Personal lines insurers are investigating emerging technologies and developing strategies and plans related to individual new technologies. Technology is advancing so rapidly that it is even difficult to define what should be considered an emerging technology. For the past several years, SMA has been tracking 13 technologies that many consider to be emerging. These include technologies such as autonomous vehicles, AI, wearables and the Internet of Things. In our recent research, five of these technologies have emerged as “power players” for personal lines insurers, based on the level of insurer activity and the potential for transformation. The specific plans by insurers for these and other technologies are detailed in the SMA report, Emerging Tech in Personal Lines: Broad Implications, Significant Activity.
Some big themes for emerging tech in personal lines stand out:
Artificial Intelligence dominates. AI is often a misunderstood and misused term. However, when specific technologies that are part of the AI family are evaluated, much activity is underway – by insurers, insurtech startups and mature tech vendors. Chatbots, robotic process automation (RPA), machine learning, natural language processing (NLP) and others are the subjects of many strategies, pilots and implementations.
The Autonomous Vehicle frenzy is cooling.There is still an acute awareness of the potential of autonomous vehicles to dramatically alter the private passenger auto insurance market. But there is also the realization that, despite the hype, the transition is likely to be a long one, and the big implications for insurers are probably 10 or more years out.
The IoT is going mainstream. Discussions continue about the transformational potential of the IoT for all lines of business. But rather than just talking about the possibilities, there is now a great deal of partnering, piloting and live implementation underway. We are still in the early stages of incorporating the IoT into strategies and insurance products and services, but their use is becoming more widespread every day.
UI Options are dramatically expanding. The many new ways to interact with prospects, policyholders, agents, claimants and others should now be considered in omni-channel plans. Messaging platforms, voice, chatbots and more are becoming preferred ways to communicate for certain customer segments.
Certainly, other trends and much emerging tech activity are happening outside these main themes. Wearables, new payment technologies, drones, blockchain and other technologies are being incorporated into strategies, pilots and investment plans. The next few years promise to be quite exciting as advancing technologies spark more innovation in the industry.
Historically, technology adoption within commercial lines organizations has been met with a wall of push-back, largely related to commercial lines being wrapped in a cloak of “art versus science” thinking. Because of risk and product complexity, commercial lines organizations believed that only highly trained and seasoned humans could be involved with processes and decisions.
Additionally, due to the predominance of large, enterprise-scale projects, characterized by protracted ROI exercises and IT resource allocation exercises, past technology choices generally brought out the “yeah buts.” (What are the “yeah buts”? This is the response to enterprise technology options, to which commercial lines product and underwriting heads promptly responded – “yeah, but that doesn’t work for us.”) In many cases, this was not an inappropriate response because of risk and product complexity. But, at long last, there is a change afoot, and it lies within emerging technologies.
SMA has been conducting research and surveys around emerging technology since 2010 to gain insight and understanding of insurance industry adoption and spending. In the past, results have predominantly trended across the P&C industry. However, the recent 2018 results reveal clear differences between commercial lines and personal lines organizations. Even more exciting, commercial lines product segment and transactional differences are emerging. As the phrase goes: Vive la difference!
So, what does all this mean? SMA’s recent report, Emerging Tech in Commercial Lines: Ramping Up Adoption, covers eight emerging technologies that hold great promise for commercial lines organizations: artificial intelligence (AI), new user interaction technologies, the Internet of Things (IoT), drones, blockchain, autonomous vehicles, new payment technologies and wearables. How are commercial lines organizations viewing these technologies? Here are some examples that show emerging technologies are being viewed uniquely by varying commercial lines segments and processes:
AI – This technology garners the highest percentage of implementations of all the emerging technologies by almost twice the other categories, with 26% indicating so. Investment in AI exceeds the next closest emerging technology by more than 24 percentage points. The difference: It can drive straight-through processing for small business and simple specialty lines and support complex decisions for middle market, large national/global accounts and complex specialty lines. “Art versus science” well managed!
New User Interaction Technologies – This is another technology that is affecting small commercial lines as this product segment goes digital. But 67% of all responders see the value in customer experience, regardless of product segment, and 50% are focused on policy servicing.
