With all this technology come new risks and challenges.
With all this technology come new risks and challenges. Dogan discusses recent insurtech trends, potential implications of artificial intelligence and machine learning for employee benefits, and pitfalls brokers can keep an eye out for when adopting new tools.
Launch From Original Source
Q: Do you see anything new on the horizon that brokerages could be investing in?
For decades, there has been a big push toward self-service on the P&C personal lines side, starting with basic functionality like bill payments and then moving towards more complex services like online binding and renewals. Commercial lines were slower to get started, but in recent years, we’ve seen the same pattern, starting with online bill pay and quoting, then digital applications, and gradually supporting more complex services like certificates.
More recently the big push in insurtech has been more advanced automation of complex insurance processes like underwriting and policy checking. But many insurtechs are focused on very narrow tasks, solving one step in a complex process rather than tackling an entire process end-to-end. It leaves brokers navigating how many vendors and products they must cobble together to have a complete solution for their team.
With this elevated level of excitement around automation, individual agencies often struggle to figure out how to make use of these advancements without a CIO or CTO and a large technology organization.
Q: Then would you say a successful implementation of technological tools would rely on more of a holistic approach?
As one of my colleagues at Patra often says, there are solutions in the technology world that should be companies, there are solutions that should be products, and there are solutions that ultimately should be features. Currently, I see a lot of companies that are more accurately positioned to be products or features.
For broad adoption to take hold, consolidation is required, where broker agencies seek more than multiple point solutions for narrow problems and look to strategic partners with broader solutions that can seamlessly integrate into an agency’s systems, operations and overall growth goals.
Q: To narrow in on employee benefits a little bit more, are there technological tools now that could be feasibly used to automate or streamline processes like eligibility transactions or agency management system updates?
Employee benefits, like all insurance, is a document-heavy industry that relies on manual review and work. Copying and pasting of data across multiple systems, sending and reviewing forms across multiple organizations, and “swivel chairing” are process inefficiencies that cause backlogs and lead to a high risk of error.
Artificial intelligence and machine learning have proven highly efficient and accurate with identifying commonalities and differences across data sets. A solution for the EB space that could review an employee roster in an HR system and in the carrier system to identify gaps across these data sets would be a huge improvement in today’s world, eliminating thousands of hours manually keeping this data in sync.
Q: What are some pitfalls when it comes to these tools?
The biggest pitfall is assuming that implementing some technology instantly creates a complete solution, wherein reality some manual work is typically still required. For example, software may automate steps three, five and seven of an eight-step process, but the other steps still require human oversight or intervention. Having the organizational maturity to integrate partial automation with human workforce management is also important.
Q: Could data protection laws pose an issue when it comes to the use of these types of technologies?
The insurance industry has tight data regulations relating to any personal identifiable information—and HIPAA is, of course, a whole other level. Automation techniques like artificial intelligence and machine learning rely on being “trained” using real data to identify patterns. For a broker partnering with a small insurtech startup, there are some risks with asking the provider to review all the data within a book of business by feeding it into their model.
But this risk is not unmanageable. To comply with data protection and privacy regulations, it’s best practice to de-identify the data before training machine learning models. For example, when uploading claims or policy information, ensure it’s scrubbed of any data that would identify an individual person. Additionally, verify you have the right protections to safeguard against transferring this information to a third party, as well as the legal and data security infrastructure to manage the information responsibly.
About Patra
Patra is the leading provider of technology-enabled outsourcing services to the insurance industry with helping retail agencies, MGAs, wholesalers, and carriers modernize and adapt to the evolving needs of their customers. Patra provides scalable and technology-driven outsourcing solutions, along with valuable data-driven insights that empower insurance organizations to stay ahead of the competition. Patra optimizes insurance processes through an innovative PatraOne platform, helping insurance organizations as they sell, deliver, and manage policies and customers through our PatraOne platform. With a global team of over 5,000 skilled professionals in data-secure environments, Patra empowers agencies, MGAs, wholesalers, and carriers to achieve profitable growth and enhance organizational value.