In brief
- “Insurtech” refers to the technologies used by insurance companies across their operations, from distribution and underwriting to claims processing and investigation. Adopting insurtech enables insurers to implement automation, leverage data analytics, accelerate and streamline processes and increase personalization
- Automated rule-based fraud detection is already standard practice. And ML-powered advanced data analytics recognizes patterns that rule-based systems ignore. However, data protection and privacy remain vital concerns when using advanced analytics and ML for fraud detection
- Telematics provides vehicle use and driver behavior data to improve risk assessment for underwriting. But car makers are pushing technology to the limit. Modern vehicles advise the owner if they’re unlocked. How long before claims are refused if the policyholder ignores car warnings? And if your car reports that you’ve switched off the collision warning system, will that invalidate your policy?
The past decade has seen tremendous insurtech innovation, from Lemonade’s two-second claim processing to new products like usage-based insurance. Insurtech funding worldwide continues to rise quarter-on-quarter, in spite of deal size reaching a six-year low in Q3 2023 .
Insurtech companies that target end consumers may look like a threat to established insurers. However, innovation forces incumbents to rethink their business models, upgrade customer experiences and transform processes for maximum efficiency.
What is insurtech?
Insurtech combines insurance and technology. The word refers to the sector, company type or the use of technology in insurance in general:
- The insurtech sector is a part of the insurance industry, comprising insurtech companies and their customers
- Insurtech companies and startups leverage cutting-edge technology to launch innovative products destined for use by industry actors or consumers, e.g., Lemonade, Next Insurance, Clearcover, GoHealth
- Insurtech in general concerns using technologies to optimize costs, create new business models and products, improve customer experience and increase carrier efficiency and resilience
I’ll be using the word insurtech in its general form for this article.
Why does insurtech matter?
With the insurtech market valued at $5.45 billion in 2022 and expected to double by 2030, insurtech innovation is impossible to overlook. The main reasons why traditional insurance companies must consider insurtech are as follows:
- Better customer experience. Insurtech-driven customer experiences are digital-first, omnichannel and more convenient, boosting engagement, loyalty and lifetime value
- Enhanced efficiency and cost savings. Insurtech reinvents workflows and processes, automating time-consuming, routine tasks and streamlining others. This improves efficiency and optimizes running costs
- Personal touch at scale. Data analytics, personalization at scale and 360-degree-customer-view tools enable insurance carriers to gain deep insights into each customer
- Data-driven insights. Data analytics provides companies with a better understanding of their customers, enabling data-driven business decision-making
- Improved agility. Insurtech is a prerequisite for flexible, highly customizable insurance offerings, allowing companies to better adapt to customer needs
- Mitigated risks. Insurtech delivers more accurate risk-based underwriting and more precise fraud detection to minimize risks
Five challenges insurtech can address
Undetected fraud
According to the Coalition Against Insurance Fraud, insurance fraud costs U.S. consumers at least $308.6 billion a year. In addition, 78% of U.S. consumers report that they‘re concerned about insurance fraud.
As keeping up with fraudsters’ methods is the number one challenge in fighting fraud, insurtech can help level the playing field. However, while identity theft is a concern, most fraudulent claims concern actual incidents with exaggerated costs.
Automated rule-based fraud detection is already standard practice. However, advanced data analytics powered by machine learning can recognize patterns that rule-based systems can’t take into account. Data protection and privacy (the number two challenge) remain key concerns when using advanced analytics and ML for fraud detection.
Complicated claim processing
Lengthy claim management workflows lead to below-par customer experiences, overblown operational costs and unnecessarily slow results. For example, 60% of car insurance companies have a resolution path with over 10 steps.
Insurtech simplifies processes and reduces manual workloads in several ways, decreasing the risk of human error. For example:
- Optical character recognition (OCR) can automatically extract data from scanned documents
- Machine learning can be used to automate claim processing, fraud detection, self-service customer support and more
- Robotic process automation (RPA) enables employees without coding skills to create and run automations
Long-winded customer experiences
When an insured incident occurs, friction points — from a lack of self-service options to having to repeatedly contact the insurer — deteriorate the customer experience. On top of that, 70% of customers want their user journey to be 100% digital or hybrid.
