Image analytics: A competitive approach for property insurers
Property image data has become an important source for insurance carriers.
Property assessment is a costly affair and may pose inaccuracy risks to insurers.
High-resolution aerial imagery can reveal the underlying risks of the properties and add to the details in application data for underwriting purposes. In fact, the global geospatial imagery analytics market size is expected to grow from $6.9 billion in 2020 to $27.9 billion in 2025, at a Compound Annual Growth Rate (CAGR) of 32.1% during 2020-2025.
Currently in the P&C industry, there is an upsurge in imagery consumption by insurance carriers as the upgradation of camera technology that has allowed drones and satellites to capture high-resolution images of properties. This advancement, combined with cutting-edge research in artificial intelligence algorithms, has opened up avenues for property insurers to leverage image analytics as a means to property underwriting and as a first filter to fraud detection in a way that not only saves cost but also increases efficiency and reduces turnaround times.
The pace of change has been augmented by the industry need for quick decision making, driving down the costs and optimizing the customer experience.
Is this real?
Imagery based data is being used to detect pools and trampolines that in turn signal an increase in exposure to liability. Similarly, dense vegetation in the surrounding area indicated in the image-based data can be sign of high wildfire risk.
The AI technology that can capture the image pixel data and convert that data into useful information already exists. Increased adoption among property insurers will give underwriters an edge not only for underwriting but also for claim processing and triaging, risk control, and auditing, and reinsurance.
With high resolution aerial imagery, some insurers also are being able to get a clear assessment of a property’s roof condition. In addition, potential hazards can be captured by seeing a property’s condition change over time. For instance, image data about overhanging trees or nearby structures that are prone to fall, as revealed by the aerial imagery analysis, can be used to assess the probability of costly damage caused to attachments such as solar panels, and thus assist in claims processing and settlement during CAT events.
There are similar success stories in the risk control and audit. Consider the assessment of new risk factors, which is a laborious and time-intensive activity, and potential repercussions on premiums. Leveraging high-resolution aerial imagery can unlock value and set up new ways of doing business.
Profitable growth through effective underwriting
It is always a tedious task to accurately estimate the roof and building conditions easily by using the owner’s data alone. Underwriters need a clear picture of the property’s condition while writing policies and deciding the premium structure. With the increasing competition, they need to do it more quickly and accurately.
Property and casualty insurance is slowly adopting remote property intelligence as a fast, reliable, and cost-effective alternative to relying on brokers who provide estimates based on incomplete data. COVID-19 lockdowns and corresponding physical-distancing protocols have double-downed the need to rethink underwriting.
Traditionally, physical surveys and audits have been the go-to approaches. These tools result in high expenses for the carrier and also lead to high turnaround time. Moreover, there is no way to verify the information’s accuracy.
Carriers need to access various details of property at every stage of the acquisition process, including information that is dimensional, geographical, situational, structural and even historical. Image-based information can help identify the most profitable customers. What’s more, advanced imagery-based machine learning algorithms make it possible to extract data such as square footage and many other features that can quickly help target personal or commercial property owners who are more likely to purchase the insurance.
Claim processing: Speed, accuracy and fraud detection
Image analytics is helping some carriers accurately assess the extent of damage to a property and thus aid in claim processing while simultaneously reducing claim turnaround time. It follows that P&C carriers are engaged in a number of claims processing optimization projects involving such variables as debris, holes in the building/roof, patches on the top, staining, change in dimensions of the building and roof, surrounding changes such as sea level, vegetation coverage, and fence present or not post the natural disaster came out to be useful to detect the extent of the damage.
Insurers are likely to increase the use of aerial imagery and advanced analytics to take advantage of the latest in artificial intelligence and machine learning advancements to allow for automated damage classification by combining this with location information such as an address and other information like the extent of the damage.
Pre and post-disaster imagery also will provide Special Investigation Units (SIU) with the intelligence needed to expose insurance fraud. Likewise, oblique aerial imagery has been beneficial in the remote inspection of all sides of the property and in determining with confidence whether reported damage occurred before or after an insurable event.
In dealing with claim surge during CAT events, high resolution aerial imagery is helpful in highlighting potentially fraudulent claims. Although capacity building for processing claims is still going to be a challenge, leveraging high resolution imagery can expedite claims processing and reduce the processing cost by at least 20%.
Risk control and audit transformation
The data captured by artificial intelligence algorithms from the high-resolution cameras can support ongoing audits and evaluate whether there is a risk change. Comparing property data between different time points can lead insurance companies to assess the difference in risk.
For example, a carrier leveraging imagery data in the audit process observes a property with new floors and a parking lot added. This necessitates a premium change for this property. Similarly, risk control audits leveraging imagery data can track specific changes such as addition or removal of chimneys, fences, air conditions or solar panels as well as changes in the surrounding areas such as removal or addition of fire stations or vegetation. This information is useful for carriers for premium readjustment, or any changes in policy terms are needed.
What’s in it for carriers?
Image analytics offer distinct capability to insurance carriers to add new property condition elements to use at point of sale and renewal. This offers a way for the carriers to expand and strengthen the underwriting process. This certainly adds to the top line of the carrier’s property book of business.
In CAT as well as non-CAT event claims, image analytics are already helping risk monitoring and risk management. In the current times, claims cost control is a competitive edge from the carriers’ bottom line perspective. Image-based detection of truthfulness and degree of damage to the property is going to increasingly help carriers. Such data together with policy information can help predict the claim count and dollar amount across geographies of interest.
Property image data has become an important data source for the carriers that can help them grow and compete in the marketplace. Image based data as well as models are becoming an important aspect to improve customer service, pricing, reserving and risk management.
As the science of image-based data and models is leaping forward, it is becoming a mainstream insurance tool for controlling costs and underwriting efficacy.
Siddharth Garg (siddharth.garg@exlservice.com ): Siddharth has excellence in transforming data to knowledge. With nearly 8 years of experience spanning multiple ladders of analytics ranging from basic reporting to advanced Data Science, he has worked extensively with P&C and HealthCare Insurance leaders to design and implement data based solutions to solve multiple business problems. Siddharth holds a B.tech degree from IIT Kanpur and MBA from ISB Hyderabad
Upendra Belhe (upendra@belheanalytics.com ): Dr. Upendra Belhe is the President of Belhe Analytics Advisory, helping businesses achieve business outcomes through data-driven insights. He serves as a Strategic Advisor to EXL Service.
See also: