Using AI for effective claims processing
One place that has desperately needed automation is data processing.
Competitive market pressures and growing customer expectations are increasingly challenging the insurance industry to innovate its processes or be left behind. Despite this, many insurers are still relying on outdated, manual processes to handle claims and process data that naturally flows through their organizations. This results in delays, errors, customer frustration and other issues that can impede growth.
Part of the reason for holding on to traditional methods is that modernization not only requires technical investments but also a new way of thinking. In addition, many firms are dealing with a legacy tech stack that has been cobbled together over several decades.
Yet the tide is turning, and leading insurance firms are following other tech-enabled industries and leveraging automation as the blueprint for unlocking new efficiencies. Automating these manual processes is enabling them to speed up claims processing, increase resource capacity and deliver the level of service and innovation the market demands.
The first mile in claims processing is the one automated
One place that has desperately needed automation is data processing. Ironically, data entry —the first step in the first mile of claims processing — is one of the last areas to become automated, and remains one of the ripest for disruption.
Insurers today continue to rely on teams of data keyers or outdated technology to read and transcribe critical information from a page and get it into a format that can be used by various systems. This is particularly burdensome for areas like claims, which require multiple pages (oftentimes handwritten) and an increasing amount of data to assess settlements, issue payments and service customers.
Despite the talk of digital-first, paper won’t be going away anytime soon, and even when documents are generated electronically, they are typically incompatible with internal databases and often need to be manually processed and re-entered. As a result of all this, businesses continue to spend billions of dollars and hundreds of thousands of hours on manual data entry to classify and extract information from documents and get it into a format for downstream processing, servicing and front-line customer interactions.
Turning to a smarter form of data entry
While automated data entry has provided some efficiencies, legacy technologies still struggle to accurately read the most challenging documents — which might contain handwriting or cursive, have been faxed, mailed or scanned, or photographed by a smartphone and uploaded to a portal. In the absence of reliable results, organizations have continued to rely on people to check the outputs, fix errors and rekey data — an inadequate, short-term solution.
Fortunately, technologies are available today to address this challenge. Increasingly, innovative firms are turning to Intelligent Document Processing (IDP) technology, which leverages AI and related technologies like machine learning, to automatically capture, classify and extract unstructured data from various documents. The right IDP solution can read messy, real-world documents with context and continues to learn and train itself in response to the data that it’s exposed to. The result is a high level of accuracy, speed, efficiency and business value.
By implementing this type of AI-driven solution, claims processors can reduce clerical errors that can result in inaccurate customer data or slow the claims payment process and improve customer experience in the process. After all, when filing a claim — oftentimes a period of great stress — customers want their insurers to alleviate their worries and burdens, not add to them. With less manual and administrative work, insurers and their employees can focus on these higher-order initiatives, and more complete and accurate data lays the groundwork for greater downstream AI, analytics and strategic digitization initiatives.
A 2019 Everest Group study reinforces the growing role of IDP across industries, estimating that IDP is expected to grow 70-80% in the next two years “and will rapidly accelerate as enterprises seek to harness the power of artificial intelligence (AI) technologies to improve compliance and governance and reduce the overall cost of processing huge volumes of information.”
- Don’t automate broken processes. Before automating anything, it’s important to review your processes to make sure they are efficient and sound. Look at how your data enters the organization, where it moves, what systems it needs to work with, and how it is used to achieve business goals and service customers.
- Recognize that people will always be part of the solution. People aren’t perfect, and neither are machines. To solve challenging, real-world problems (like reliably capturing the data contained in documents), machines need supervision. Once you understand that, the key is to investigate how a solution involves employees in the process. Invest in a system that reliably automates the repetitive, tedious tasks and is smart about when to involve your employees, giving them more time to focus on customer service, upselling customers or providing strategic value in other ways.
- Focus on incremental change. Regardless of the technology being added, it’s important to implement change incrementally so that you can finetune new processes, allow employees to become accustomed to the new way of doing things and adjust as needed.
According to a McKinsey report, “over the next decade, next-generation capabilities have the potential to completely transform the claims process.” Thanks to new automation technologies powered by AI, both customers and employees alike will be freed from the data bottlenecks and outdated processes that have been a key challenge to more strategic business growth.
Charlie Newark-French (charlie@hyperscience.com) is the COO at Hyperscience, the automation company that enables data to flow within and between the world’s leading firms in insurance, financial services, healthcare and government markets.
Related: