The pros & cons of automated underwriting systems

The insurance industry is often seen as a dinosaur with very few changes over the years. However, the situation is very different under the surface, with new technology, artificial intelligence (AI) and Digital Coworkers making a massive difference to the insurance underwriting process.

Robotic Process Automation (RPA) has led to significant cost savings and efficiencies, but the introduction of AI and Cognitive Process Automation (CPA) has taken automation to a whole different level. Otherwise known as Digital Coworkers, AI-fuelled bots carry out many mundane and monotonous tasks associated with risk rating.

The automated underwriting system of today bears little resemblance to that before the turn of the century. This multi-faceted solution has many benefits, but it’s also common for carriers to have concerns.

Pros of automated underwriting systems

RPA has been around for several years. It is fair to say that the insurance industry, like many others, was initially wary and took time to embrace such technology. However, once major insurance company decided to bite the bullet, it was apparent that the rest would have to follow to maintain a competitive edge.

While insurance underwriting covers a whole host of different subjects, the concept is the same. Information is received, the risk is calculated, and the relevant insurance policy is presented. When using Digital Coworkers, some of the main benefits include:

Faster turnaround

While time is often of the essence in any business, you need to appreciate speed against accuracy when introducing new technology. This means that your Digital Coworker needs to be as efficient and fast as possible without losing accuracy. Taking a conservative view, the straight-through processing rate for cognitive automation is more than 95%. Compared to human error rates of between 10% and 30%, it is fair to say the benefits are enormous.

Historically, many RPAs were created with specific tasks in mind, although there was limited connectivity. Digital Coworkers with inbuilt CPA and AI can manage the underwriting process from application to delivery and everything in between. Historically, many of the more repetitive mundane tasks were outsourced to overseas third parties. This option may have been cost-effective, but there were limited efficiency gains with a potential risk when giving third parties access to client data.

See also: Insurance Rating Software vs. Robotic Process Automation in Insurance

More engaged employees

There is no doubt that engaged employees are more proactive when their work is challenging and rewarding. Consequently, where their role involves repetitive actions that require little grey matter, it is easy for them to become disengaged. A recent poll by Gallup found that a staggering 48% of America’s working population were actively seeking alternative employment. These are employees who are either not engaged or disengaged from their role in the workplace.

The use of Digital Coworkers means that employees can be utilized elsewhere to fulfil their potential and enhance their value to the company. It is difficult for anybody carrying out repetitive tasks to see any enhancement in their career. There is also the fact that non-engaged/disengaged employees are more likely to be tempted to leave for pastures new. Disruption to business operations, let alone the added cost of recruiting, can considerably impact the bottom line.

See also: Balancing Employee Engagement and Digital Disruption with RPA

Easier new business & renewals

Many industries seem to focus on attracting new business while often taking existing customers for granted. In the insurance sector, existing business is hard to keep but vital to the lifeblood of the company.

It is relatively easy to automate the insurance renewal process using Digital Coworkers and take a more proactive approach to new business. New, enhanced optical recognition technology allows Digital Coworkers to scan structured and unstructured insurance documents, identifying specific information and taking the appropriate action for clients. This includes requesting loss runs and other underlying policies, cross-selling additional lines of business, or even collating policy materials and signed applications.

The renewal process tends to be relatively straightforward. However, this ability to be fully automated with minimal human interactions creates vast time and cost benefits. In addition, due to the AI nature of Digital Coworkers, changes to circumstances, missing data and even potentially fraudulent activity would be flagged and addressed.

Insurance companies only have a handful of opportunities to communicate with their existing policy holders each year. When cognitive automation handles the time intensive, mundane work, it creates more opportunity for human-to-human connection and interaction, that can be spent building relationships, educating policyholders about new products, and cross-selling additional lines of business.

More focus on new lines of business

The opportunity to remove employees from monotonous, repetitive tasks to more proactive and engaging activities has many benefits. First, it allows insurance companies to enhance their customer-facing services, something which customers are demanding, and be more proactive regarding new business opportunities.

A proactive incentivized workforce would be more driven to address complex issues to benefit the overall business. This may include commercial partnerships, third-party arrangements or new business lines involving complex enterprise risks. Over time, elements of these activities may also be automated with the use of digital coworkers. This will enhance efficiency, crystallize cost savings and improve profitability – the cycle continues.

Cons of automated underwriting systems

While no system is perfect, programmable Digital Coworkers with the ability to think themselves are a huge benefit. However, many of the potential cons associated with automated underwriting systems involve early technology such as RPA, which does not include advanced components such as AI, Machine Learning, and Intelligent Character Recognition. Therefore, it is essential to be aware of possible downsides to automated underwriting approval systems and appreciate recent advances.

RPA was heralded as a game-changer in its time and had a considerable impact on business costs and efficiency. However, it is essential to clarify that this first generation of automation isn’t a viable solution when compared with cognitive automation.

RPA processes are rigid

Comparing stand-alone RPA technology to cognitive automation is akin to comparing night to day. Historically, due to the inflexible nature of RPA systems, it was relatively easy for if-then statements to be broken. In some circumstances, the potential knock-on effect could deem RPA systems to be more detrimental than helpful to a business. When business owners lose confidence in automated systems, this isn't easy to regain.

The use of AI-fuelled Digital Coworkers, meticulously trained before release, offers a degree of flexibility and on-the-job learning never seen before. This has removed the inflexible nature of historic RPA technology while at the same time enhancing the value of cognitive automation to any business. While many believe the ability of bots to think for themselves is futuristic, and many years down the line, this is not the case. It is here and happening today across the insurance industry.

See also: Accelerating Policy Submissions, A Case Study

Upfront costs can overshadow long term benefits

Historically, RPA is a cost-heavy investment. Insurers would buy a software license for a few thousand dollars, and then quickly become overwhelmed when fees for additional licenses, developers, maintenance, cloud security, and more start to add up. When all is said and done, the time and payroll spent on development and updates is more than the time and money saved with RPA.

Here at Roots Automation, we take a hands-on approach to creating the optimal Digital Coworker for each business. Working with customers, we identify the underwriting insurance processes for automation, train the bots, and release them into the business environment within 6 to 8 weeks.

When we deliver your Digital Coworker, they will quickly bolt onto your existing systems and take on their new role. There is no expensive ongoing introduction, no involvement of your IT department or in-house training program. Even though we deliver the finished article, don’t forget that your bot will start learning using the latest artificial intelligence from day one.

On average, a Digital Coworker can carry out the work of four to eight humans. The ongoing costs equate to the average salary of one worker, therefore, creating substantial cost savings from day one. Moreover, the payback period tends to be measured in months, not years, with an average 250% return on investment.

See also: Roots Automation CEO and Co-Founder Chaz Perera for Forbes, The Bot Revolution Stalls as AI Fails to Scale.

Partially optimized automation is a false economy

The first generation of RPA has one massive failure: the bots don’t understand how to fail gracefully. When the bot is unable to log in to a system, can’t find a contact, experiences a slow down, or runs into any general issue – it stops completely.

Because the bot halts the entire process until a human can troubleshoot the issue, the organization loses valuable uptime and production. Additionally, traditional RPA never learns. No matter how many times the bot runs into that issue, it will never learn to solve it on its own.

Setting the foundations for AI learning is essential for long-term success. On average, it takes between six and eight weeks from initial contact to the delivery of your pre-programmed personalized Digital Coworker. During this period, we will introduce an array of different characteristics (from our vast library) to ensure that every bot is unique and focused on your requirements. Then, when delivered, your new "employee" will be ready to hit the ground running.

Fail to prepare, prepare to fail is an apt description of our philosophy and approach to customer services. Unless Digital Coworkers are optimized, focused and able to "learn on the job", you will find any new automated technology will quickly become a short-term band-aid. As a result, the value to your business will decrease, ongoing costs will increase, and the initial euphoria will soon dissipate.

Employee acceptance is crucial

In earlier renditions of RPA, executives were so excited about the technology that they forgot to instil that excitement in the workforce. Instead, employees began to fear the bots, rather than accept them and maximize their own internal growth.

For companies to get the most out of automated underwriting systems, everyone needs to be on board. Digital Coworkers should be seen as complementary to the human workforce, removing tedious, repetitive tasks, leaving them to focus on more proactive and engaging activities. It is crucial that your employees are engaged as a disengaged employee can cost your company around 18% of their salary per year.

See also: No, RPA Won’t Replace Phil in Accounting

Summary

There are many elements of the insurance underwriting process which have been automated in isolation. Robotic process automation insurance has been around for some years, but the introduction of AI and CPA has made an enormous difference. We can train Digital Coworkers to carry out specific tasks while also taking a proactive approach to the use of AI. This means that while the initial efficiency and cost benefits are significant, they are constantly enhanced by AI. Less time on mundane, repetitive tasks, more time for employees to be fully engaged and proactive creates a win-win scenario for everybody.

If you’re ready to explore an automated underwriting solution, you can experiment with our on-demand demo. You’ll teach a Digital Coworker to rate a Directors & Officers liability policy, then watch as it learns from you and begins rating all on its own.

You can access the on-demand demo here.