Beyond Human: AI's Impact on Insurance Underwriting

The insurance industry is undergoing a seismic shift. No longer confined to actuarial tables and manual processes, it's embracing a future driven by data and artificial intelligence. Your next underwriting colleague might be an AI agent, capable of outperforming human counterparts in speed, accuracy, and efficiency.

AI Agents: The New Underwriting Powerhouse

AI agents are transforming the underwriting process in several key areas:

• Enhanced Risk Assessment: Leveraging advanced algorithms, AI agents can analyze vast datasets, including historical claims data, weather patterns, and emerging risks, to provide more accurate risk assessments. This empowers insurers to price policies precisely and competitively.
• Fraud Detection: AI's ability to identify patterns and anomalies in claims data makes it a formidable tool in combating insurance fraud. By detecting suspicious activities early on, insurers can protect their bottom line and maintain customer trust.
• Streamlined Operations: Routine tasks like data entry, policy generation, and document verification can be efficiently handled by AI agents, freeing up underwriters to focus on complex cases and strategic initiatives.
• Personalized Customer Experience: AI-powered chatbots and virtual assistants can provide instant responses to customer inquiries, improving satisfaction and loyalty.
• Product Innovation: By analyzing customer data and market trends, AI can help develop innovative insurance products tailored to specific customer segments.

The Future of Underwriting: Humans and AI Collaboration

While AI brings immense potential, human expertise remains indispensable. The future of underwriting lies in a collaborative model where humans and AI complement each other's strengths. Underwriters will leverage AI insights to make informed decisions, build strong customer relationships, and drive business growth.

Challenges for Insurance Leaders

Embracing AI in underwriting comes with its own set of challenges:

• Data Quality: Ensuring the accuracy and completeness of data is crucial for AI models to deliver reliable results.
• Model Explainability: Understanding how AI arrives at its conclusions is essential for building trust and complying with regulations.
• Talent Acquisition and Development: Building a workforce with the skills to effectively collaborate with AI requires investment in training and development.
• Ethical Considerations: Implementing AI in a way that is fair, unbiased, and transparent is paramount to maintaining trust with customers and regulators.
• Cybersecurity: Protecting sensitive customer data from cyber threats is a top priority in an AI-driven environment.

By addressing these challenges and embracing AI, insurance leaders can position their organizations for long-term success. The future of underwriting is bright, and those who harness the power of AI will be at the forefront of industry innovation.

Are you ready to transform your underwriting operations with AI? [Include a call-to-action, such as inviting readers to attend a webinar or download a whitepaper on AI in underwriting]

The Let’s Talk Payments article, How APIs Are Obliterating Technology Challenges, suggests the phenomenal increase in APIs is because we are “used to getting what we want – instantly – and, whether you know it or not, you have APIs to thank for that.“

IBM recently published the whitepaper, Identifying API use cases: Banking industry, that summarizes the need for an API economy as “a business API extends an enterprise and opens new markets, application developers can easily leverage, publicize and aggregate a company’s assets for broad-based consumption.”

Thanks to the iPhone, Uber, and AirBnb, consumers today are conditioned for, and are demanding, frictionless access to all the tasks they perform. They want a user experience that is designed, simple, easy to use, and innovative. Both FinTechs and financial institutions are striving to provide consumers with what they want. But better tools are needed to help them create apps that solve real problems and drive business value with applications, infrastructure, and services.

Recently at IBM InterConnect in Las Vegas, IBM announced IBM Cloud for Financial Services.  It’s a platform that provides developers with the tools they need for agility and speed in the development, deployment, and testing of apps that tap into the FinTech ecosystem.

For a financial services developer to build a next generation app, IBM Cloud for Financial Services is the one place they can come to develop and test their apps, run their applications with high security and reliability. And it’s a place where large institutions and FinTechs startups alike can monetize their offerings.

IBM envisions a vibrant marketplace where financial institutions and FinTech startups can buy and sell different digitized offerings. These offerings can be APIs, datasets, perhaps chain code that runs on HyperLedger, trained neuro networks, and more. The sky’s the limit and the time is right to push boundaries of what’s been tried before.

It wraps all the elements of a successful platform that is specialized for financial services.

In addition to brand new APIs from IBM, IBM Cloud for Financial Services includes apps from partners including Xignite, Accern, Plaid, FinLeap, Ripple, and Monetise, making it possible to mix and match. Starter kits are available to make it easy for a programmer to get going quickly. A starter kit is a bundle of APIs and datasets that have been handpicked and put together in a single, easily consumable bundle.

