Straight-through processing (STP) is becoming more common in insurance underwriting and claims, especially in personal lines, individual life and small commercial. Adoption rates in more complex lines and in claims processes remain relatively low, so STP is by no means universal across the industry. However, it’s likely to remain a priority for insurers seeking to improve the ease of doing business for their distribution partners and create more convenient customer experiences for their policyholders.

Generally speaking, STP refers to the ability for insurer systems to automatically process transactions without manual intervention or input. The system ingests data digitally and completes the transaction based on decisions governed by algorithms, including predictive models and simple business rules. STP offers insurers benefits in speed, consistency, productivity and application throughput while making the customer experience quicker and more convenient.

Technological Factors

Technological improvements over the past two decades have been a major factor in the rise of STP. The most fundamental of these has been the internet, which has both enabled the connectivity underlying STP as well as shaped consumer expectations for user experience and speed.

More recently, modern insurance core systems with the ability to automatically adjudicate applications based on configured business rules, combined with modeling capabilities that can capture underwriting factors and predict outcomes, have made it possible to process applications without human oversight.

Insurers also now have access to a wealth of high-quality, third-party data, which can enable pre-fill, eliminate unnecessary questions or inform insurers of potential risk factors. AI and machine learning capabilities to refine algorithms and flag potential fraud have also contributed to insurers’ confidence in STP.

STP in Underwriting

STP in underwriting is most common in personal lines and individual life—lines that are under cost pressures and, increasingly, sold online. More than 80% of insurers selling these lines have at least some level of automated underwriting, and many personal lines insurers process straight through more than three-quarters of the time.

Large commercial and specialty lines typically don’t have high rates of STP, because these lines are generally sold through agents and brokers. Even where insurers have supported some level of automation (for example, via portals with rating components), most policies aren’t written straight through. Instead, insurers are focusing on distribution connectivity, which is itself a prerequisite for effective STP.

See also: The Digital Journey in Commercial Lines

Small commercial and workers’ compensation lines occupy a middle point. Many of these products still require manual underwriting, but they’re also seeing increasing direct sales activity, often directed at niche market segments. Carefully defining sales targets in this way allows insurers to facilitate STP for these lines, as they can design tailored products governed by specific business rules that rule out more complex risk scenarios.

STP in Claims and Digital Claims Payments

For most insurers, though, STP in claims is fairly uncommon. Nearly 60% of insurers have no STP in this area. On average, fewer than 10% of claims are processed straight through in any line. It’s most common in personal lines and (for payouts) in annuities.

Claims STP is likely to become more common, especially in personal lines and individual life, as insurers continue to improve their core system capabilities and as the availability and quality of third-party data improves. Where coverage limits are relatively low, insurers can increase their levels of automation to create faster and more convenient claims processes. 

Insurers have achieved more substantial automation in digital claims payments. A third to half of insurers process and deliver claims payments digitally (depending on the line of business), and 10% to 20% of insurers do so most of the time. Digital payments are likely to be a priority area for insurers. Since COVID-19 forced many insurers to send some workers into the office to print and mail checks, manual and paper processes of all kinds are under intense scrutiny.

The STP “Sweet Spot”

Generally, STP is most effective when four factors apply to a particular line of business or transaction:

• Risks are well understood, which makes modeling easier
• Data is easily accessible and generally reliable
• Speed is at a premium to be competitive
• Margins are thin, so productivity and throughput drive profitability

Figuring out where to enable STP isn’t always a question of looking at specific lines or products and determining whether these factors apply. Insurers can also use these principles to design new products, especially for direct distribution—for example, by defining the allowable risk profile for a particular product more narrowly so it’s limited to the cases that are most likely to be profitable.

The Future of STP

While the industry as a whole is trending toward greater automation, most insurance will never be completely straight through; there will always be some complex claims scenarios or unusual risks that will require human intervention and review. That itself is part of STP’s value, though: When technology handles the easy processes, humans have more capacity to focus on higher-value work.

See also: Insurance Outlook for 2021

Enabling STP has an upside for those human actors, as well. Investing in better data creates resources human underwriters can use, and better connectivity eases integration and improves ease of doing business for distribution partners.

Even just the process of implementing STP can have benefits. Creating the business rule framework or algorithm to adjudicate an application—or even figuring out if a particular process can be done straight through—requires insurers to examine their workflows, understand what really matters and justify what is done and why. That can lead to process and product improvements that wouldn’t have surfaced otherwise, as legacy mindsets can hide in all kinds of places.

For more on STP, please see Novarica’s recent report, Straight-Through Processing in Underwriting and Claims.