Blockchain – While personal lines organizations are generally assessing the applicability of blockchain, commercial lines have found use cases and pilots. 42% of survey respondents believe that policy servicing and billing are the significant value areas. Global and complex lines of business are the first target areas.
Other emerging technology examples and spending projections can be found in SMA’s commercial lines report. But the big takeaway for me is that, happily, the “yeah buts” are disappearing across commercial lines of business and products as executives search for and find emerging technologies that can improve business outcomes. Because of the way emerging technologies are being delivered by incumbent and insurtech providers, discreet value choices can be made without having to launch enterprise-level projects. Vive la difference!!!
Artificial intelligence stands out as one of the hottest technologies in the insurance industry in 2018. We are seeing more insurers identifying use cases, partnering and investing in AI. 85% of insurers are investing time, money and effort into exploring the AI family of technologies. The focus is not so much on the technology itself as on the business challenges AI is addressing.
For companies looking to improve internal efficiency, AI can assist through machine learning.
For those working to create a dynamic and collaborative customer experience, AI can assist with natural language processing and chatbots.
For those seeking an edge in data and analytics, AI can help to gain insights from images with the help of machine learning.
Through our annual SMA Innovation in Action Awards program, we hear many success stories from insurers throughout the industry that are innovating for advantage. AI was a key technology among this year’s submissions. The near-ubiquity of AI was even more obvious among this year’s insurer and solution provider winners, many of whom are leveraging some type of AI to solve widely variant business problems. They have provided some excellent use cases of how insurers are applying AI and how it is helping them to succeed.
Two AI technologies, machine learning and natural language processing, fuel Hi Marley’s intelligent conversational platform, which West Bend Mutual Insurance piloted in claims with outstanding results. The Marley chatbot lets West Bend’s customers text back and forth to receive updates, ask and answer questions and submit photos. Its use of SMS messaging means that communication can be asynchronous and done on a customer’s own schedule, eliminating endless rounds of phone tag.
Natural language processing allows Marley to communicate with customers in plain English – both to understand their needs and to respond in a way that they will understand.
Machine learning enables Marley to continue to improve. The platform analyzes every conversation and uses it to shape how Marley responds to specific requests, refining its insurance-specific expertise for future interactions.
Natural language processing is also a critical tool for Cake Insure, a digital workers’ comp MGA with a focus on making the quoting experience easier for direct customers. One of the hurdles that would-be customers had to overcome in obtaining workers’ comp coverage was answering a multitude of questions regarding very specific information that a layperson is unlikely to know about or understand.
NAIC codes, for example, are required for every workers’ comp policy, but the average small business owner would be baffled if asked about them. Cake circumvents this by asking usera to type in descriptions of their companies in their own words. Natural language processing parses this plain-language description and searches for its approximate match in the NAIC data sets. This back-end process occurs without the user’s awareness and without exposing potentially confusing content.
As with Hi Marley’s chatbot functionality, natural language processing is paired with machine learning to improve its ability to respond to specific phrases and content.
Machine learning can also be deployed in conjunction with other AI technologies. Image analysis and computer vision are combined with machine learning in Cape Analytics’ solution, which can automatically identify properties seen in geospatial imagery and extract property attributes relevant to insurers. The result is a continually updated database of property attributes like roof condition and geometry, building footprint and nearby hazards.
Computer vision helps turn the unstructured data in photos and videos from drones, satellite and aerial imagery into structured data.
Machine learning allows the solution to train itself on how to do that more effectively, as well as higher-level analysis like developing a risk condition score for roofs.
We are only scratching the surface of how AI can be applied across the value chain. The incredible variety of AI’s potential applications in insurance is difficult to overstate. QBE knows that well: It won a company-wide SMA Innovation in Action Award for wide-ranging activities in emerging technologies and partnerships with insurtech startups, but AI in general, and machine learning specifically, are their top priorities. In addition to partnering with dozens of insurtechs, QBE has also pushed itself to deploy each insurtech’s technology somewhere within its business – meaning QBE has dozens of different creative AI applications in play at once. For example, in partnership with HyperScience, QBE is improving data capture from paper documents through machine learning and computer vision.
