Location, location, location – what value does it bring?

One of the key pieces of information in Property & Casualty insurance (P&C) is location. This applies both to pricing and risk assessment, at the level of a single insurance policy as well as the insurer’s entire portfolio. Due to reinsurance capacity and capital requirements, ensuring the right data quality and continuously monitoring exposure have also become extremely important.

Modern technologies increasingly enable insurers to collect larger amounts of information, automate processes and manage their portfolios more efficiently. Among them are technologies and systems that use broadly defined geospatial information – data linked to geographical coordinates such as longitude and latitude.

The key condition that must be met before an insurer can start using this technology is assigning geographical coordinates to the portfolio of insured locations. This can be challenging, as the quality of data for each location must be ensured. Individual address elements, such as street name, building and apartment number, postal code and locality, should be stored as separate fields in the database.

Data quality can also be significantly improved by introducing dictionaries or controlled vocabularies in the source system where address data is entered, helping to ensure accuracy and consistency. Such solutions may seem obvious because we encounter them almost every day, but this has not always been the case, and they are not always easy to implement.

This is an unavoidable challenge for insurers in markets where long-term agreements are signed or tacit renewal applies. If the policyholder does not terminate the contract, it is continued for another insurance period, usually one year. From a technical perspective, this is often handled by copying data from the previous period or extending the insurance period within the existing policy. As a result, data and its structure may remain unchanged for many years, and address data quality may require automatic or manual correction.

To fully benefit from the potential of geospatial technologies, insurers need to ensure sufficient geolocation accuracy. This depends on both the quality of address data and the coverage provided by the geolocation service provider in a given area.

In line with data quality expectations, such a service should locate as many addresses as possible at the “rooftop” level of a particular building. With lower geolocation accuracy, for example only to the middle of a street, locality or postal code area, the results of the analysis and the benefits of using geospatial data may be significantly affected.

Using GIS in insurance portfolio analysis

Once the insurer has data enriched with geographical coordinates, there are many ways to use it. GIS (Geographical Information System) software makes it possible to analyse and visualise insurer-owned data such as premiums, sums insured, the number of insured items or claims amounts, for example in the form of a heat map.

One example is portfolio monitoring: verifying the total value of insured objects, meaning the sum of sums insured, within a specific area or distance from one another. Each insurer has reinsurance protection within a certain limit. When signing individual insurance agreements, the insurer verifies whether the sums insured exceed that limit.

Without GIS technology, verifying the entire portfolio and the dependencies between individual agreements is very difficult and imprecise. The lack of such analysis, or an error in it, may be very costly for the insurer.

If an insurer covers several clients within the same location and the total sum insured across all agreements exceeds the reinsurance capacity, the problem may only become apparent once claims are reported. In the worst-case scenario, if total losses occur, the reinsurance protection may prove insufficient and the insurer may have to cover a larger share of the losses than planned. This could even affect the insurer’s solvency.

Another use case is monitoring trends in the number of claims reported within a given area. Detecting an increased number of claims may require increasing the number of experts or establishing cooperation with a new contractor, such as a car repair shop or medical facility.

A spike in claims in motor insurance, for example, may also indicate fraudulent activity in a particular area.

Geospatial data in catastrophe and claims management

Crisis situations such as catastrophic events are another area where geographical data can be applied. With information about the location of a flood, hurricane or earthquake, insurers can calculate the number of customers exposed to potential losses and estimate their value.

This makes it possible to plan for a larger team servicing the helpline or deploy more appraisers in the affected area. In other words, insurers can maintain operational capacity and avoid the negative consequences of being unprepared for unexpected events.

An insufficient number of claims handlers may negatively affect the Net Promoter Score (NPS) of affected customers, which in turn can weaken customer trust and future sales. Without enough experts, claims handling takes longer, exposing the insurer to additional costs related to increasing damage or potential fees and interest caused by delays in claim settlement.

Geospatial data in underwriting and pricing

In addition to analysing geographical data, GIS technologies and tools make it possible to create custom data layers on maps. These layers connect areas that share a common feature or value and can be used in underwriting or pricing during the sales process.

Insurers most commonly use layers related to natural hazards such as floods, hurricanes or earthquakes, often prepared by external partners. There may be one or multiple layers for a particular type of risk.

For floods, one layer may show areas exposed to an event occurring once every 100 years, while another may show areas exposed to a flood occurring once every 50 years. Other recurrence intervals can also be used. For hurricanes, layers may be created based on different wind speeds.

When an insured location overlaps with a particular layer, the insurer may calculate a specific premium or use the information in business rules. For example, the terms of insurance may change or an individual assessment by an underwriter may be required.

Layers can also be used for other purposes, such as generating leads for new customers based on areas assigned to particular agents, or directing health insurance customers to the nearest medical facilities.

