RPA to IPA: How Insurers Can Add Intelligence to Their Automation Efforts
From property and casualty, through life and annuity, insurers of all stripes need to transcend task-based robotic process automation and holistically embrace more powerful intelligent process automation to improve their performance.
In an environment characterized by low interest rates, market overcapacity, tougher competition and growing regulations, insurers are increasingly looking to automation as the answer. For many, robotic process automation (RPA) has appeared to be the best bet for cutting costs quickly. One would think that in a data-driven industry, with high-volume, repetitive manual processes, RPA would be the ideal solution. Unfortunately, that is not entirely true. For insurers that applied RPA as merely a cost-cutting tool capable of fixing any process, the technology turned out to be an automation hammer when their processes actually needed a more refined Phillips-head screwdriver.
Intelligent process automation (IPA), meanwhile, enhances software bots with cognitive technologies that mimic human perception and judgment such as artificial intelligence (AI) and machine learning (ML). But IPA isn’t proving to be an easy transformation for the industry either. In our experience, too many insurers see robotic automation as a silver bullet for reducing headcount and improving profits.
Insurers wading through tough automation decisions must keep in mind that crossing the chasm will not be easy for them and their employees. But companies that develop a holistic strategy that combines RPA with cognitive technologies can cut costs now and achieve benefits that position them for success in the years ahead.
Why insurers must get automation right
Both life and annuity (L&A) and property and casualty (P&C) insurers seeking the benefits of automation need to develop a holistic, customer-centric approach to IPA. Over the next few years, RPA and more advanced cognitive technologies will be infused into a broad array of business processes.
For example, if ML is added to a bot, the software will continually improve at performing a task. Guidewire’s predictive analytics tool uses both RPA and ML to assess claims in real time, flagging unexpectedly large or suspicious submissions, fast-tracking small claims and managing workflow. Amelia, IPSoft’s virtual agent, is used at insurers such as MetLife and Credit Suisse to combine ML with natural language processing (NLP) to make decisions based on real-time conversations and suggest ways she can improve her performance.
IPA will allow insurers to automate not just mundane tasks or parts of a process, but also end-to-end business operations. Companies that fail to effectively automate their processes will have difficulty achieving long-term growth.
Finding the right IPA capabilities
Comparing the many IPA tools available to insurers can be overwhelming. A good starting point is categorizing IPA technologies into the following three main capabilities, realizing that as tools advance, these abilities will be integrated:
Developing an effective automation strategy
How should insurers analyze the capabilities of new technologies to determine the best way to improve their processes and realize benefits? We advise an integrated, customer-centric approach to IPA that includes the following steps:
Analyze the process you want to improve holistically, from end to end. Senior executives making automation decisions need a deep understanding of the process from end to end. Take a customer-centric view of the process whenever possible, so that automation enhances the customer experience.
Analyze how various types of automation can enhance your organization’s process. In addition to a deep understanding of the process, executives must also fully know the capabilities of the automation tools on the market.
Optimize your process as you automate it. We have found that roughly one-third of processes being reviewed for automation require changes beforehand. Insurers should eliminate unnecessary work and/or actors wherever possible. A simple change such as making an optional field on a form mandatory can allow a bot to process it.
Integrate front- and back-office processes, putting customers at the center. A holistic approach to automation integrates the customer-facing part of a process with the back office. Automating a process may reduce costs or ease workloads, but it’s important to consider the impact it will have on the customer.
Prioritize your processes in alignment with business priorities. The top goal for many insurers tackling automation is cutting costs, largely by reducing employees. This minimizes many other automation benefits (see below) that can foster long-term growth.
Look beyond cost reduction. Benefits offered by automation go beyond cost reduction. These include: improved customer experience, improved quality, enabling sales through personalized marketing and improved compliance.
Establish an automation center of excellence. An automation project should be undertaken as part of an integrated strategy supported by both the business and IT functions. An automation center of excellence (CoE) will help ensure that automation is just one aspect of a mature continuous improvement strategy.
Using systems of engagement to know your customer better
Most data management initiatives are driven primarily by regulatory mandates and the desire to cross-sell, and on building a single view of the customer across lines of business reflecting transactions, investment holdings, and basic, compliance-driven demographic information. This data has traditionally been the primary source of client information and the basis for advisors’ formulation of client investment plans.
Firms that have completed these integration efforts now seek opportunities to leverage their systems of record data for more strategic purposes, such as building analytics to respond to or predict client behavior. These are ambitious goals. In our experience, however, it is unproven that systems of record data alone can generate enough insights to lead wealth management firms to these end points.
A more promising path forward is to enrich this core data set with additional elements sourced from various systems of engagement. Examples of these sources include data already at hand within enterprise systems such as customer relationship management (CRM) records, service logs and call-center interactions, and data derived from natural language analytics.
Using RPA for a quick performance boost
RPA can quickly add value for a company by performing a wide range of repetitive, structured tasks governed by simple rules. A software bot can perform hundreds of sequenced actions day and night, with no errors or biases.
We have found that roughly one-third of the insurance industry processes we assess for clients can easily be automated by robotics, offering a fast return on investment. For example, bots can dramatically reduce the time it takes to process a claim by inputting FNOLs, notifying loss adjusters and assigning the case to claims handlers. However, close to 70% to 80% of RPA investments by insurers that we observe tend to focus on making claim and policy processing more efficient.
New business/underwriting also contains many processes that can benefit from bots taking over human tasks, such as gathering and processing applicant data from internal and external sources to assess risk. An Oxford University study found that insurance underwriting tasks are among the most likely to be subsumed by bots and AI.
With the additional insight that systems of engagement data offers, advisors will have the wherewithal to understand clients beyond wealth tier and life stage — the traditional methods of market segmentation. They will be able to generate forward-looking hypotheses and assess predictive outcomes. Informed by data beyond systems of record, advisors can connect with clients in a timelier and personalized way.