These winners’ stories demonstrate the myriad ways that insurers are applying AI to improve business operations. Notably, its deployment helps them to significantly improve the customer experience – or, in the case of data capture, the internal employee experience. The need for this kind of seamless customer experience in the digital world cannot be overemphasized. AI, which struck many as a science-fictional concept, has proven its real-world worth by enabling insurers to transform their customer journeys and experience.
With full-scale implementations popping up across the insurance industry, as well as the pilots and limited rollouts that we have seen in previous years, it is easy to lose sight of the fact that we are seeing only the very tip of the iceberg in terms of how AI can transform the business of insurance. Applications of more advanced and advancing AI technologies, as well as the combination of AI with emerging technologies such as drones, new user interaction technologies, autonomous vehicles and IoT, are unexplored territory that is bright with promise.
See also: 3 Steps to Demystify Artificial Intelligence
This much is clear: AI will change the face of the insurance industry. In fact, it’s already happening.
For more information on the SMA Innovation in Action Awards program and this year’s winners, please click here.
To download a free copy of SMA’s white paper AI in P&C Insurance: Pragmatic Approaches for Today, Promise for Tomorrow, please click here.
In a 2017 report titled “Drones: Reporting for Work,” Goldman Sachs estimated the addressable market opportunity for drones globally between 2016 and 2020 to be $100 billion, of which the insurance claims drone market was estimated to be $1.4 billion.
And the report did not address the wider opportunities in personal and commercial property insurance: underwriting, pricing, risk prevention, traditional and virtual claims management, fraud detection and product marketing. The report also didn’t cover the use of images from satellites and fixed-wing aircraft, including streaming video.
Whatever the actual size of the total insurance market opportunity, the impact of aerial and drone images in insurance will be enormous.
Industry observers are just beginning to recognize the transformation in property insurance underwriting and claims that is emerging through advanced analytics, artificial intelligence and machine learning tied to neural networks and integrated with data from aerial and drone images.
Property claims investigation costs the industry an average of about 11% of premiums – automated inspection can reduce that expense substantially. And automated property inspection cycle times can average two to three days, compared with 10 to 15 days using traditional methods – lowering costs and increasing customer satisfaction.
Providers will transform the property insurance industry through the convergence of these sources of better images, expanding numbers and types of connected home technologies, customer self-service and aggregated property risk data (historic and real-time).
Follow the money
Venture and private equity investment activity in emerging technologies is a good indicator of potential growth opportunities – these professionals typically engage subject matter experts and conduct deep market research and diligence in a highly disciplined and proven evaluation process prior to investing. Since 2012, almost $2 billion has been invested in more than 370 drone company deals, and the current run rate is more than $500 million in announced deals annually, according to CB Insights research, which states that ”19 of the 24 smart money venture investors have backed at least one drone company since 2012.”
Within just the past two months, four such insurance-related transactions were announced;
Nationwide Ventures made an investment in Betterview, a machine learning insurtech startup focused on analyzing data from drones, satellite and other aerial imagery for commercial and residential property insurers and reinsurers. This follows a September 2017 seed round funding of $2 million.
DroneDeploy, the world’s largest commercial drone platform, raised $25 million of Series C venture capital, bringing total funding to $56 million.
Cape Analytics raised $17 million to grow its AI and aerial imagery platform for insurance companies, led by XL Innovate.
Clearlake Capital Group acquired a significant interest in EagleView Technologies alongside Vista Equity Partners, which had purchased EagleView in 2015. (Vista also owns the majority of Solera, parent of property and auto insurance claims services and information providers Enservio and Audatex.)
In 2017, Genpact, a global professional services and insurance claims solutions provider, acquired OnSource, which provides 24/7/365 full service on-demand drone property inspection claims and settlement services across the U.S. Earlier that year, Genpact acquired BrightClaim and National Vendor, providers of integrated claims solutions to the U.S. property insurance market
In 2016, Airware, a global enterprise drone analytics company, closed a Series C round of $30 million to bring its total funding to $110 million. Early in 2016, Verisk Analytics formed the Geomni business unit to specialize in image sourcing and analysis and has since acquired a number of U.S.-based aerial survey companies and their aircraft fleets. Verisk also owns Xactware, the dominant industry provider of property insurance claims solutions and third party products. The Geomni fleet is expected to include more than 125 fixed-wing aircraft and helicopters by the end of 2018, operating from 15 hubs located throughout the U.S. Verisk expects to invest approximately $100 million in Geomni through 2018.
Competition and differentiation
The space has attracted a large number of participants in the past two years, and there are no signs of slowing. Competitors are taking innovative paths to differentiation, including: drone manufacturing, drone operating software for use by field staff and contractors, ground-based roof and wall measurement technologies and full-service, virtual property inspection and property damage reports using drones.
Insurance industry adoption and barriers
The insurance industry’s use of images from satellite and fixed-wing aircraft is fairly well-established, particularly in catastrophe response planning and claims. The North American property/casualty insurance industry has been cautious and conservative in its testing and adoption of drone use for property claims and in using aerial images for underwriting.
Until recently, FAA rules had made it onerous for carriers and industry vendors to obtain licenses and permission to use drones for property inspections. However, after extensive industry lobbying efforts, assisted by more pro-business policies, that obstacle has eased significantly, and several carriers have trained staff and hired contractors to use drones for property claims inspections. Obstacles remain, including restrictions on use near airfield perimeters and outside of operators’ line of sight.
Carriers are split into two roughly equal camps (by market share) on more recently introduced third party services that provide virtual property inspections: those that do not believe that drone image and damage identification technology is sufficiently accurate as yet to manage claims leakage as effectively as their own staff field adjusters – and those that do. Both groups acknowledge that drones are not appropriate for all property claims. Furthermore, customer satisfaction and therefore retention is thought to be higher when insurance company staff visit the property and the homeowner in person.
The future of property insurance
For claims, virtual methods of inspection will include not only drones but claims reporting that involves customers. Claim self-service, including smartphone images and video, which has seen impressive adoption and results in auto claims, is beginning to penetrate property insurance claims, particularly for reporting home interior and exterior wall damage. New, accurate 3D smartphone image measurement technology combined with higher image resolution and the expected expanded availability of much faster 5G wireless broadband will drive adoption.
Other methods of property inspection, particularly following extreme wind or hail events and catastrophes, will most certainly incorporate the use of drones, whether operated by insurance staff, managed repair network contractors or third-party inspection services. Also, autonomous drones performing roof inspections not requiring an operator on site may be expected soon.
Finally, on the property underwriting side, we expect high-resolution geospatial image data from multiple sources, artificial intelligence and machine learning to transform that process. Real-time feeds of comprehensive property attributes such as measurements and condition of roofs and other property on the target site will enable instant and more accurate pricing, quoting and binding/renewal of property insurance.
Aerial imagery, mobile technologies, artificial intelligence and computer vision will continue to transform property insurance products and processes, leading to better pricing accuracy, more profitable operations and, above all, better customer experience for policyholders.
Across the insurance industry, claims organizations have made significant progress in modernizing their core processing systems in the last several years. Typically, the objectives of these programs are to increase speed, improve accuracy and reduce risks in all phases of claims handling. Given that claims interactions are “moments of truth” in customer relationships, insurers have good reason to ensure that the experience for policyholders is smooth and satisfying at every step of the process.
No matter where insurers are on this continuum, robotic process automation (RPA) can help them achieve their business objectives while leveraging existing technology and boosting returns on previous and current transformation investments. In seeking the best path forward, claims leaders will want to consider:
Why robotics is well-suited for use in claims and how it complements other enabling technologies
Key components of the business case and value proposition
High-priority opportunities and common use cases
for deploying RPA
Applying the principles and techniques used by successful early adopters as they develop their own implementation approach
Why RPA? Why now?
RPA involves the use of virtual workers, or software robots, to perform business tasks similar to human users. The main appeal for insurers is the ability to handle high-volume and complex data actions at exponentially greater speed than in the past.
RPA is also notably flexible, which makes it both business-enabling and IT-friendly. It can be deployed alone or with other technologies across the claims value chain. For example, robotics can:
Automate discrete tasks or activities
Work in concert with other systems on transaction processing, data manipulation, communication and response triggering
Facilitate straight-through or “no-touch” processing, working alongside analytics tool sets and other cognitive technologies, such as machine learning and natural language processing
The cost of entry for RPA in terms of financial commitment and deployment requirements is low, compared with other technologies. There is no disruptive “rip and replace” with RPA; proofs of concepts are straightforward to launch, which helps IT and business leaders get past their “not another technology” reluctance. And many benefits can be unlocked without large-scale process re-engineering.
