Insurance technology is evolving and applications of digitisation are only expanding further. One of the most basic problem statement which insurers face is, reducing physical intervention in system driven/ assisted processes. In course of handling day-to-day insurance operations like underwriting, claims processing, policy servicing, etc., insurers are flooded with overwhelming levels of routine, repetitive and operational task; it’s automating these tasks that we need to work on.
The product or outcome expected is always about adding value. The value in this case is about time of processing and obviously saving cost because of lesser manual intervention. Here, I feel Robotic Process Automation (RPA) seems to be the technology that will drive the insurance sector towards better efficiencies and customer centricity.
What is RPA?
It’s like as a software BOT, which can be programmed to take up repetitive, predictive, rule driven tasks with a clear-cut decision tree and outcomes. RPA aims to complete repetitive tasks more quickly, and more accurately than humans, without foreseeable downtime. This, in turn, frees up human time allowing them to focus on tasks that require reasoning, judgment etc. In simple terms, it’s like an advanced version of an Excel MACRO which interacts with multiple systems, based on rules.
Usage of RPA in Insurance
RPA can add value across the entire insurance value chain, some of the use cases are mentioned below.
New Business and Underwriting:
RPA’s role in underwriting can be amalgamating data from different sources to accurately assess the risk associated with any insurance. Underwriting requires evaluation of risk and exposures of clients, which can be a lengthy process especially for medium and large risks with data scattered across multiple sources and reports. RPA can read this data, organise it and accumulate it in one place for underwriter’s convenience. By using a set of rules, it can go one step further and make recommendations about accepting the risk by assigning an acceptability score, which can be used for decision making and pricing.
Policy Administration and Servicing:
Policy issuance systems old and new serve similar functionalities where a bunch of users are involved in issuing variety of policies based on certain defined processes and norms. Scalability unfortunately is not practical and what makes it cumbersome is multiple systems a policy punching employee must go through, which slows down the process even further, leading to inefficiencies and mistakes. RPA can allow the issuance system to read data and process it from other systems, for example vehicle inspection module in case of motor insurance or receipt in case of finance. It essentially automates transactional and administrative parts of activities such as accounting, settlements, risk capture, etc. RPA can also minimise time for policy servicing requests such as modifications and cancellations. Additionally, it can not only help keep track of proposals, but also track delivery status by reading/ interacting with external mediums such as courier companies and postal authorities.
The gains of economies of scale to be derived are massive and that too without any operational risk per se. Freed from mundane and routine tasks, back-office staff can be redeployed into front-office roles where they can focus on complex customer demands and generate growth.
Sales and Distribution:
RPA can be leveraged in sales administration by preparing automated customisable and gamified scorecards that read sales figures/ reports. This can help any sales team in understanding where they stand in terms of their budgets, especially when it comes to certain lines of business. Don’t think of it as MIS reports, but rather customisation of sales performance indices for empowering customers. RPA can also provide better prospect data based on models built internally by putting up a clear propensity index.
Regulatory Compliance and Process efficiencies:
Tracking and measurement are hallmarks of RPA. The audit trail provided by RPA helps with regulatory compliance which further supports process improvement. Therefore, it helps better claims and customer service response time, in turn benefiting customers with streamlined applications.
Since insurance is a regulated sector it needs to adhere to various guidelines for documenting work and creating audit trails. Some of the use cases where RPA can significantly contribute in improving regulatory processes are validating customer information, flag non-compliance real-time based on defined set of rules/ parameters, generate regulatory reports.
Claims Registration and Processing:
Processing a claim requires a lot of information from various sources like policy admin, finance etc. The amount of manual processing and checking makes it unviable to handle large volumes. This leads to an adverse impact on customer experience and their perception about the industry as whole. We have already started using AI within the claims management process to cut down the processing time. RPA can further help integrating different claim processing information from multiple sources. In context of insurance industry, it can automate the manually intensive processes like extraction of data, complex error tracking, claim verification and more; consequently, speeding up the process and creating a better customer experience.
Fraud flagging and prevention:
The essence of RPA is interconnectivity of systems and information flow between the same. Fraud flagging and identification can be facilitated at all points of processing within an organisation be it pre-sales, sales or claims. External sources of information such as an industry level data on serial offenders or other industry collated information sources can be brought together as well to tackle frauds in a better manner.
What are the Challenges of implementing RPA?
Usage of RPA in insurance sector is at a nascent stage. It’s essentially automation of a click stream or journey to speak, and other technologies like AI and ML have to support an end to end RPA solution. A solution must go beyond tackling mundane tasks to providing business analytics and insights. Most proposal forms in the Indian insurance space are handwritten and recognition of this is not yet a refined technology. One could argue a case for OCR based forms, but that is simply not practical and could be time consuming. Another challenge in this regard is, if the form is filled in local language it would be tough for an English handwriting recognition software to read. There are a few start-ups working aggressively on this space and we should see some headway soon. A lot of insurers have legacy systems and most often sitting in their independent silos with varying levels of tech preparedness and integration capabilities, thus it is often advised to get core systems in place before moving onto full automation.
The use cases are clear, but Insurers need to decide on their level of preparedness on core systems, and prioritise technological innovations as ‘must have’ and ‘good to have’. One feels AI, ML and analytics are ‘must haves’ as of now, but RPA will move into this category in the near future.
Authored by KV Dipu, Head – Operations & Customer Service, Bajaj Allianz General Insurance