RPA is no news in the banking and insurance sector and many organizations have passed through the verification and implementation process in the last four years. However, it is not always easy to decide where and why to deploy a robot. Such a decision requires a thorough consideration of all involved aspects. The automation of business processes is a tool that helps banks, insurance companies, institutions and businesses to boost their competitiveness, to reduce costs and, last but not least, to enhance customer satisfaction. Proper implementation brings along a wide scale of benefits, including those mentioned above.
The general purpose of RPA is to leave manual tasks with low qualification requirements to robots and to free the staff up for processes with higher added value and higher qualification demands. The implementation usually entails savings in terms of processing costs (FTE) and the robot considerably improves processing efficiency and effectiveness. In theory, it can work 24/7, but in practice, its “working hours” are limited depending on the availability of automated systems or data sources. In any case, robots can handle many times the volumes of processes that could possibly be dealt with by humans. Thanks to a lower error rate as compared with humans, RPA strengthens the security of sensitive data processing, enables instant scalability and improves the overall data processing quality and consistency. With RPA, processes can run three to five times faster and leave behind a clear audit trail.
And finally, RPA allows for a faster and cheaper development compared with equivalent IT solutions integrating various robotic systems. Still, it is necessary to consider the costs of implementation and development when introducing an RPA. Operating costs would normally include the following:
- Licences required for the use of the robotic process,
- HW and SW involved in the automation platform,
- The cost of operations, i.e. running the robotic process,
- Minor repairs and maintenance of the robotic process.
Which Tasks Should Be Automated and Where
A robotization normally focuses on back office processes that are often triggered automatically without user interaction, such as service modifications or setup. But robots can also help to improve front office processes. An employee may initiate a robotized task by entering the required data and is then free to serve the client while the robotic process runs in the background. Call centres, for instance, may work in the same way.
Copying structured data by hand from one data source system to another (loan documentation > loan processing system) is a typical candidate for robotization. This type of tasks results from the lack of integration between systems, because the integration would be either too expensive, or too difficult to implement. The demand for robotization of tasks involving unstructured data is on the rise. To fulfil such a requirement, the robot needs to involve artificial intelligence or natural language processing features. The input normally comes in the form of written documents, e.g. Word documents, with information that needs to be extracted and then entered into or searched up in IT systems.
Another process suitable for robotization is the migration of data between systems with comprehensive validation functionalities, where simple data migration would be too complicated or inefficient. Yet another is data checking that cannot be done simply at data level. This may involve, for example, checking data in stored files and comparing them with data existing in applications.
The Next Step in Decision-Making
A process pointed out for automation must be clearly defined, including its inputs and decision-making patterns. As we know from our experience, clients across sectors most often expect that automation will save them capacities (FTE) engaged in routine tasks. Employees should then pass over to higher-added-value operations and become more creative. Business case calculations are essential for this specific target and they are very easy at the same time. They include the assessment of costs and benefits of implementing a robot. First of all, an organization should look for answers to the following questions:
- Is the process initiated with enough frequency to justify its automation?
- Is the process lengthy and labour-intensive?
The next decision should be whether to start the robot manually, or automatically. Either an employee may activate it when needed, or batch processing may be triggered automatically. As was already said above, a clearly defined and structured process with deterministic decision-making pattern is a necessary prerequisite. What next?
The potential for savings is not the only benefit. The advantages include speed, precision, 24/7 efficiency and effectiveness, scalability, and auditability.
The Work of Banking Consultants Is Transforming
Process automation is not just a matter of saving time and relieving back office of routine tasks. A robot deployed at a branch desk, for example, will deeply transform the entire working pattern of banking consultants. While the robot processes the required tasks, the consultant is free to attend to the customer’s needs and requests, saving time both for the customer and for the consultant. Robots do not necessarily replace humans. The two may work in synergy, allowing bank personnel to dedicate more time to customers and to processes with higher added value.
Clearly defined processes with precisely specified decision-making rules
Robotization is suitable for clearly defined processes with decision-making steps independent of human judgment.
Applications and data input structures that are not prone to frequent changes
Robotization is an ideal choice for stable systems without repeated modifications. Otherwise, the robot would require too many adjustments.
Similarly as applications, the processes to be robotized should not be prone to frequent modifications.
Processes with digital inputs
Data must be provided in a digital format, because the robots are in fact software. Otherwise, it will be necessary to prepare the data in the first stage, e.g. to scan paper documents.