As part of a talk given at the Verisk Elevate Virtual conference, I want to highlight why a solar storm might be a major concern for the insurance industry. The large losses that will likely occur in the Deep South are just a small taste of what could happen.

This link includes a summary of the 30-minute talk. To watch the entire recording free simply register and stream.

More on this subject is below:

2020 or 2021 for that matter showed the insurance industry that it should 100% expect the unexpected. Multiple hurricane landfalls in Louisiana and the large uncertainty with COVID-19 losses going forward were just a few examples. The year was also big for out-of-this-world events: two solar eclipses, comet NEOWISE, Jupiter's and Saturn’s great conjunction, Japan’s asteroid sample returning to Earth and, most importantly for the insurance industry, the sun's awakening with solar storms. If strong enough, the storms could lead to a large amount of uncertainty for the insurance industry.

Solar storms (sometimes referred to as coronal mass ejections (CME), solar flares, geomagnetic storms or cyclic plasma eruptions) release a torrent of charged particles arriving in just hours at the outer limits of Earth's ionosphere, where they interact with Earth's magnetic field, causing a geomagnetic storm and creating spectacular auroras for days. As beautiful as these displays are, they could also be a substantial risk. To make it more complicated, the risk is increasing as the world develops more reliance on electronic communications and satellite-powered location technologies such as GPS and a more complex energy distribution network. Even more concerning is that we have very few historical events to understand what might happen when a large solar storm hits with little notice.

Current Sunspot Cycle:

The sunspot data shown below suggests that the current sunspot cycle, widely known as solar cycle 25, is just now starting to ramp up. With the end of solar cycle 24, the sun has gone quiet. Solar cycles 23 and 24 had fewer sunspots observed compared with past cycles. The overall solar cycle is suggesting fewer sunspots, which would correlate with fewer solar storms, but it only takes one geomagnetic event to be disruptive to the insurance industry.

Let's use the example of a hurricane season: In an active hurricane season, one might expect a higher chance of named storm landfalls, and, in an inactive hurricane season, one might expect fewer landfalls. But we know that in a season like 1992, which only had seven named storms, it only took Hurricane Andrew to have a substantial impact on the insurance industry. It only takes one event.

Over the next few years, the solar cycle 25 will approach its peak, and the number of solar storms will increase, which also increases the chances of a disruptive event to the insurance industry.

See also: What the Recent Deep Freeze Portends

The Day the Sun Brought Darkness

According to NASA, five different categories of solar flares depend on their brightness in the X-ray wavelengths. For example, M-class flares can cause brief radio blackouts in the polar regions and provide occasional issues for satellites. The events that need to be watched are X-class flares. These are the most severe and can trigger radio blackouts around the whole world and long-lasting radiation storms in the upper atmosphere. On average, solar flares of this magnitude occur about 10 times a year and are more common during the solar maximum than the solar minimum. An X-class flare could produce enormous geomagnetic distortions, not only to all satellites in space but also to electrical grid networks -- especially high-voltage substations and transformers, which are very sensitive to this overload of geomagnetic energy from the sun.

In fact, the only well-known impact on our modern society came from the March9, 1989, solar storm. It caused some satellites in polar orbits to lose control for several hours. They experienced short-wave radio interference, including the jamming of radio signals. The most significant impact was related to power outages; circuit breakers tripped on the Hydro-Quebec power grid, all within 90 seconds. Millions suddenly found themselves in dark office buildings and underground pedestrian tunnels, as well as in stalled elevators. People woke up to cold homes. The blackout also closed schools and businesses.

The March 1989 event was relatively weak, but much larger events have been known to come from the sun. They are just rare, or the Earth has not been in the direct path of the most intense part of the storm. For example, while an event on July 23, 2012, did not receive much media attention, it likely could have been twice as bad as the March 1989 event and could have had similar results as the 1859 Carrington event, but it missed Earth with a margin of approximately nine days. The sun rotates around its own axis with a period of 25 days, and the section where the event occurred faced mostly away from the Earth at the time.

At some point, the Earth will be in the direct path. In 2013, Lloyd’s estimated the return interval of a Quebec event to be 50 years, with a reasonable range of 35 to 70 years. A stronger event like the Carrington event of 1859 has an estimated reasonable range of 100 to 250 years. This is similar to other extreme hazard scenarios, such as large earthquakes and explosive volcanic eruptions. However, because we do not have a model of impacts, the industry might just be surprised one day.

Since the 1989 Quebec event, a lot has changed, and this needs to be considered. The U.S. electrical grid has grown and become more complex in distribution, which is an essential part of our technology‐dependent society. The freeze in Texas showed just how complex the U.S. electrical grid can be and all the different dependences that can quickly cascade into failure. The complexities can lead to a large uncertainty about what impacts an extreme geomagnetic storm would have on modern power distribution systems.