More than just overhauling the most routine administrative tasks, robotics creates capacity and expands the art of the possible in claims. While many assume robots simply replace human resources, RPA can – and should – be viewed as an enabler and a win-win for insurers and their workers.
RPA ROI: building the business case
A significant number of insurers have already implemented robotics, though few have done so at scale. ROI cycles for RPA can usually be measured in months rather than years. Most early adopters start with multiple functional “pilots” or proofs of concept that are completed in as little as 30 to 60 days. Broader, first-generation programs may take six to 12 months.
Increased capacity and focus on high-value work: Robotics can free knowledge workers from the burden of routine reporting, documentation and maintenance tasks. Instead, they can focus on areas where they can provide the most value, such as managing exceptions and dealing with high-risk and complex claims. A common approach is to use RPA to support straight-through processing for claims under a certain dollar threshold. RPA may also be used to handle basic data entry tasks for claims of any amount. Industry research has found that turnaround times for these types of claims may be reduced as much as 75%–85%, with 50%–70% of repetitive tasks effectively eliminated.
Higher quality and accuracy: Robots processing claims will no doubt be able to increase accuracy and reduce errors, whether related to sophisticated fraud or simple “fat-fingering,” for the vast majority of routine claims. Indeed, robots are uniquely qualified to assist quality assurance (QA) staff, given their ability to scan large quantities of data and transactions almost instantaneously. For example, RPA can help identify potentially fraudulent claims by flagging data outliers. Further, in the realm of compliance, RPA helps strengthen and streamline adherence to standard audit, risk, privacy and security policies and protocols.
Increased scalability: RPA is a natural solution for insurers that need to add temporary capacity to deal with seasonal spikes in claims activity or after catastrophes. The virtual workforce can scale to peak loads without overtime and establish 24/7 processing. For example, RPA enables insurers to increase the amount of new loss intake capabilities without a corresponding increase in first notification of loss (FNOL) processing staff. The easy scalability also makes RPA a highly useful tool for insurers exploring shared services models for claims.
Higher customer satisfaction: In identifying processes that can be automated, leaders should also look for opportunities to enrich the customer experience. Speed, accuracy, transparency and level of service are what matters most to claimants. RPA helps on all those fronts by allowing claims professionals to focus on the “art” of claims adjusting and customer experience, as opposed to the transactional aspects. RPA can also accelerate innovation programs in customer engagement and experience. Business rules can be configured directly into the robotics to align with customer expectations for personalization and timely communications.
Strategic data usage: The quality gains and capacity improvements from RPA enable claims teams to shift from simply processing data to exploiting it for more accurate and timely reporting and insight generation. In this sense, RPA can actually be an empowering force, rather than a discouraging threat, to a claims workforce.
RPA in action: where to start the journey
The use of robots and automation can take many forms in claims, including both customer-facing and back-office functions and tasks. The following represent the most common and promising use cases across the industry:
Streamlining vendor applications and estimating: Most current estimating processes require adjusters or others to rekey data from one form or system to another. Robotics along with enabling technology such as optical character recognition (OCR) can eliminate that duplicate effort by bridging the gap between claims systems, vendor apps and third-party estimating systems.
Capturing and managing claimant data: RPA can be on the receiving end of claims submissions, especially those that typically include photos from customers. Robots can ensure the right information ends up in the right systems and attached to the right claims. As such, they ensure human representatives have the information they need to move claims forward and respond to customer inquiries. Customers who prefer self-service also benefit when submitted information is more readily accessible.
Streamlining, automating and enhancing communications: Claimant communication remains a largely manual undertaking, requiring adjusters or other claims staff to initiate and, in some cases, monitor the process. RPA can help operationalize smart rules so the right letter (e.g., one required to be sent 30 days after a loss is reported) reaches the right claimant at the right time through the right channel. For instance, robots can pull data from claims submission forms and pre-populate letters that are typically housed in other systems and map distribution to customer preferences.
