June 2, 2021
RPA in insurance: let bots take care of paperwork
RPA Business Analyst
Fostered with massive investments and fueled by Murphy's infamous law ("Anything that can go wrong will go wrong"), the global insurance market was valued by Statista at around $ 5.8 trillion in 2020. After a brief decline directly attributable to the COVID-19 pandemic, the trend for the next few years should be positive and lead to exceeding $6.9 trillion by 2023.
The other side of this steady growth is that insurance companies are increasingly burdened with massive amounts of time-consuming and dehumanizing paperwork, traditionally handled manually by stressed-out employees.
Luckily, the rise of the latest automation technologies has made it possible to delegate a growing number of tasks to machines and allowed human professionals to focus on more motivating and high-value duties, such as business analytics, data-driven decision-making, and customer care.
One of the pillars of this new approach, which is set to augment virtually every branch of the modern economy, is known as robotic process automation (RPA).
Robotic process automation involves the deployment of software bots programmed to mimic and replace humans in performing high-volume and repetitive functions such as filling out forms or sending emails.
These virtual helpers are immensely faster and more precise than real operators. According to our partner and automation leader UiPath, the implementation of RPA in insurance can lead to a 100% accuracy in data entry, 50% shorter call times in customer service call centers, 5x speed increase in performing some key activities, and significant cost savings.
RPA solutions are pretty versatile as they can take different forms based on the purposes for which they are developed and implemented. Old-gen bots, for example, follow a strictly rule-based approach and can handle structured data like numbers quite well. On the other hand, they may have a hard time facing exceptions and extracting unstructured information from sources like images and videos.
That's why today's trend is to combine RPA tools with artificial intelligence and cognitive automation technologies so that bots can gain a deeper understanding of the data they process and perform more advanced tasks.
However, such enhancements are not always necessary, as the insurance industry is highly bureaucratized, and, as a result, a relevant portion of its business workflows is rather straightforward and follows a pure "if/then" logic. This makes RPA adoption particularly useful even in its most basic configurations, especially when it comes to carrying out tasks such as:
All of these processes can be automated with relative ease and low investment, thanks to the non-invasive and highly flexible nature of RPA. In fact, bots can be integrated with legacy systems without changing or replacing pre-existing setups, which would require time, money, and know-how.
Despite the clear benefits of implementing robotic process automation in insurance, the statistics on the actual adoption of this technology are quite mixed.
Looking at the BFSI macro-group as a whole, which includes the banking, financial, and insurance sectors, the stats are very promising. According to Grand View Research's Robotic Process Automation Market Report 2020, the BFSI segment ranked #1 with a share of more than 29% of the global revenue in 2019.
However, when focusing only on the insurance sector, numbers are not that encouraging. As pointed out by PwC in its 2019 Actuarial RPA survey report, this industry is constantly lagging compared to other financial services in adopting RPA, even in processes that would benefit immensely from it. For example, the actuarial function, which is an integral aspect of all insurers' operations and heavily involves data collection and reporting, would be a perfect candidate for automation.
Fortunately, this trend may soon reverse. Based on Juniper Research’s predictions presented in its 2019 Robotic Process Automation in Telecoms and Insurance study, insurers will increase their spend on RPA solutions to $634 million by 2024, up from $184 million in 2019. This means that we may see an investment growth of 245% over five years.
North American and European insurance companies are expected to lead the market, with over 65% of providers investing in robotic process automation services by 2024 to improve their overall efficiency, cut operational costs, and remain competitive.
The range of RPA use cases is growing relentlessly in virtually every sector. The insurance industry, despite the delays we have previously described, is no exception to this positive trend.
According to PwC's aforementioned report, RPA's most popular use cases in insurance mainly involve some sort of data processing and reporting tasks. For example, we can count on bots for sending outputs (usually emails), running ETL (Extract-Transform-Load) processes, creating audit trails, compiling reports for stakeholders, and acquiring data from Oracle, SAP, or other enterprise systems.
Other important and fairly widespread use cases relate to activities that are not exclusive to the insurance sector but common among all industries, namely finance and accounting processes such as reconciliations, GAAP reporting, and so on.
After this premise, let's take a look at some of the most interesting ways to leverage RPA in insurance.
The first process of our overview—claims handling—probably covers one of the tensest moments between customers and insurers, as it's the stage where the customer asks for compensation right after an unfortunate event. In such a context, managing the situation as quickly and efficiently as possible is critical to delivering a great customer experience.
However, entering customers' data into the corporate database to process claims or sending quotations manually can be challenging and time-consuming. As you may expect, manual processes are error-prone, especially when it comes to collecting and entering data in multiple formats and from different sources. Among them, we may deal with pictures of damaged vehicles, police reports, medical records or certificates, and many more based on the type of policy.
Not to mention the fact that data collection is often delegated to outsourced personnel, increasing the number of steps in the entire process and, consequently, the error rate.
What if we leave this responsibility to robots? UiPath's experience can provide us with some useful insights on this. As described in one of its case studies, a leading global insurance corporation has implemented RPA in its business workflows, leveraging robots to speed up data extraction from customer correspondence and match it with the right claims forms.
While employees previously spent an average of four minutes on a single matching, now a bot accomplishes this task in 42 seconds. After automating this activity and several additional functions, the company was able to save 18,000 person-hours and about $195,000 in six months.
Similar results are certainly impressive and promising, but bots’ performance can be further improved by combining them with artificial intelligence or computer vision, which broaden the range of sources from which data can be extracted (including photos and videos).
The second business function that greatly benefits from bots' data extraction and processing capabilities is underwriting. This process involves gathering a significant amount of information from many different sources to estimate customers' risk rates with the utmost accuracy and set the premiums before taking out an insurance policy. The whole activity can take up to two or three weeks.
RPA can boost data collection and data entry into the corporate systems to speed up underwriting. Subsequently, the most relevant information may be extracted to compile a wide range of reports (including loss runs containing a customer's entire claim history) and provide product pricing recommendations.
Again, bots can be enhanced with artificial intelligence, especially machine learning solutions, to acquire deeper analytical skills. ML algorithms can analyze historical data, detect patterns and relationships between them, and forecast customer exposure with increasing accuracy. This allows a more precise and personalized estimate of insurance premiums.
Did we mention robots taking care of the paperwork? Well, policy administration is the real bottomless pit of the insurance bureaucracy. This makes it a perfect field in which to deploy RPA.
Even before the advent of robots, insurance companies already relied on traditional policy administration systems to streamline their workflows. These tools have certainly proved useful, but they are also quite expensive to maintain and update. Furthermore, they are not scalable as quickly as they need to be to meet constantly growing customer demands.
This is where RPA comes into play. Robots are fast, efficient, and highly scalable. They can take care of many back-office functions including form registration, payouts, account updates, and policy cancellations. They are also very efficient in classifying cases by type and processing or routing requests related to insurance policy reviews.
Another important aspect of administration involves fiscal and accounting processes. Specifically, we're talking about tax reporting as well as payment processing and invoicing.
Once again, RPA bots can help out with key data extraction and form filling faster and more accurately than humans, avoiding unpleasant delays or penalties.
We shouldn't forget that carrying out administrative activities with proper accuracy is also an essential element of regulatory compliance, as it deals with sensitive issues such as personal data (think about HIPAA privacy rules or PCI standards) and constantly evolving tax legislation. In this regard, a tireless robot with a close-to-zero error rate may be an excellent assistant for administrators and accountants.
One last interesting application of RPA in insurance relates to customer service, and we may consider it as a direct consequence of all the previous use cases. After all, speeding up business procedures and reducing the error rate via RPA robots is the first step to improve customer care's overall quality.
A clear example of customer service automation with RPA is offered by PZU, one of Europe's top insurance giants in Europe. This historic Polish company was struggling to maintain high standards of customer experience in the face of a steadily growing customer base. By adopting RPA, agents have benefited from a considerable reduction in bureaucratic workload while being able to focus on front-office activities more.
Speaking of numbers, robots have led to an overall improvement in consultants’ performance and a 15% increase in the number of decisions issued per person just two months after the implementation. Also, call center times have been reduced by 50% (from an average of six to three minutes) and data entry accuracy has risen to 100%.
The potential of RPA in insurance is huge, but far from being unleashed on a massive scale. However, many analysts predict a positive trend in terms of implementation rates in the years to come. According to Novarica's 2020 Emerging Technology in Insurance report, RPA has already been adopted by more than 40% of the companies surveyed and continues to expand.
To catalyze this structural shift, insurance companies will need to take into account a handful of key factors, including:
Taking these steps will allow insurers to maximize the positive impact of RPA and further expand the spectrum of bots' applications.
RPA in real estate is a valuable ally but still far from its maximum potential. Find out how to leverage it to the fullest.
Learn how the application of RPA in healthcare can alleviate the administrative burden and increase operational efficiencies.
Learn the most potent use cases for RPA in banking and figure out the best implementation strategies with Itransition.
Discover the contribution of RPA to finance and accounting, its top use cases, and the good practices to implement it properly.
Learn about the most relevant RPA use cases by industry and some limitations and requirements to take into account.