Also since 1989, the forecast of solar storms has gotten better, and grid system operators are more aware of the problems based on forecast models. On the other hand, the evolution of open access on the transmission system has fostered the transport of large amounts of energy across the power system to maximize the economic benefit of delivering the lowest-cost energy to demand centers. This has increased the risk of potential power failures. The power disruption occurring in the Deep South also shows the dependencies on models and the fact that the only way of knowing of impacts without an event is based on models, which of course can be wrong. Therefore, given the large uncertainties in the modeling of impacts along with the complex distribution network, it is still possible to have power disruptions that may likely result in a global catastrophe that could lead to large losses to the world economies as well as unspecified insured losses.

Insurance Implications

A major solar storm that would result in the loss of power would cause a physical damage trigger. Even in a case of pure voltage collapse without equipment damage, the incapacity of the grid itself could be deemed physical damage, because it is unable to perform its essential function. Business interruption is likely to be only one aspect of potential insurance exposure. Winter property loss of power could lead to frozen pipes and heat loss to buildings that could be substantial over a much larger area. Similarly, in the summer during a heavy rain event back-up, sump/sewer issues might be compromised, leading to large claims. There would likely be a disruption to supply chains that might trigger contingent business interruption covers. Major disruption to the power network could also lead to wide-scale cancellation of events. It is conceivable that major power outages could result in liability claims. And how about companies/organizations viewed as not taking appropriate preventative action; they could be susceptible to D&O claims.

Many of the aspects listed above are similar to the lessons that we have learned from COVID-19, which were largely known but unknown risks. The same could be true for a potential solar storm impact on the insurance industry.

See also: 2021, We Can’t Wait to Get Going!

Conclusion

Economic and insured losses are on the rise, and our world economies rely more and more on satellite, GPS and electricity. Because we have limited solar storms to learn from in the recent past, the future impact is unknown and difficult to quantify. There is no model to understand the type of losses that could be likely.

If you were following solar forecasts over the last decade, you would realize there has been an over-hyping and misunderstanding of solar cycles 23 and 24. A severe solar storm is likely less probable, but at any given time a large solar flare could hit Earth regardless of the current state of the solar cycle. The solar cycle simply allows us to understand periods of increased activity. Given that we are approaching the peak of the current solar cycle, there is an increased concern for geomagnetic storms that could affect Earth’s communications, logistics and power systems, including an unspecified amount of insured loss.

An Odd ‘New’ Retronym Paul Carroll, Editor-in-Chief of ITL Retronyms have always intrigued me: those new formulations for long-used terms that arise because of some advance, usually related to technology. My beautiful old wristwatches are now “analog watches” because so many of you sport digital watches. A war used to be a war, but then the Cold War came along; now, when people shoot at each other, we call that a “hot war.” (A friend who consults with the military recently used the euphemism, “sending kinetic energy downrange,” which I love but somehow doubt will replace “hot war.”) A century ago, cars just had transmissions; now, those that require the driver to change gears are “manual transmissions.” And so on. I’m now starting to see a lot of uses of a sort of retronym that I never expected: “human.” The actual retronym is “human employee,” which is increasingly being used to distinguish those of us with flesh and blood from the artificial intelligences that are being employed in business settings. But the term almost always gets shortened to “human,” which makes the implication even starker: We’re at an interesting spot in our deployment of AI, maybe even at a tipping point.... continue reading > STRATEGIC PRIORITIES 2021 REPORT Majesco research examines the gap between Leaders and Laggards and the critical changes Leaders made to stay on top Read More SIX THINGS What Future Will We Choose? by Francis Bouchard The industry needs to stop wishing others could see the critical role we can play in preparing for climate change and just start playing that role. Read More Are Solutions in Tune With Today’s Needs? by Bernhard Klein Wassink Developing products around new customer priorities, and reaching new demographics in need, are key to keeping the industry relevant. Read More ‘An AI Walks Into an Electronics Store…’ by Marty Ellingsworth Customers may prefer interacting with a smart-bot--no judging, no fatigue, no bad days. There is empathy in any process that respects our time. Read More The Insurer’s Customer Acquisition Playbook sponsored by Data Axle The right approach to data analytics can cut wasteful spending in customer-acquisition programs while attracting high-value clients. Read More Digitally Challenged Miss Opportunities by Paul Ford Cloud-based AI can compare thousands of variables in a few hours, enhancing pricing, risk assessments and customer acquisition. Read More To Post or Not to Post? Choose Wisely by Maria Grimm Here are six tips on how agencies should use social media, to join important conversations while protecting against reputation damage. Read More Leaders Rise From a Year Like No Other by Denise Garth In the first six months of 2020, e-commerce in North America as a percentage of overall commerce increased more than in the entire previous decade. Read More MORE FROM ITL A Conversation on Corporate Strategy with Amy Radin In-depth with ITL's Thought Leaders Join ITL's editor-in-chief Paul Carroll as he sits down to discuss corporate strategy with director, advisor, author and thought leader Amy Radin. Watch Now March's Topic: Strategic Innovation   Strategy is what you don’t do. That was the dictum of the late, great Mel Bergstein, who way back in 1994 founded the pioneering digital strategy firm Diamond Management & Technology Consultants. (It became part of PwC in 2010.) I heard Mel’s line a lot, as a partner with Diamond from 1996 through 2003, and I think his are words to live by in the insurance industry these days. Everyone seems to have gotten the memo about the need to digitize insurance and to explore innovative ideas, but the present typically creates a real drag that slows movement toward the future. Take Me There GET INVOLVED Write for Us Our authors are what set Insurance Thought Leadership apart. Get Started Partner with Us We’d love to talk to you about how we can improve your marketing ROI. Learn More SPREAD THE WORD Share Share Tweet SUBSCRIBE TO SIX THINGS