One of the new APIs from IBM is Cognitive Market Sensing. This uses powerful quantitative algorithms from the IBM Algorithmic product in the form of smaller, consumable APIs. It allows modern apps to contain elements of formerly monolithic applications and combine them with offerings from one of the FinTech partners.

APIs from IBM’s Apple Apps team are included to help with mobile front-end functionality, as well as Watson cognitive APIs around natural language processing, parsing sentiment in news articles, and Watson Conversation for enabling chatbots.

Even if you have never done anything with IBM before, you will be able to go from zero to code in 15 minutes or less. That includes signing up for a free account on Bluemix, getting access to a marketplace and starter kits, building the code, deploying it, and launching it.

All too often, at the pace today’s market moves and the scale on which market leaders operate, we forget that frequently all it takes to get ahead is just a minor shift in tactics or perspective. For example:

Facing down increasing competition and growingly oversaturated neighborhood markets, big-box retailer Target didn’t open more super-sized stores in suburban locales. Instead, the company launched pint-sized City Target stores in fast-growing (and fast-moving) cities that sold household items in smaller sizes—and equipped salespeople with mobile scanners to make it easy for customers to enjoy storewide checkouts. The concept proved so successful that Target quickly doubled the number of these locations nationwide. Target also dropped the “City” branding, making these half-sized locations a standard part of its core retail strategy.

When medical device leader Medtronic wanted to expand its already successful business throughout Western Europe and beyond? It didn’t double down on cutting-edge devices. Medtronic reinvented its business model instead, expanding its offerings to include clinical and consulting services and establishing new business units that partnered to put owned-and-operated labs inside hospitals. Not only has Medtronic increased its business, but it has provided partners with significant improvements in customer service and cost-savings by doing so. Having earned their trust, the company has also built a sizable business around ancillary services such as supply chain management and performance benchmarking.

After sales started falling on drugstore chain Walgreens, the company didn’t bump up ad spending or start offering more paid incentives. Instead, it asked its executives to rethink the chain’s approach to customer care. Strategies included bringing pharmacists out from behind the counter to offer counseling services and building API middleware tools that let third-party developers create hundreds of apps that let patients order from its 8000+ pharmacies. Sales rapidly recovered.

When French telecom giant Orange wanted to double the size of its innovation initiatives but didn’t want to invest millions in R&D or hordes of high-priced working professionals? It decided to outsource the entire process, and offered APIs—plug-and-play back-end software solutions—both to internal employees and external developers so that they could create new uses for Orange’s technologies. Using just one of these solutions, the company has been able to seamlessly integrate social and second-screen experiences from hundreds of film and TV companies into many of its services in under a year.

When Newell Rubbermaid’s Contigo brand wanted to find a way to differentiate its products in the hugely-crowded and -contested market for portable containers and cups? It didn’t invest a fortune into dozens of abortive product roll-outs, attempting to guess what working professionals on the go would want. It simply studied today’s busiest travel sites, where commuters tended to congregate and, after discovering that passengers were constantly wiping off their mugs’ mouth guards on napkins, sleeves, and handkerchiefs, Contigo introduced a new line of travel mugs with special covers designed to keep out dirt.

And when MasterCard needed a new idea for a mobile payment app? It simply put the call out to employees at Innovation Express, a global series of hackathon events where businesspeople, designers and software developers team up to create new business plans and products in record time. Two days later, Qkr, which can let you order food from your seat at a stadium or preorder school lunches for children right from your pocket without ever setting foot in a cafeteria, was born.

While it’s not always obvious to the casual observer, innovation is far easier than you think. All it takes to successfully outmaneuver the competition or overcome a problem is simply a greater sense of perspective and greater willingness to be more creative with how you apply the tools at hand.

More than 100,000 individual and enterprise developers from the financial services industry already turn to IBM monthly for insight and support. Now these and more will have access to APIs, data, and content to ­­­­­­­­­­­­build and monetize cognitive-enabled financial services apps quickly and at scale for FinTechs, banks, wealth management firms and insurance companies.

Initially available in beta, developers will also have the ability to build in customer insights, regulatory compliance analytics, security, privacy and compliance readiness to help reduce the time needed for development and testing. The new tools can simplify the time-consuming tasks of selection, mapping and data integration, allowing developers to use IBM services or combine them with their own data. ""We have extensive expertise in financial services with 97 percent of the world's largest banks relying on IBM. At the same time, IBM has become the leading AI platform for business,"" said Bridget van Kralingen, Senior Vice President, Industry Platforms. ""Our experience across financial services and in the commercial deployment of AI with enterprise and start up builders has informed these new tools. We're excited to see how they put them to work to quickly create the latest solutions powered by cognitive computing and blockchain for everything from new payments directives support, and even regtech solutions.""

