In conversation with Guidewire Analytics: From systems of record to systems of insight
Can you give us some background on Guidewire Analytics’ genesis?
The mission of Guidewire Analytics group is to make the world more insurable. We do this by enabling Guidewire customers to leverage data from their Guidewire core policy, claims and billing centres as well as third-party data. This provides them with insights to understand, price and manage risk from underwriting to claims settlement. With these insights, insurers gain experience underwriting new lines of business with potentially more preferable terms for the policyholder.
What does your role as General Manager and SVP entail?
I am responsible for growing Guidewire Analytics’ existing business which includes those which have been acquired. Acquired businesses are Predict, an advanced P&C-specific machine-learning tool that allows insurers to make data-driven decisions with predictive analytics; HazardHub, a comprehensive P&C risk data set that equips insurers with more than 1,000 property risk factors; and Cyence which stores cyber threat data for insurers to model the likelihood of cyberattacks at both individual and portfolio levels. I also manage the growth of businesses developed in-house, theses include Canvas, a claims analytics tool that helps insurers prepare for and manage catastrophes; Explore a business operations solution for both policy and claims; and Compare, a tool that enables insurers to monitor their claims performance and also compare it against benchmarked peers. My role includes supporting the development of new products to add to the Guidewire Analytics portfolio and evaluating potential partnerships
How do insurers and MGAs use Guidewire Analytics’ products?
Insurers and MGAs can use Guidewire Analytics’ products throughout the policy lifecycle. These products can be used in tandem with Guidewire’s core systems or integrated into external workflows. For example, the Predict product, which was formed following the acquisition of EagleEye, is used by customers to build and customise underwriting models using their own data as well as third-party data. These models, and their assumptions, are then integrated into existing workflows. Insurers and MGAs are using these models to inform underwriting and claims decisions.
Another product, Cyence, is used for cyber underwriting and portfolio management. Guidewire recognised that insurers and MGAs find it challenging to quantify cyber risk due to unknown perils and the risks associated with unfamiliar territory. However, cyber threats represent a huge risk and opportunity to the insurance market. The 2019 Worldwide Threat Assessment Report cited cyber as the number one threat to USA security, above nuclear weapons. The decision to acquire Cyence has allowed insurers and MGAs to become familiar with pricing cyber risk and enabled us to enter this growing market. Cyence customers have access to cyber threat data from over 400 sources and 1,000 data points, allowing them to model the likelihood of cyberattacks at an individual and portfolio level.
Where does the opportunity for brokers lie by using Guidewire Analytics?
I predict that the role of the broker will increase in 2023 as new and existing policyholders seek to secure economically favourable policies against rising premiums. Brokers are using products such as Cyence and HazardHub to build models on behalf of their customers to glean a better understanding of policyholders’ risks and suggest lower-priced policies or methods of reducing premiums.
Why did Guidewire Analytics acquire HazardHub?
Guidewire acquired HazardHub in August of 2021 to enhance its own property and casualty offering to insurers, using HazardHub’s geographic information systems. Through HazardHub’s integration with PolicyCenter, underwriters can now access HazardHub’s scores and models without needing to visit another website or key multiple API calls. The data is also automatically updated and is available to insurers in the instance of a mid-term adjustment or claim. HazardHub’s risk data set features perils such as fire, wind and manmade hazards as well as area characteristics and business operations. In California insurers are now able to determine wildfire risk down to individual parts of ZIP codes, previously they were only able to determine wildfire risk at an entire ZIP code level.
How does Guidewire Analytics select its partners and acquisition targets?
Different relationships will work for different vendors and Guidewire is open to collaboration with companies that operate in similar spaces. When evaluating technology vendors, it is important to discuss a company’s areas of focus, customer feedback and future product development plans. For companies that align closely to Guidewire Analytics’ mission or solve an existing customer problem, we work with them to create a joint workflow. An example of this is Guidewire Analytics’ relationship with Shift Technology and FRISS to create fraud detection solutions which can be integrated into Guidewire’s PolicyCenter and ClaimCenter.
What are some success stories you can share from insurers using Guidewire Analytics’ applications?
Ventus Risk, a subsidiary of Arch Insurance North America, approached Guidewire Analytics to resolve the challenge of differentiating and selecting risk between commercial properties. The insurer, was in search of additional information about individual commercial buildings needed at the underwriting stage. Guidewire Analytics worked with Ventus to build an API that feeds into HazardHub and prefills applications with property data to facilitate enriched underwriting at scale.
Why has the insurance industry been slow to embrace new analytics solutions?
Most areas of the insurance value chain already use analytics in some way, whether that is insights in how to market products or price risk. Adoption of new analytics technologies has been slower than in other industries such as fintech. However, the insurance industry is right to be cautious with new analytics as its primary role is to transfer and safeguard risk. Analytics software not fit for purpose can lead to insurers failing to understand the risk they have underwritten or how it might change over time.
Guidewire’s whitepaper, Building Trust in Analytics, highlights the risk of insurers moving too quickly. Our whitepaper suggests that the insurance industry should assess whether analytics lead to unfair outcomes by reviewing these outcomes through an actuarial, societal and business lens to draw their own line between accuracy and fairness.
What does the future hold for insurance analytics?
Given financial constraints in 2023, the investment arms of insurance companies are more likely to be biased towards solutions with a proven track record and strong return on investment. Insurers need to ask themselves what it is they want to change about the nature of their business and selectively target analytics vendors that can facilitate these changes.
What emerging technologies do you think will improve the quality of data analytics?
Terms such as AI and machine learning are powerful tools but are oversaturated and have led to desensitisation amongst insurers, MGAs and brokers. If a problem can be solved with a cheap and simple solution it will work equally as well, in the eyes of the customer, as one which uses expensive AI and machine learning. Businesses that create a tool using technologies that are fit for purpose and simple to execute will be differentiated and succeed in the next few years.
What recent product updates has Guidewire Analytics launched?
Guidewire Analytics has expanded HazardHub’s geographic reach to provide scores and variables across all Canadian provinces. We are keen to further HazardHub’s capabilities and have updated the product’s fire and wildfire models. These models have been calibrated with billions of dollars in real-world claims to enhance their accuracy. In the longer term, we will also launched our Tune capability within the Predict product which will enable rate scenario testing and integration.
If you would like to learn more about Guidewire Analytics’ products or would like to discuss the topic of building trust in analytics, please contact Roger Arnemann at [email protected].