By finding opportunities for digital transformation within the claims function, insurers may reduce costs and increase performance.

The reduction in profitability due to rising claims rates is creating pressure. In 2019, gross claims payments in the U.S. alone accounted for around $1.57 trillion, 5.5% more than the gross claims payment in 2018. However, price competition has made insurers reluctant to increase premiums. Insurers must instead look for opportunities to reduce indemnity spending by going after fraud cases, leakage, supplier costs and recovery performance. 

Traditionally, claims processing requires a significant amount of manual labor, complicated workflows and incomplete and disorganized data. The secret is to unlock the potential of the large quantities of unstructured data streaming through the claims function. By closely analyzing internal procedures, in-house and third-party data sources and turning raw data into usable information, insurers can produce insights that improve both claims quality and efficacy.

The approach to reducing indemnity spending can be divided into four strategic pillars focusing on data, processes, people and technology.  

Sourcing the Right Data

Claims organizations must assess the quality, accuracy and availability of their data. Through sharing the correct datasets across functions, ensuring that data definitions are consistent and resolving any specific privacy issues, organizations can create useful insights from these data.

Insurers have large volumes of internal data, including reports on customers, quotes and prices, policy specifics, claim details and past examples of fraud. This can be used to make informed, easier decisions and streamline the claims workflow. For internal data that is unstructured and resides in the form of PDFs, conversation recordings or emails, insurers need intelligent programs that leverage algorithms such as natural language processing (NLP) or optical character recognition (OCR) to convert them into usable formats. 

External data -- such as public domain information on demographics and weather -- can be sourced and linked to the claims dataset, enriching the available information and increasing decision accuracy. Third-party proprietary data sources, such as ISO ClaimSearch by Verisk, also offer specific data for sale in areas including claim analysis and fraud detection.

See also: Claims Development for COVID (Part 1)

Application of the Correct Process

Traditional claims systems focused primarily on the experience of claims handlers, with minimal use of data and analytics. With improved availability of better evidence, machine learning and AI, three types of templates can help handlers make better decisions.

• Estimation models forecast maintenance costs, treatment costs, legal costs and other fields. Insurers may use these forecasts to help track success and focus their efforts. 
• Classification models provide binary or multi-class decision flags to group similar claims. Insurers can then devise strategies for these specific groups of claims. 
• Propensity models predict the probability of an event occurring. These models typically provide a probability percentage that can be used for preparing to take the appropriate action for likely future events.

Apart from these three types of models, some improvements within the recovery process could also boost efficiency and cycle times. An optimized chase policy can be accomplished by bilateral agreements with third party insurers. In addition, changes should be made to the fraud detection process. Through applying analytics, insurers can detect subtle or non-intuitive trends, increasing accuracy and coverage and maximizing referral rates.

Engaging the Right People

The insurers with lower claims indemnity are the ones with the right resources and adequate data training infrastructure. Even if the insurers have to pay a premium, they get the best data engineers, modelers and business analysts. These insurers also have robust upskilling, cross-training and retention programs to create a multi-skilled talent pool that fuels the carrier’s data and analytics capabilities.

Technology enablement

Using the right technology enables the claims process to operate at maximum potential and generate valuable data.

• Analytical tools can improve areas including data management, model development, business intelligence, reporting and visualization. 
• Claim-specific tools can bring built-in advanced analytical models trained on proprietary third-party data, but will often lack insurer-specific information. 
• IOT devices for loss prevention and claim avoidance can reduce claims frequency and severity using real-time monitoring, through smart devices such as water damage sensors, home exterior sensors and connected cars. 