With the support of the Monetary Authority of Singapore (MAS), the central bank of Singapore, IBM also developed the Singapore FinTech Hub, which is linked to IBM's financial services developer cloud. The Hub will serve as a catalyst to support Singapore FinTechs by providing information about local policies and developments in the ecosystem, access to cognitive and other APIs, insight to accelerate innovation, and a platform to showcase made-in-Singapore solutions to the world. ""The launch of the hub will help accelerate the growth and influence of Singapore-based FinTech companies to be part of the worldwide ecosystem of API providers. Their solutions and APIs will become available to the global developer community, expanding their market reach beyond Singapore,"" said Sopnendu Mohanty, Chief FinTech Officer, MAS.

IBM Bolsters Partner Ecosystem for Financial Services
IBM is extending its current Business Partner program by providing additional benefits and further support for partners aligned with the financial services industry. Initial FinTechs offering their financial services focused APIs on the financial services developer cloud include Accern, Actiance, Bondevalue, Dow Jones, Eigencat, Envestnet|Yodlee, Opentopic, Plaid, PolicyPal, Quovo, Riskspan, TagniFi and Xignite. ""Partnering with IBM brings us one step closer toward our goal of building an inclusive financial services ecosystem, giving industry institutions and their customers access to critical infrastructure by marrying IBM's institutional knowledge with Plaid's technical expertise,"" said Sima Gandhi, Head of Business Development and Strategy at Plaid. As clients place increasing value on industry-specific as-a-service solutions, ISVs play an important role as they collaborate on the IBM Cloud to deliver cloud native applications and cloud-enable existing applications. These companies are working with IBM to help extend their data, microservices and content to reach more clients in the industry. Advisory services help ISVs enhance their intellectual property. Technical and sales support, reference architectures and support for developing and using APIs also help IBM partners transform the industry. ISVs participating in the program and building and deploying innovative solutions on Bluemix include Avoka, Majesco, TCS BaNCS and others.

IBM Business Partner Solution Hubs such as those in London, New York, San Francisco, Singapore, Dallas, Paris, Nice, Silicon Valley, Stuttgart, Sao Paulo, Bangalore, and Shanghai will provide direct access to IBM expertise to help partners commercialize financial services solutions with cognitive and cloud technology from IBM. The hubs bring together the best in IBM Design Thinking, Lean Startup, Agile and Extreme Programming to help transform the banking, capital, and insurance markets. This will enable IBM clients, partners and developers to come together to build new relationships with innovators and foster co-creation.

Developers and partners can access the IBM Cloud for Financial Services at https://developer.ibm.com/finance.

About IBM Watson Financial Services
IBM is working with organizations across the financial services industry to use IBM Cloud, cognitive, regtech and blockchain technology to address their business challenges. Banking, wealth management and insurance are some of the areas poised for dramatic change by using cognitive and AI capabilities provided by IBM Watson Financial Services.

For more information about IBM Watson Financial Services, visit https://www.ibm.com/watson/financial-services/.

InterConnect is IBM's cloud and cognitive conference where more than 20,000 developers, clients and partners are being introduced to the latest advancements in cloud computing through 2,000 sessions, labs and certifications. IBM is positioning both enterprise and startup clients for success with a complete portfolio of cloud services and marquee partnerships, supporting a wide range of applications including: big data, analytics, blockchain and cognitive computing. For more information, visit: https://www.ibm.com/cloud-computing/. Engage in the conversation through @IBMCloud and #ibminterconnect.

Sixty-four percent of banks have reported lost deals and revenue due to problems with their onboarding.[1] The top 100 institutions have a 25%-40% rate of attrition, losing $400 in revenue with each customer[2].

The bank with the best app wins

Today’s customers expect banks to have fast, reliable, and secure services. They chose institutions that offer the best apps and most convenient banking.

But many financial institutions fail when it comes to speed and convenience.

For example, customers may apply for loans online, believing that the online application will be quick and easy. However, customers don’t know that complicated manual processes are hiding behind the “quick and easy” application.

Since the web application doesn’t communicate with the bank’s other back-end systems, customers may need to wait days/weeks for someone to manually review their loan. It’s no surprise that the abandonment rate for new account applications can be as high as 90%.[3]

It’s not just about collecting data … it’s about taking action based on your insights

To retain customers, banks must automate their back-end functions. Automation improves your onboarding processes, allowing you to provide customers with faster, more convenient services.