Insurtech has the power to reinvent customer journeys. Take claims management for example. Insurtech can automate:
- Data verification by comparing data from multiple sources
- Fraud detection by scanning for red flags
- Compensation calculation by following predetermined rules
Straightforward claims can be processed with little human involvement, as in Lemonade’s two-second claim settlement. However, auto-rejecting claims based on the verdict supplied by insurtech solutions can open the company to legal action (e.g., Cigna).
Slow underwriting
The days when an underwriter had to manually request, enter, organize, verify and analyze customer data from multiple sources are long gone.
Today, data points can be collected via mining or automation without human involvement. Data comparison and verification can also be automated. Risk assessment can be streamlined with advanced analytics algorithms that evaluate the individual’s risk profile and determine the most suitable pricing and insurance coverage.
This can take underwriting from a protracted process into an immediate one. Because Lemonade’s chatbot includes underwriting algorithms in its code, it can insure customers instantly.
Inflexible pricing
At a time when natural disasters intensified by climate change are compromising existing insurance business models and risk management is critical for survival, one-size-fits-all pricing is no longer efficient.
Hence the rise of highly flexible products like pay-as-you-drive, pay-how-you-drive and usage-based car insurance (e.g., Geico DriveEasy, Progressive Snapshot). These insurance offerings use telematics to monitor driver behavior in real-time, estimate associated risks and adjust pricing.
On the other hand, data analytics helps health insurers make sense of a sea of data, tailoring premiums and deductibles according to the patient’s requested coverage and medical history.
Finally, insurtech can power pick-and-choose plans where consumers customize services, coverage, length, deductibles and more to arrive at a plan best suited to their needs.
How insurtech is transforming the insurance industry
Artificial intelligence (AI) and machine learning (ML)
AI and ML power predictive analytics, intelligent automation, NLP, OCR and generative AI and are transforming the insurance industry in the following ways:
- Fraud detection: Identifying irregularities in submitted claims and flagging them as suspicious for further review by a claim manager (e.g., Quantexa)
- Risk management: Assessing the individual’s risks and selecting the appropriate plan with minimal human involvement to accelerate underwriting (e.g., Planck)
- Demand forecasting: Predicting product demand and premium volumes for sales and distribution teams
- Claim processing: Handling FNOL via a chatbot (e.g., Lemonade) and automating processing, adjudication and reporting
As AI becomes increasingly entrenched in cross-industry company operations, regulators are setting the framework for its use. The European Union is the frontrunner in this regard, with an AI Act in development.
RPA
Automation today comes in two main flavors: Robotic process automation (RPA) and intelligent automation.
RPA is software that allows users to create bots, with little or no coding required, to interact with applications and perform repetitive tasks like a user. RPA handles simple actions based on predetermined rules (e.g., UiPath and Flobotics).
Like AI and ML, RPA transforms insurance company operations from A to Z. For example:
- First notice of loss (FNOL): Automating data extraction from scanned documents (in combination with optical character recognition), verifying data via cross-referencing
- Claims management: Aggregating data from multiple sources, including the claim itself, to verify the data, calculate compensation and pay claims in simple cases
- Underwriting: Automatically pulling data from multiple sources and cross-referencing it for verification purposes, calculating policy rates in simple cases
- Policy management: Generating contracts, automatically updating policies, notifying customers about changes
Intelligent automation
While RPA tools are potent automation drivers, they have one significant downside. They can’t handle exceptions to the predetermined rules.
Leveraging AI, ML and RPA, intelligent automation (IA) solutions learn from data, make predictions, work with unstructured data, understand natural language, generate responses and improve over time.
Intelligent automation expands automation in the following ways:
- Claims management. Intelligent automation analyzes and triages claim complexity. Simple claims are automatically analyzed, scanned for signs of fraud and resolved
- Customer service. Intelligent automation automatically triages opened tickets based on priority. IA tools like Clarifai can also recognize the intent within text data via NLP
- Underwriting. IA tools identify and extract data, analyzing it for risk. Clarifai can scan for red flags in text data, including information from social media, news outlets and medical and police records. In contrast, Smart Submission extracts unstructured submission data from emails, etc.
Big data
Big data refers to data sets that are too big for traditional data-processing software to handle and analyze. Making sense of massive data sets requires big data tools that identify patterns and dependencies.