How to Optimize Indemnity Spending  

Insurers should use a project prioritization framework that accounts for competing factors when transforming claims processes. This would measure the efficiency of an initiative across several areas. For each project, a detailed analysis should be performed on the potential ROI. The model and analysis should be clear, and the ease of implementation can be calculated on the basis of complexity, the need for data assets and other variables.

See also: Surging Costs of Cyber Claims

Other considerations include the length of time it would take an intervention to deliver the desired outcome. It is also necessary to examine the scalability of an intervention, with special consideration given to the achievement of long-term objectives while accounting for short-term goals. Synergy with existing initiatives also needs to be analyzed.

Conclusion

The transformation of the claims process requires factoring in the different decision metrics, dependencies and process flows. This analysis helps to ensure that the transformation process does not harm existing claims processes. With the right data and analytical strategy in place, insurers can achieve significant compensation savings.

The critical measures in underwriting are operational exposures, hazards affecting those exposures and controls that manage the risk. Those measures, combined with an historical view of prior losses, provide the insight that underwriters need. Unfortunately, a long-standing inefficiency in the underwriting process slows engagement with customers and increases the time it takes to make decisions about adding business to fragile underwriting portfolios. 

Underwriting needs a vision to advance in an ever-changing business environment, especially in commercial lines. That vision can be partitioned into five basic considerations and exercises to create a tactical road map for moving forward.

1. What is your desired trading model? How do you wish to engage your policyholders?  

While there are numerous web- or app-based solutions in the market, the degrees of data collection and integration vary. So, a key initial step in evaluating a platform is to make sure it satisfies data extraction requirements from multiple file types. That flexibility allows the design of a fluid, low-friction process for customer engagement. 

The more effort for customers on the first touch, the lower the probability of conversion from marketing to quote to bind.

2. What underwriting data can be foreseen and what proxies exist to capture that data without requiring customers to provide it? 

Enhancing customer engagement and limiting unnecessary, onerous data-gathering tasks are critical elements of this strategic assessment phase. New technologies exist today for integrating bundles of third-party data sources — to ease data integration into legacy systems without the need for building new foundations.

Technology is pervasive in the personal lines space, but commercial lines adoption has been slow, even though far more effort is often needed to gather and assess risk data in this segment. 

See also: The Future of Underwriting

3. How can you abbreviate the supply chain to build a more responsive initial indication/pre-quote?  

This question illustrates an acute pain point for brokers and agents.

The foundation of the portfolio and pricing is predicated on a pre-formulated target set of classes that can be both underwritten and adjusted (post-claim) with the appropriate subject-matter expertise, to bind new business at a profitable rate and an expected loss ratio. A qualified decision must be made as quickly as possible — to determine if the risk profile is worth pursuing, before incurring substantial time and expense in the underwriting process.

At this stage, there is a heavy dependence on data analytics and presentation in support of decision-making.

4. Have you implemented a process to match available coverage products with the operational exposures and hazards of commercial accounts?  

A matching process of all available products is a critical differentiator for creating a competitive advantage and achieving profitable growth.  Through a matching algorithm to assess enhancements for existing coverages, as well as complementary additional coverages or endorsements, underwriters can provide more comprehensive, custom coverage to brokers and their clients. 

Insurance continues to transform into a customer-centric rather than product-centric industry, all but eliminating unnecessary coverages, terms and conditions. By definition, the industry is moving toward “pay for what you need” models. While simple and pragmatic, this concept has eluded the insurance industry for a lifetime in the commercial segment. However, the concept has been adopted in personal lines.

5. Have you built a feedback mechanism to monitor performance of the policy to take immediate and decisive action to mitigate upside and downside risks?  

While there are significant decisions to be made prior to binding the policy, there are equally critical decisions post-bind, as the business moves from the “new” book to the “renewal” book. Underwriters should be enabled to monitor immediate changes to operational exposures, hazards and risks; and they should be able to exert suitable control on the potential erosion of profits. 

As an example, the Internet of Things (IoT) continues to provide great advances in the real-time monitoring of exposures, be they man-made or inanimate. Creative manuscript endorsements and stronger terms and conditions can use the IoT to manage the risk of losses.

Transformation in underwriting is key for any insurance organization. And transformation is more possible than ever before due to the increasing impacts of modern technology — which makes it possible to get strategic insights from data for better real-time decision-making and results.

Retronyms have always intrigued me: those new formulations for long-used terms that arise because of some advance, usually related to technology. My beautiful old wristwatches are now "analog watches" because so many of you sport digital watches. A war used to be a war, but then the Cold War came along; now, when people shoot at each other, we call that a "hot war." (A friend who consults with the military recently used the euphemism, "sending kinetic energy downrange," which I love but somehow doubt will replace "hot war.") A century ago, cars just had transmissions; now, those that require the driver to change gears are "manual transmissions." And so on.

I'm now starting to see a lot of uses of a sort of retronym that I never expected: "human."

The actual retronym is "human employee," which is increasingly being used to distinguish those of us with flesh and blood from the artificial intelligences that are being employed in business settings. But the term almost always gets shortened to "human," which makes the implication even starker: We're at an interesting spot in our deployment of AI, maybe even at a tipping point.

I'm hearing the distinction between human and AI "employees" primarily among friends and former colleagues at the big consulting firms, which tend to lead the way on language in business. They don't always get the shifts right -- I'm waging a battle against "decisioning" and "solutioning" at the moment -- but I suspect the increasing need to specify when someone is talking about a human employee suggests that we're at the beginning of a great acceleration in the use of non-human ones.

For years now, I've described the likely employee of the future as a "centaur," part human and part machine (as opposed to the half-human, half-horse of mythology). I suspect that the recent need to distinguish between human and computer employees suggests that we're getting there.

The shift should help address two big issues in insurance: operational efficiency and the talent gap. The efficiency gains are obvious and are already reflected in all the interest in robotic process automation, chatbots and other forms of AI.

The effect on the talent gap should be no less significant. While people have worried for years about all the talented people set to retire soon, and about who will replace them, AI will solve much of that problem -- not by flat-out replacing those retiring but by augmenting the skills of others and helping them cover for those leaving the industry. AI will also help with recruiting talent. Think about how different the pitch to a standout recent college graduate will sound: Instead of promising to make the person an expert in sorting through the fine print of an insurance contract or in using actuarial tables, you can hold out the prospect of making her an expert in using AI.

For now, the new way of thinking will likely stay incremental. We'll still have the traditional processes for underwriting, claims, etc. We'll just increasingly have AIs slotted into the process to do some of all of a traditional task, perhaps gathering and then exchanging information with all the relevant parties. The processes will become centaurs even before the individual jobs do.

In time, processes will be reinvented in ways that move past the traditional roles, so we won't think in terms of having an AI or a centaur slotted in where a human employee is doing the work today. There will still be a blend of human and AI input, but the flow of work will be very different.

Rest assured that the humans will still hold all the key decision-making (no, not decisioning) roles, so we won't have to bow to any robot overlords any time soon.

In the meantime, we can sort through the potential need that the pandemic has created for other retronyms. A meeting used to be a meeting, but now we have Zoom meetings. If you and I sit down a few feet from each other and have coffee after we're all vaccinated, does that become an "offline" meeting, an "in-person" meeting, a "face-to-face" meeting, or what -- noting that Zoom meetings could be described as "in-person" and even "face-to-face" if we have the video on?

I'd appreciate your help solving (no, not solutioning) such questions.

Stay safe.

Paul

What Future Will We Choose?

The industry needs to stop wishing others could see the critical role we can play in preparing for climate change and just start playing that role.

Are Solutions in Tune With Today’s Needs?

Developing products around new customer priorities, and reaching new demographics in need, are key to keeping the industry relevant.

‘An AI Walks Into an Electronics Store…’

Customers may prefer interacting with a smart-bot--no judging, no fatigue, no bad days. There is empathy in any process that respects our time.

Digitally Challenged Miss Opportunities

Cloud-based AI can compare thousands of variables in a few hours, enhancing pricing, risk assessments and customer acquisition.

To Post or Not to Post? Choose Wisely

Here are six tips on how agencies should use social media, to join important conversations while protecting against reputation damage.

Leaders Rise From a Year Like No Other

In the first six months of 2020, e-commerce in North America as a percentage of overall commerce increased more than in the entire previous decade.

Event Name: The Future of Risk

Event Dates: May 18-20, 2021

Event Location: Virtual

Event Link: futureofrisk.cventevents.com

Description:

The future brought to you no matter where you are. There’s no better time than now to spend a few hours learning and preparing for the future. Join industry leaders and innovators at The Institutes’ The Future of Risk™, to learn how disruption, innovation, and technology are changing the way we work and the risks we insure. This three-day event connects you directly with thought leaders and executives who are shaping the future of risk management and insurance.
 
Throughout the three days, we’ll feature industry experts addressing some of the hot topics facing our industry today and in the years to come.