The most successful banks are implementing solutions that allow them to achieve straight-through processing (STP) on new accounts. This reduces onboarding delays and allows your bank to be more responsive. By providing better, faster services, you can reduce attrition and increase the lifetime value of each customer. Automation also simplifies your back end functions – reducing the costs of processing new customers.

Get a 360° view of your operations with process intelligence

To automate your back end operations, you first need to see what’s going on. You can gain a clear picture of your operations through process intelligence (PI).

PI collects data on your business processes and workflows, so you can identify the weak spots in your onboarding and customer engagement.

For example, PI can show you which customers are diverging from your onboarding path and where they are getting lost. This allows you to correct the process, so you can delight and retain your customers. With PI, you can also:

• View your onboarding process from start to finish – such as from the moment a customer first applies for a loan until they are approved or denied
• Examine the flow of applications through the entire work process
• Streamline your processes by removing unnecessary steps
• Discover exceptional pathways
• Leverage your existing systems to simplify your onboarding processes and give customers a better experience
• Determine which offers are successful and which ones aren’t
• Compare region to region
• Improve your compliance and reduce your risks

And the benefits of PI don’t end after you’ve successfully onboard a customer. PI analytics continue to work behind the scenes – collecting data that will help you improve even more of your processes. While most analytics tell you what happened in the past, PI gives you visibility into what is currently happening, so you can take action in near-real time.

Here are just a few of examples of how banks leveraged PI

PI gives you a clear picture of your processes, so you can streamline your onboarding and retain more customers.

Whether you believe it’s one of these or something else altogether, it’s clear that the industry is not performing as well as it – or its investors – would like.

Sustained low interest rates and the ongoing operational costs – in particular the cost of compliance – are compressing profit margins.  In response, most banks are cutting costs and expenses aggressively.  But they’re doing it tactically, knowing all the while they cannot save their way to prosperity.

At the same time, commoditization is hitting both commercial and retail banks.  Revenue and profits are stagnating while investment in fintechs is surging.

These findings are part of a new IBM Institute for Business Value (IBV) comprehensive study that will be released on September 20, the week before Sibos begins in Geneva.  The study, called “The Cognitive Bank,” is based on interviewing and surveying more than 2,000 banking and financial industry C-suite executives across five continents.

The cognitive bank is built on cognitive computing, which IBM popularized in 2011 when its Watson computer defeated the top human contestants on the televised quiz show, Jeopardy.  (To learn about Watson, click here.)

A cognitive computing system is based on four key principles:

• It learns and improves with each new piece of information
• It builds speed and scale by using machine learning to handle complex, repetitive tasks
• It uses subject matter experts to collate and curate human intelligence, for rapid reuse or decision support
• It interacts using natural language, context and reason

Cognitive computing – applied to banking – solves a lot of problems.

At the highest level, it allows banks to switch from tactical cost-cutting to begin operating at what I like to call a strategically lower cost.

Perhaps most importantly, cognitive computing finally will help banks exploit their biggest advantage against upstart competitors: data.  Banks are awash in data but are only just beginning to use it and recognize its power to drive a competitive edge.

Imagine a system that understands your entire global client base individually, has complete knowledge of both existing and proposed banking regulations across continents, countries, states and provinces, and keeps perfect records for 100 percent compliance.  Almost sounds too good to be true, doesn’t it?

Cognitive computing also can help expand a bank’s reach.  The difficulty in extending wealth management services for only the ultra-rich out to the mass affluent typically has been one of scale.  In India alone there probably are a quarter of a million people who would enjoy wealth management services, but don’t qualify for financial advisors.

To address this, the cognitive bank can use machine learning to review millions of data points, including the client’s threshold for risk, investment horizon and other unique attributes to create a personalized portfolio recommendation and other automated services.  Potentially, it could even use the client’s public social media record to learn more about his or her personal preferences and experiences.

This is what I mean by operating at a strategically lower cost.  And it’s a much needed influx of innovation.

Ultimately, the cognitive bank is about generating wealth, providing liquidity and protecting assets.

Not surprisingly, the study finds that the highest performing banks already are on their journey towards becoming cognitive banks.  Yet only about a quarter of bankers are familiar with cognitive banking and fewer than one in five believe their organization is prepared to embrace it.

My team and I look forward to seeing you at Sibos to talk about the new IBV study, how the cognitive bank is being deployed and how you can prepare for it.  To learn about more about IBM activities at Sibos, visit ibm.com/Sibos.

In line with these trends, financial institutions are seeing strong growth in consumer adoption of online and mobile banking products and services. 41% of consumers use mobile banking (including 71% of millennials). Customers’ expectations are also growing. Over 34% of consumers indicated that they would be very or extremely likely to use TouchID and 22% currently use Mobile Check Deposit (Raddon Research Insights, 2016).