Big data comes from data mining and can be structured or unstructured. In the insurance industry, big data examples include:
- Telematics. Real-time driver behavior data can inform policy pricing and coverage, powering new product types like usage-based insurance. Combined with predictive analytics, vehicle data can also be used to notify policyholders when preventative maintenance is necessary
- Personalization. The massive amounts of customer data can be used to determine the likelihood of policyholders ending their contracts or which products and services they’re most interested in. These insights enhance marketing efficiency, customer retention and policyholder satisfaction
- Process optimization. Leveraging previously collected data can streamline specific processes. For example, applications and notices can be prefilled using available data
China’s largest online insurance company, ZhongAn, leverages big data to offer dynamic pricing, risk tracking and exceptional customer experiences.
Chatbots and virtual assistants
Chatbots and virtual assistants have been around for quite some time. Lemonade opted for AI-powered chatbots for claim processing back in 2017.
Chatbots and virtual assistants can:
- Fully automate the FNOL process by replacing a claims form with a chatbot or virtual assistant (e.g., Lemonade already accepts insurance claims via its chatbot, while Hi Marley provides this functionality for insurers out of the box)
- Resolve common customer queries, reducing the strain on customer service representatives (self-service options can reduce calls by 12%, on average)
- Personalize customer onboarding with virtual tours and AI-powered chatbot capabilities (e.g., HDFC ERGO, an India-based, non-life insurer, leverages AI for this purpose)
GenAI
GenAI has the potential to elevate chatbots and virtual assistants to a new level, but that’s not its only application. It can also boost employee productivity.
This type of AI refers to technology that generates or manipulates text, images, or videos that read or look realistic. Large language models (LLMs) are a type of GenAI designed for generating text in natural language. Think ChatGPT and Google’s Gemini .
GenAI can transform claims management, reducing claims payouts by 3%-4% and reducing loss-adjustment expenses by 20%.
While its use in a new iteration of chatbots and virtual assistants is more or less inevitable, GenAI tools also improve employee productivity by automating tasks like summarizing, translating and generating and correcting text, images and codes. For example, AXA is rolling out one such tool to its employees.
Computer vision
Computer vision is the technology that leverages ML and deep learning to allow programs to analyze images and videos.
In an industrial setting, a computer vision solution automatically detects product flaws on the manufacturing line. In insurance, computer vision can:
- Estimate damages. Solutions like Tractable estimate vehicle and home repair costs by analyzing the level of damage done to the property
- Extract more data for claim processing. Computer vision solutions can analyze dashcam footage to identify weather conditions and even potholes
- Detect fraud more accurately. Automated damage assessment can be instrumental in catching claims with overestimated repair costs
Data analytics
Data analytics analyzes and derives insights from swaths of data. While not necessarily implying ML, this technology paves the way for advanced data analytics, such as predictive and prescriptive analytics.
Several concepts already described above are impossible without data analytics:
- CX personalization
- Pricing personalization
- Telematics for usage-based insurance and other innovative products
- Intelligent automation
- Fraud detection
Also, data analytics can help established insurers:
- Refine the risk assessment process with geospatial and weather analytics (e.g., CAPE Analytics, Betterview)
- Enable straight-through processing (STP) for specific types of claims
- Automatically triage claims
- Micro-segment customers to improve marketing and sales efforts
- Automatically determine the most suitable inspection strategy and method for every claim
- Identify bottlenecks and inefficiencies in internal processes
Most traditional insurers are in the “Adopter” or “Explorer” stages of the data analytics journey, with only 6% of surveyed companies rating themselves as “Pioneers.”
Blockchain
Blockchain is a distributed ledger technology mostly known for powering cryptocurrencies like Bitcoin and Ethereum. However, that’s not why insurance companies are interested in it.
Blockchain stores records securely and is immutable. That’s why enterprises beyond insurance, such as IBM, are experimenting with the technology.
In addition, blockchain powers smart contracts, transaction protocols that automate contract execution when specific conditions are met. In insurance, smart contracts can automate damage payouts when the policyholder’s region suffers a natural disaster. For instance, Chainlink pulls data to trigger smart contract execution.
The Internet of Things (IoT) and telematics
IoT refers to devices with sensors that collect and process environmental and hardware data. According to NTT Data, IoT is one of the technologies with the highest growth in the insurtech sector.