Highlights include an opening panel with industry CEOs and a risk transfer solutions panel. Session topics include:

• Customer Experience: Leveraging Data for Success
• Blockchain 2021 -Evolution To A Standard
• Embrace the Future: The Open Playing Field of Platform Ecosystems
• COVID-19 Pandemic & Personal Lines Insurance Customers

Strong demand from Americans fleeing apartments or upgrading to more space pushed home prices from an average price of $247,000 in February 2020 to $271,000 a year later, according to Zillow. Massive shifts like this present insurers with a once-in-a-generation opportunity to sell their products, and using data is one of the best ways to focus on the right targets and the right time -- as they make their purchase decisions. 

However, the movement toward data is often accompanied by an unhealthy euphoria. The solution is not simply “more data,” despite what many providers seem to be pitching.   

In an economy with 260 million smartphone users together creating billions of new data points every day, the challenge for today’s insurers isn’t buying or gathering more data; it’s making sense of existing data.

Scores of companies credibly claim to have excellent predictive data. The question is: How do you find a partner that can connect insurers with the right data and then make sense of the information, turning it into actionable campaigns that drive business outcomes. In a world with too many choices, strategic counsel is the most valuable resource. 

Ask the right questions

Selecting the right data partner is mission-critical. Below are five questions to ask a data partner before engaging:

1. Do you source data from other providers?

Reputable data purveyors will be willing to source data from other providers so they can provide clients with the best possible leads. Even if a provider doesn’t have a relationship with the owner of a particular data set, the provider should be willing to engage on your behalf to ensure a consistent data selection process that’s not redundant. There is no reason a company has to manage multiple data provider relationships itself. This is especially important for the insurance industry, where there are multiple providers of excellent “life event” data – and no single provider can claim to have a monopoly.

2. Can you provide case studies and client references that speak to the data’s performance?

Case studies are not always available – especially for fresh data that’s never been used – but they can be a helpful predictor of a data set’s value. If a case study isn’t available, ask for a client reference. In general, the more a provider is willing to share about the data they’re selling, the more confident they are in its performance. This “transparency test” is an important barometer of any potential partner.  

See also: The Right Counsel for the Right Coverage

3. What data points are you able to flag? Do you have a first mover advantage? 

The key point here is to go as deep as possible into the data you’re buying so there are no surprises after the purchase is made. The most valuable data include information that competitors haven’t purchased already. Is the provider offering a new way to reach expectant parents, for example? Ask your provider whether they’re offering anything unique. Beyond that, it’s important to understand exactly what you’re buying. For example, a provider might be trying to sell you a data file that identifies new movers, new homeowners, pre-movers and newly engaged, but you may only need  a subset of those identifiers. When you know what somebody is trying to sell and what you already have, it gives you greater bargaining power and avoids duplication after the purchase is completed. 

4. Has your data been tested – and were under-performing sources removed?

How data should be tested depends on the individual needs of your organization and what attributes are most important. The data provider should be able to explain to you how they ensure continued accuracy of their data. They also should have steps in place to validate their data sets on an continuing basis and have the necessary monitoring in place. If they know a source isn’t quality, they shouldn’t be selling it to you. 

5. Will your provider provide a timestamp for the data?

Timing is everything when it comes to data, and providers should be transparent about when the information they’re selling was collected. This is especially important when buying data for time-sensitive moments, like marketing mortgage protection insurance to a new homeowner. On the other hand, it may be wise to wait a bit to offer life insurance to a new parent instead of targeting them as soon as they have a child. A few weeks can be the difference between a campaign’s success or failure. Can your provider certify that the information has the right timestamp identifying when the event occurred – or at least when the provider first became aware of the event? 

When you have a deep understanding of what you’re buying and how it connects with your existing needs, the odds of maximizing the value of the asset after the transaction go up exponentially. Insurers should seek out partners who can do the hard work of identifying the right data – and then come to the table ready to turn that information into successful campaigns that drive sales.

A new computer algorithm has just written the world’s funniest joke.... [Punchline at the bottom.]

During the last year, rapid advances in data, AI, and cloud solutions have accelerated, and a portfolio of AI assets is making everything more DIY than ever before.

More customers are more satisfied, while expenses go down and profits go up. Data gets more accurate and more precise when it is collected right the first time, by folks with direct involvement at the right place, at the right time and in the right process.

Sure, the potential for a diabolical AI to warp the insurance value chain and create illicit data trafficking is the salacious fodder of tabloids of the techno-Luddites. But, name-calling aside, letting people interact directly in a protected fashion with self-service channels in a personalized fashion has only seemed to widen hours of availability, lower operating expenses, increase transparency, improve communication, decrease cycle time, eliminate travel, enable social distancing, establish authentication and verification of identities and data, create faster payments and improve satisfaction.

It seems that the more tasks customers can do themselves, the less they must wait for expensive, slow, serial, synchronous, manual and human-error-prone, analog processes to do for them.