As institutions quickly adopt digital channel marketing, they often overlook some key factors. Institutions who are considering spending more time and energy on digital marketing this year, may find the below list helpful.

1.  Marketing emails and websites are not mobile-optimized

Over 50% of emails are read on mobile devices, and this percentage increases each year (see growth chart on Litmus.com). The best practice is to create mobile-friendly or responsive email templates that will resize or show different content when opened on mobile devices. It used to be an after-thought to even review and test the mobile design. With the rise of “mobile-first” mentality, companies are beginning to create for the smaller screen first – both for websites and emails.

Mobile and tablet internet usage now exceeds desktop usage worldwide. In 2015, Google started ranking sites based on “mobile-friendliness”.  And last year, Google started to experiment with making search indexing mobile-first – meaning that their algorithms will pick up content from your mobile site instead of the desktop version. This could be an issue if your mobile site has less content or is strictly built around online banking.

These changes along with the fact that consumers are using the internet as their primary source for financial services product information, make it more important than ever to evaluate the mobile user experience. Take a look at how your current website, mobile app(s) and email creative are viewed and function on mobile.

2. Not using targeting and segmentation

Another common mistake is forgetting to use targeting and segmentation in your digital marketing campaigns. Why would you need to target specific customers with a product offer? Why not send it to everyone since the email channel is inexpensive compared to direct mail? Even when using digital channels to reach your customers, you should segment your customer base, determine the best message to send to each person or household and determine which channels they prefer. You don’t want to waste a “touch” by sending a blast marketing email with the same message to everyone.  Sending a message not tailored to a user is a quick path to being unsubscribed, meaning you might never email that customer again.

Another concern is that many financial institutions have email addresses for fewer than half of their customers. Your institution should have a plan in place to reach these customers through other channels and develop campaigns to get them to opt-in for marketing emails.

3. Not looking at the complete picture: the conversion funnel

If consumers see your online advertising, will they be able to open a new account online? Your website needs to be follow best practices for Search Engine Optimization (SEO) so that consumers searching for your brand can find you and the information they are looking for. You may need to send traffic to landing pages that you test and optimize to get the most conversion. If you don’t have an online application, or if the one you do have has a very low completion rate, then you should have the phone number and local branch address front and center on the landing page along with any additional product or rate information they may need.

If you are advertising home loans online, consumers should be able to reach a landing page or product page that has home loan information and a strong call-to-action. It’s also beneficial to have interactive calculators and rate tables on this page.

4. Tracking and reporting are not in place

When you launch campaigns in multiple digital channels, you run into the same measurement complexities that you have when running TV, radio and billboard advertising. How do you know which channels are most effective and which ones are working together to drive the best results? The first step is to enable web analytics on your site such as Google Analytics or Omniture/Adobe Analytics. Each digital marketing channel will have its own cost and response/conversion metrics that you need to establish upfront to bolster your reporting.

For your email campaigns, you should be able to access reports with opens, clicks, and unsubscribe rates by campaign. But you will also want to look at your email member base over time to see who is opening and who is inactive in this channel. You can hold out control groups (customers whom you pull out from receiving marketing emails for a certain period of time) to measure against or you can take the entire mail list and run a match-back process to see who responded during a 30-day window. How many times have you seen a preview of an email, not opened it, but visited the website or store within the next week because it was top of mind?

5. Lack of social followers

You post to Facebook, Instagram, Twitter and your blog on a regular basis. How many people are seeing these posts? How many followers do you have? You may need to build your follower base before you can depend on getting reach through this channel. More importantly, monitoring and responding in these channels to customer services concerns or consumers asking general questions may be of benefit to both you and your customers

Start by promoting your social properties to customers. During the onboarding process, you can send them a series of welcome emails with links to find/follow you on social media and include content from your blog or snippets from your posts. Include social links and call-to-actions in your email communications. Drive readers from your e-newsletters to your blog. It’s helpful to have a content calendar in place – this can drive your social posts, blog articles and e-newsletter content.

We hope these pointers on how to avoid common mistakes in digital marketing will help you as you set your 2017 marketing initiatives. Raddon Advisory Consulting offers marketing strategy engagements and can help support your digital marketing efforts. If you would like a digital channel assessment or help with your digital marketing strategy, please send an email to comments@raddon.com

Interested in Leaning More? Join our Raddon Strategic Consultants on April 25, 2017, for a no-cost webinar on Uncovering Secrets to Successful Marketing. This webinar will be useful for institutions considering spending more time and energy on digital and direct marketing this year and may need to refine their efforts to achieve stronger results.