Telematics, in its narrowest sense, refers to IoT systems within vehicles. Vehicle telematics enable integrated driving assistance systems and location tracking, for example.
For insurance, telematics provides data on vehicle use and driver behavior to improve risk assessment for underwriting. It’s the technology that drives usage-based insurance.
But vehicle manufacturers are pushing technology to the limit. Modern cars advise the owner if they’re unlocked. How long before claims are refused if the policyholder ignores car warnings? And if your vehicle reports that you’ve switched off the collision warning system, will that invalidate your policy?
But IoT applications aren’t limited to new vehicle insurance offerings:
- Homeowner insurance. Moisture sensors and thermal cameras can detect mold in walls and alert homeowners about a potentially leaky pipe, averting damage. Smart smoke and fire alarms can automatically alert the fire department if no one is home. This is linked to automatically closing fire doors
- Health insurance. Wearable devices that constitute medical IoT can supply data for remote patient monitoring, assessment and risk management. This data can also be used to incentivize policyholders to counter health risks by leading an active lifestyle, like Vitality in 2023
As IoT deals with vast amounts of potentially sensitive data, privacy and cybersecurity should be primary considerations.
Drones
According to Federal Aviation Administration data, the insurance industry already accounted for 17% of commercial drone use in 2018. Today, drones are used in both pre- and post-loss workflows.
Before the claim is ever filed, drones can be used to assess the property features and risks for underwriting purposes. Once the insured event occurs, drones can facilitate the investigation process by:
- Providing an aerial view of the damaged property. No need for employees to endure hazardous conditions
- Speeding up inspections
- Making the adjudication process more objective and preventing potential fraud with footage and photos that can also be analyzed by computer vision software
- Enabling generalists to perform field assessment, reducing the overall costs
What does the future hold for the insurance industry?
Insurtech will continue disrupting the legacy-driven insurance industry. However, insurtech companies face challenges of their own, from changing investor priorities in light of economic uncertainty to climate change demanding higher adaptability and better risk management.
Besides the technologies described above, APIs will continue to power embedded insurance, a smart distribution model with a $3 trillion market potential. GenAI will find both internal and external applications growing.
Migration to public clouds is also on the books for most insurance leaders, according to a McKinsey survey. Adopting public, private or hybrid cloud is critical for leveraging technologies like predictive analytics, computer vision, NLP, IoT and chatbots.
Many insurtech companies will continue to power new products and offerings, from small business insurance (e.g., Pie Insurance) to pet insurance (e.g., Trupanion).
Frequently asked questions
What does insurtech stand for?
Insurtech stands for “insurance and technology”. The term refers to the range of technologies used by insurance companies across their operations, from distribution and underwriting to claims processing and investigation.
What is an example of insurtech?
Here are three examples of insurance technology:
- The Internet of Things and telematics power usage-based insurance
- Predictive analytics streamlines risk assessment during underwriting
- Data analytics algorithms scan and flag suspicious claims to prevent fraud
Insurtech companies that provide insurance products for customers include Lemonade, ZhongAn and Hippo. Bold Penguin, Shift and Chain Link are B2B insurtech companies.
How does insurtech differ from traditional insurance?
Traditional insurance is legacy-driven and paper-reliant, which results in longer customer journeys for everything from getting a quote and underwriting to settling a claim.
Leaders in insurtech adoption leverage automation, data analytics and other technologies to accelerate processes, make them more efficient and personalize customer experiences.
Which business models do insurtech companies use?
An insurtech company typically relies on minimal overhead and maximum operational efficiency to generate revenue. The exact business model depends on the company’s goals, offerings, target market and audience. Some of the new business models include customer-centric insurance, Insurance as-a-Service and insurance ecosystem orchestrators.
In conclusion
Insurance companies revamping offerings and workflows must set clear goals.
So, if greater efficiency is your ultimate goal, then IA and RPA demand close consideration. Alternatively, CX personalization needs predictive data analytics.
Find out more
Sharpening your competitive edge in a digital-first world is impossible without insurtech. And leveraging cutting-edge technologies requires deep domain and technical expertise.
If you’d like to learn more about how Zoreza Global and insurtech could help your company become more competitive and shape the future of the industry, contact us.