Same for agents, underwriters, employees, claims handlers, vendors and executives. How else could we all be functioning after a year of working at home?

We all still like a timely, helpful, empathetic and friendly voice when we want one (or even a live-video channel), but many find that an avatar or smart-bot can in many ways be e-better. No judging, no fatigue, no “bad days,” just the facts. There is implicit empathy in any process that respects our time.

As for ethics, it’s not even a required course in the MBA curriculum, but, because these robot processes establish a permanent digital record, it is easier than ever to audit them and not devolve to a “he/she/they said contest.” Even a gender-neutral machine can repeat the facts on record from the DIY process. Got a problem? E-litigate it (if you can’t settle with escalation). Facts/data can change, and adding them to the process should influence the outcome as appropriate. Misrepresent the facts, and that is on permanent record, too.

We have not yet seen “pick the robot voice and language you prefer,” but that technology to remove the potential for implicit bias of a gendered voice is on the horizon.

Facts can be recorded, with an ever more attractive set of options for responding, such as:

• “Would you please review the pre-filled data and say 'OK' to continue?"
• "Say 'yes' to opt in your data from your smart-car or smart-phone to complete your submission”
• “Use the hotlink just texted to your smartphone and take a picture
  • for a claim we will ask for a few pictures,
  • for your odometer-based billing, please also type in what you see on the picture as digital authorization for immediate processing,
  • for underwriting, please follow the prompts for your car, home and valuables."
• "Say, 'I have completed the task' as your voice signature, or simply push the green 'agree/OK' button."

A digital transcription of the session will be available on your app, in your portal and via e-mail, and your integrated communication will be available for agents, adjusters, repairers/vendors, underwriters, customer service, billing, etc.

See also: Despite COVID, Tech Investment Continues

When it works well, we are empowered.

When we need to exit to a human, we want the connection now, and we want the human to start where the machine left off. The machines don’t get frustrated, but customers do, and we hate repeating ourselves after creating a permanent record already. Our time is valuable. Simply notify us when you are ready to talk?

The only thing for sure – “easier, faster, more trustable, cheaper, convenient and more delightful” is a one-way street for customers.

They are voting with their feet for personalized risk rating and automated channels, especially during low-mileage driving and long hold times common during COVID.

People really are counting their mileage for usage-based pricing, but proven claim-free drivers may not need their every movement traced to just get miles verified -- we already can use third-party data and smart car data; why is that not already DIY, too? Get the picture, and you will get the joke.

The punchline “…but the algorithm's joke is only machine-readable.“ [Machine laughter: "Harharhar...."]

What happens when a key data source becomes less available, reducing carriers’ ability to evaluate risk? This has happened during the pandemic in workers’ compensation.

In workers’ compensation, OSHA is one of the top data sources that underwriters use. In particular, underwriters will look at OSHA inspections and violations to measure some aspects of the risk. 

Here at Convr, our focus has been to help carriers with the right insights at the right time for better decision-making, and we found, using a vast data pool, that planned inspections dropped 48% in 2020.

One reason is fewer claims; as operational capacity was reduced or suspended for many industries in 2020, workers’ compensation claims dropped by over 20%. As accidents declined, inspections that normally would have followed weren't needed. In addition, OSHA reduced the number of planned inspections for the safety of their inspectors.

The reduction in inspections has led to a lack of reliable information for workers’ compensation carriers to evaluate businesses -- but this is where technology comes in. With the help of artificial intelligence and advanced analytics, carriers can still determine the risk of a business by looking at past patterns.

These past patterns include types of structured and unstructured information that data scientists refer to as “features” in machine learning models. Often, significant features are high-dimensional nonlinear combinations of company and property characteristics, such as the size of the business, the year it was established and prior violations. Other features include social media information and product and services data.

See also: 9 Months on: COVID and Workers’ Comp

Applying AI to our data lake, which is informed by over 2,000 data sources, Convr has determined that, in place of the normal volume of OSHA inspections, carriers can use a workplace safety model to accurately quantify risk. A workplace safety model consists of a machine learning model that predicts how safe a workplace will be based on OSHA data and the different data sources mentioned above.

Companies labeled as the riskiest 10% by Convr’s proprietary workplace safety risk scoring model observed three times as many future violations as those labeled as the median risk.

COVID-19 has proven that circumstances can change unexpectedly, and carriers have to become adaptable and flexible enough to implement alternative solutions to minimize the impact. Advanced AI models, like the one Convr has created to quantify workplace safety, hold tremendous value for carriers, enabling them to better understand risks even when traditional sources of information are limited or unavailable.

When armed with technological advancements such as these, carriers are equipped with the right tools for better decision-making and optimal underwriting results.