External data sources and automation

More and more data resources can now be treated as geospatial and used as either a primary or supplementary source for analysis. They can also support processes typical for insurance companies.

One example is cadastral data, which may include information about land parcels, the number of buildings, their surface area or structure. Such data can be used in property insurance sales. On the one hand, it simplifies the process by reducing the amount of information the customer has to provide. On the other hand, it enriches risk analysis and can improve the accuracy of premium calculation.

Weather station databases are another example of data that can support claims handling. Based on this information, insurers can automatically verify whether rain or wind conditions capable of causing damage occurred in the area where a claim was reported.

Based on such data, an insurer may decide to route verified cases through a simplified and faster fast-track claims process or even fully automate the process, significantly reducing processing time.

There are, however, limitations when using external databases. Access to specific types of data may vary from market to market. Data available in Germany, for example, may not be available in the UK or Poland.

GIS technology can also help in situations where the required data is not available in any existing dataset by enabling insurers to create it themselves.

Satellite imagery and image analysis

High-resolution imagery is widely available commercially from several providers. The frequency with which satellites revisit the same point on Earth has also increased significantly.

This creates opportunities to develop advanced satellite image analysis algorithms and extract data that was previously unavailable. There may be many potential applications for such image analysis, helping insurers reduce costs, increase sales or improve NPS.

One example is claims processing after catastrophic events such as floods or hurricanes. With up-to-date images of an affected area, it may be possible to identify damaged buildings and, in some cases, estimate the scale of the damage.

A reported claim can then be confirmed quickly, while compensation can potentially be estimated and paid without direct human involvement. Going one step further, an insurer could proactively contact the insured person with information about potential damage and offer compensation or additional assistance, such as cleaning or transport services.

For more complex losses involving corporate clients, image analysis can also support loss mitigation.

Assuming that a production hall was flooded, information about water depth obtained from recent satellite images could help identify which machines or components were damaged. Obtaining this information early, before the water subsides and the site can be inspected in person, may allow replacement parts to be ordered sooner, shorten recovery time and reduce business interruption losses.

For corporate clients, this type of technology may also support risk assessment.

Supporting underwriting with image analysis

The amount of information analysed by underwriters in corporate insurance is vast and often unstructured or even inaccessible. This may include the number of buildings, their structure, technical condition and other characteristics.

Such information can be generated through image analysis. Satellite images can also provide information about hazards in the surrounding area or the way flammable materials are stored near buildings.

Access to this data may significantly simplify and automate the underwriting process and reduce the need to involve loss control engineers in on-site assessments compared with remote assessments performed from the office.

The time saved can be used to evaluate more customers, improve risk selection and, as a result, increase the profitability and quality of the portfolio.

A practical example

Sollers Consulting had an opportunity to cooperate with a customer that implemented this type of technology. It was used to collect information about the size of insured agricultural buildings and their spatial location on the insured plot, making them easier to identify in the event of damage.

Before introducing this solution, measurements and farm plans were prepared by insurance agents. As a result, the sales process was time-consuming and complicated.

Automating the process enabled agents to focus on sales activities that generate additional value in the form of premiums. From the insured customer’s perspective, the process of signing an agreement became simpler and no longer required manual measurements, positively affecting the customer journey and the overall perception of the insurer.

The future of geospatial technologies in insurance

Geospatial technology is developing dynamically, and Sollers Consulting’s experience shows that such solutions are being implemented more and more frequently.

The impact on the insurance sector will be particularly visible in the simplification and automation of processes. The amount of location-related information that customers or claimants need to provide will be reduced to a minimum.

More processes supported by data obtained through geospatial technologies will become STP (straight-through processing) processes, where direct contact between an insurer’s employee and a customer is no longer necessary. This may be particularly visible in personal lines insurance.

The potential impact of the technologies discussed above can be illustrated with the example of insuring a house. In the future, taking out a policy may require little more than providing personal details and the property address.

The address can be geocoded, while geographical coordinates allow additional information to be collected from external databases and images. The property’s surface area can be calculated, as can its value based on local real estate prices.

The technical condition and surroundings of the building may also support the assessment of the insured property and its contents. A doghouse visible in an image could even lead to an insurance offer including pet coverage.

The risk of natural hazards can be estimated based on maps showing recurrence intervals of such events. The premium for theft insurance may be determined based on local burglary statistics.

If a hurricane damages the roof, the insurer could inform the customer about a proposed compensation amount, provide details of local repair companies and suggest nearby accommodation for the duration of the repairs.

Geospatial technologies have been presented here mainly through examples from non-life insurance, where they are likely to have the greatest impact. Nevertheless, technology used to analyse satellite images can also be applied to traditional photographs, including those taken with a smartphone. This opens further possibilities for practical applications, for example in motor claims handling.

Author: Jakub Młodzikowski, Senior Consultant at Sollers Consulting