Scanning, indexing and converting forms and data: RPA has proven especially proficient at pulling data from standard fields on medical bills, from claimant name and address, to provide information to coding details. Standard in name only, these forms are a common source of errors. Similarly, RPA can transfer and convert data across older claims systems that may be used by individual product lines or regions to newer enterprise systems.
Validating payments: Conventional wisdom holds that 3-5% of claims payments are inaccurate, though no one knows for sure, given the difficulty and expense in auditing all claims. The key is robots’ ability to quickly and cost-effectively run QA on entire populations of forms and payments, rather than just a small sample. For example, rather than auditors discovering a $5,000 payment on a $500 settlement months after a customer has cashed the check, robots can flag the disparity beforehand. Further, they can help deliver the information and intelligence so that human analysts can investigate anomalies proactively.
Customer-facing enhancements: RPA can alleviate the need for time-consuming and costly adjuster input by supporting customer-friendly apps for capturing photos of fender-bender car accidents and submitting all claims submission forms with just a few taps and swipes. Chatbots, another automation tool easily integrated with RPA, are already handling many routine communications tasks, including notifications of settlements and customer inquiries into claim status.
Integrating other enabling technologies: RPA will become more prevalent, especially as claims groups adopt other enabling technologies. For instance, AI-powered bots will likely handle the inputs from drones conducting standard property inspections or surveying damage after catastrophic storms. Integrating RPA with machine learning and natural language processing (NLP) can enable the initiation of new claims and issue first notice of loss (FNOL) communications by scanning and analyzing unstructured communications, including emails from agents or even voice interactions. Robots will also be used widely in the real-time review of social media streams to assess claims severity and reduce fraud. RPA will receive and route advanced telematics data (including video imagery) that will be instantaneously captured during automobile accidents and downloaded from the cloud, automatically
triggering an FNOL entry.
Suggested approach and lessons learned: following the leaders
Significant numbers of insurers are already using RPA in their claims organizations. In designing the business case for robotics, claims leaders should seek an incremental approach, adopting more ambitious use cases once they have built momentum and demonstrated results through initial and targeted deployments. With RPA, there’s no need to try do too much too fast, which may be attractive for insurance executives seeking to minimize risk and disruption in their adoption of enabling technologies. Further, an incremental approach can help organizations overcome their natural wariness toward RPA in terms of its workforce impacts.
The following lessons learned come from early adopters:
Target the opportunities: In developing a business case and tangible ROI model, specific tactical questions can lead to the right strategy as well as clarify the highest priorities for near-term automation. Finding answers may require a robust assessment of current capabilities and the completion of a cost-benefit analysis, given that the candidates for automation may number into the dozens.
Engage IT early and often: To ensure a smooth implementation and integration with other systems, there are many important infrastructure, governance and security questions to address. IT leaders reluctant to deploy another technology in the claims “stack” should consider how RPA can support strategic platform upgrades and those mandated by regulatory change. Most RPA tools are product- and platform-agnostic and work with existing IT architecture.
Find the right partner: External vendors and suppliers – including insurtechs, consultants and systems integrators – will be part of the solution, so it’s important to choose wisely. Beyond technical expertise, look for those firms with deep technical and operational claims knowledge, including a clear understanding of how it affects the customer experience.
Don’t overlook the organizational factors: As with other “digital” initiatives, claims leaders must invest time and resources in education and, if necessary, evangelization regarding the use of RPA. The delicate matter of robots taking over jobs should be addressed, most likely in the context of the need to reskill claims workers, as the role will evolve to become more analytical and more focused on customer needs and the most complex claims.
The bottom line: RPA is critical to the evolving claims process
The time for adopting robotics in claims has come, due primarily to the compelling business case and imperative for claims leaders to enhance performance and contribute more value to the business. Robotics can serve as a foundation in supporting true, end-to-end automation when integrated with other advanced technologies, such as OCR, chatbots, machine learning and NLP.
Indeed, as multiple early adopters have made clear, RPA is ready to help claims organizations advance and enhance outcomes in the digital era through increased automation, higher productivity and increased capacity and strategic focus for claims professionals.
RPA is among the top enabling technologies insurers should consider adopting in claims, as well as other parts of the organization, due to:
The path to ROI
Manageable deployment requirements
Flexible use cases
For the full report on which this article is based, click here.