastructure in organisations might be sufficient for certain transactions or processes, but for RPA, the requirements are more specific. Information gathering is a very critical step, since the subject matter expert (SME) or team leader’s failure to provide the appropriate information to an RPA developer can significantly hinder the success of the coding and logic involved in it. This is why, it is extremely necessary for team leaders and managers to be involved intricately in the processes, with regular internal reviews being conducted to gauge the impact.”
To ensure a smooth end-to-end implementation, Degaonkar’s team also interacted with relevant business teams, and suggested a number of necessary changes in the upstream; including adherence to a particular format for the inputs they send. If these inputs had not followed the specific format, our bot would not have been able to work properly. By informing them of what was being implemented, the processes it would impact, the way it would impact them, and the expected outcome, Degaonkar was able to ensure that the procedure was concluded smoothly.
“One of our significant learnings was the importance of having an output or goal-based project, keeping in mind the primary objective, be it efficiency, quality, customer experience, or compliance, along with how the people available to us could be utilised. This entire planning and detailing phase has to be a part of project management. The UAT testing time is also particularly crucial, because in the absence of the right UAT testing scenario, when the bot enters a hyper-care mode, one might have to keep developing or coding it for various exceptions. We looked at what the exception percentage was, right from the beginning, to ensure we knew what the expected output is. Thus, it became a deciding factor for including the process in the scope of RPA,” she says.
Degaonkar adds, “When we had received an output from the bot during the hyper care, we had our teams work manually to look at it, in order to ensure that it was absolutely correct. This gave us confidence regarding the accuracy of the bot, so as to ensure the maximum output, right from the time that the bot goes live. Detailed planning, and a watertight project management plan, along with a team that could coordinate within the organisation and other stakeholders, while also looking at all the various milestones, raising red flags as necessary, and ensuring adherence to the appropriate time constraints, have together helped us achieve our first milestone on our RPA journey.”